THE TREES OF GREAT BRITAIN & IRELAND - VOL. IV

VOLUME IVThe TreesofGreat BritainIrelandBYHenry John Elwes, F.R.S,ANDAugustine Henry, M.A.Edinburgh: Privately Printed*

THE TREES OF GREAT BRITAIN AND IRELAND

'.f

NATIVK SCOTS PINF. AT INVERGARRVft cm a Drawing by Jlfiss Ruth Brand

MfM.4'*•»••fl*Great BritainIrelandBYHenry John Elwes, F.R.S.ANDAugustine Henry, M.A.VOLUME IVEdinburgh: Privately PrintedOyMCMIXw!k,|/ X'I,

CONTENTSLIST OF ILLUSTRATIONS ....ARIES ......ABIES PECTINATA, COMMON SILVER FIRABIES PINSAPO, SPANISH FIRABIES NUMIDICA, ALGERIAN FIRABIES CEPHAI.ONICA, GREEK FIRABIES CILICICA, CILICIAN FIRABIES NORDMANNIANA, CAUCASIAN FIRABIES WEBBIANA, HIMALAYAN FIR .ABIES PINDROW, PINDROW FIRABIES SIBIRICA, SIBERIAN FIRABIES SACHALINENSIS, SAGHALIEN FIRABIES FIRMA, JAPANESE FIRABIES HOMOLEPIS .....ABIES BRACHYPHYLLA, NlKKO FlR .ABIES UMBELLATA .....ABIES VEITCHII, VEITCH'S FIRABIES MARIESII, MARIES' FIRABIES GRANDIS, GIANT FIR ....AWES CONCOLOR, COLORADO FIRABIES LOWIANA, CALIFORNIAN FIR .ABIKS AMABILIS, LOVELY FIRABIES NOBILIS, NOBLE FIR ....ABIES MAGNIFICA, RED FIR, SHASTA FIRABIES BRACTEATA, BRISTLE-CONE FlRABIES LASIOCARPA, ROCKY MOUNTAIN FIR .ABIES BALSAMEA, BALSAM FIRABIES FRASERI, SOUTHERN BALSAM FIRABIES RELIGIOSA, MEXICAN FIRPSEUDOTSUGA .....PSEUDOTSUGA DOUGLASII, DOUGLAS FlRCASTANEA ......CASTANEA SATIVA, SPANISH OR SWEET CHESTNUT .Hi71372073 273773974474675°755758760762764765768768771773777779782786792796800803806808811814837839^——————ÔN1VE,LIERAR

The Trees of Great Britain and Ireland Contents VMCASTANEA CRENATA, JAPANESE CHESTNUTCASTANEA DENTATA, AMERICAN CHESTNUT .CASTANEA PUMILA, CHINQUAPINFRAXINUS ......FRAXINUS EXCELSIOR, COMMON ASHFRAXINUS ANGUSTIFOLIA, NARROW-LEAVED ASHFRAXINUS OXYCARPA ....FRAXINUS SYRIACA .....FRAXINUS ELONZA .....FRAXINUS WILLDENOWIANA ....FRAXINUS DIMORPHA ....FRAXINUS XANTHOXYLOIDES ....FRAXINUS POTAMOPHILA ....FRAXINUS RAIBOCARPA ....FRAXINUS HOLOTRICHA ....FRAXINUS ORNUS, FLOWERING ASH, MANNA ASH .FRAXINUS FLORIBUNDA ....FRAXINUS BUNGEANA ....FRAXINUS MARIESII .....FRAXINUS RHYNCHOPHYLLA ....FRAXINUS MANDSHURICA ....FRAXINUS CHINENSIS ....FRAXINUS OEOVATA ....FRAXINUS PUEINERVIS ....FRAXINUS SPAETHIANAFRAXINUS LONGICUSPIS ....FRAXINUS NIGRA, BLACK ASHFRAXINUS ANOMALA, UTAH ASHFRAXINUS QUADRANGULATA, BLUE ASHFRAXINUS AMERICANA, WHITE ASH .FRAXINUS TEXENSIS, TEXAN ASHFRAXINUS BILTMOREANA, BILTMORE ASHFRAXINUS LANCEOLATA, GREEN ASHFRAXINUS PENNSYLVANIA, RED ASHFRAXINUS OREGONA, OREGON ASH .FRAXINUS CAROLINIANA, SWAMP ASHFRAXINUS VELUTINAZELKOVA ......ZELKOVA CRENATA . . . ...ZELKOVA ACUMINATA ....8548568578598648?98828838838848848858858868878878908918928928938958958968978978989009009019 5906907908910912912914920CELTIS .....CELTIS AUSTRALIS, NETTLE TREE .CELTIS CAUCASICA ....CELTIS GLABRATA ....CELTIS DAVIDIANA ....CELTIS OCCIDENTALS, HACKP.ERRV .CELTIS CRASSIFOLIA, HACKBERRYCELTIS MISSISSIPPIENSISALNUS .....ALNUS GLUTINOSA, COMMON ALDERALNUS INCANA, GREY ALDERALNUS CORDATA, ITALIAN ALDERALNUS SUBCORDATA, CAUCASIAN ALDERALNUS FIRMA ....ALNUS JAPONICA, JAPANESE ALDER .ALNUS NITIDA, HIMALAYAN ALDER .ALNUS MARITIMA ....ALNUS RUERA, OREGON ALDERALNUS TENUIFOLIA ....ALNUS RHOMEIFOLIABETULA .....BETULA PUBESCENS, COMMON BIRCHBETULA VERRUCOSA, SILVER BIRCH .BETULA DAVURICA ....BETULA CORYLIFOLIABETULA MAXIMOWICZIIBETULA ERMANI ....BETULA ULMIFOLIA ....BETULA LUMINIFERABETULA UTILIS, HIMALAYAN BIRCH .BETULA pApyRiFERA, PAPER BIRCH, CANOE BIRCHBETULA POPULIFOLIA, GREY BIRCH .BETULA NIGRA, RED BIRCHBETULA LUTEA, YELLOW BIRCHBETULA LENTA, CHERRY BIRCH, BLACK BIRCHBETULA FONTINALIS ....DlOSPYROS .DlOSPYROS VIRGINIANA, AMERICAN PERSIMMONDIOSPYROS LOTUS, DATE-PLUM9 2 592692892992993°93293393593794594995195 295395495595 695795895996296697497597697797998098098398798899°991992995996999\C7

ILLUSTRATIONSNative Scots Pine at Invergarry (from a drawing by Miss Ruth Brand) . . FrontispiecePLATE No.Silver Fir at Cowdray ... ..... 208Silver Fir at Longleat . . . . . . . . .209Silver Fir at Roseneath . . . . . . . . .210Silver Fir at Tullymore . . . . . . . . .211Spanish Fir in Andalusia . . . . . . . . .212Spanish Fir at Longleat . . . . . . . . .213Greek Fir at Barton . . . . . . . . . .214Himalayan Fir in Sikkirn . . . . . . . . .215Japanese Fir in Japan . . . . . . . . .216Giant Fir at Eastnor Castle . . . . . . . . .217Giant Fir in Vancouver's Island . . . . . . . .218Californian Fir at Linton . . . . . . . .219Lovely Fir in British Columbia . . . . . . . .220Noble Fir in Oregon . . . . . . . . .221Red or Shasta Fir at Bayfordbury . . . . . . . .222Red or Shasta Fir at Bonskeid . . . . . . . .223Bristle-cone Fir at Eastnor Castle . . . . . . . .224Rocky Mountain Fir in Montana . . . . , . . .225Mexican Fir at Fola ......... 226Douglas Fir on Barkley's Farm . . . . . . .227Douglas Fir Forest in Vancouver's Island . . . . . . 228Douglas Fir at Eggesford . . . . . . . . .229Douglas Fir at Lynedoch . . . . . . . . .230Douglas Fir at Tortworth . . . . . . . . .231Spanish Chestnut Grove at Bicton . . . . . . . .232Spanish Chestnut at Althorp ......... 233Spanish Chestnut at Thoresby . . . . . . . 234Spanish Chestnut at Rydal . . . . . . . . 235Spanish Chestnut at Rossanagh . . . . . . . .236vii

Vlll The Trees of Great Britain and IrelandPLATE No.Japanese Chestnut at Atera, Japan .Weeping Ash at Elvaston CastleTall Ash at Cobham Park .Twisted Ash at Cobham ParkTall Ash at Ashridge .Ash at Woodstock, KilkennyAsh at Castlewellan .Diseased Ash at Colesborne ; Deformed Ash at CirencesterNarrow-leaved Ash at Rougham HallWhite Ash at Kew .Biltmore Ash at Fawley CourtZelkova crenata at Wardour Castle .Zelkova crenata at GlasnevinZelkova acuminata at CarlsruheCeltis occidentalis at West Dean ParkAlders at Lilford .Alders at Kilmacurragh .Italian Alder at Tottenham House, SavernakeBirch at Savernake Forest .Birch at Merton Hall .Birches in Sherwood Forest .Gnarled Birches in GlenmorePaper Birch at Bicton .Yellow Birch at Oriel TempleDiospyros virginiana at Kew . .Fraxinus; leaves, etc.Fraxinus; leaves, etc. .Fraxinus ; leaves, etc.Fraxinus; leaves, etc.Fraxinus; leaves, etc. .Moms, Celtis, and Zelkova; leaves, etc.Alnus.; leaves, etc. .Betula; leaves, etc. .Betula; leaves, etc. .2372 3 8239240241242243244245246247248249250251252253254255256257258259260261262263264265266267268269270fABIESAîes, Linnaeus, Gen. PL 294 (in part) (1737); Bentham et Hooker, Gen. PI. iii. 441 (1880);Masters, Journ. Linn. Soc. (£ot.) xxx. 34 (1893); Hickel, Bull. Soc. Dendr. France, 1 907, pp.5, 41, and 82 ; 1908, pp. 5 and 179.Picea, D. Don, in Loudon, Arb. et Frut. Brit. iv. 2293 (1838).EVERGREEN trees belonging to the order Coniferae; bark containing numerous resinvesicles; branches whorled. Buds, with numerous imbricated scales, with or withoutresin, usually two to five at the ends of the branchlets, the central bud terminal andlargest, the others surrounding it in a circle on upright shoots, whilst on lateralbranchlets those on the upper side are not developed ; buds also occur rarely andfew in number in the axils of the leaves on the branchlets below. Branchlets of onekind, usually smooth, but in certain species grooved, with raised pulvini; eachseason's shoot * marked by a sheath at the base, composed of the persistent budscalesof the previous spring.Leaves on fertile and barren branchlets, often different in length and thicknessand in the nature of the apex ; arising from the branchlets in spiral order, radiallydisposed on vertical shoots, but variously arranged according to the species onlateral branchlets; persisting for many years and giving the tree a dense mass offoliage; leaving as they fall circular scars on the branchlets ; sessile, but usuallynarrowed just above the expanded circular base; linear, flattened and thin in mostspecies, quadrangular in section in a few species; ventral surface always with twogreyish or white stomatic bands, one on each side of the raised green midrib; dorsalsurface with or without stomata, which when present are either in continuous lines,as in the quadrangular-leaved species, or are confined to near the tip of the leaf inthe middle line, as in some flat-leaved species; apex acute, acuminate, or obtuse,notched or entire, spine-pointed in one or two species; resin-canals 2 two, constantin position for each species in the leaves on lateral branchlets, but in some species 3differing in position in the leaves on the upright or fertile branchlets, either median,1 In A . bracteata, all the bud-scales usually fall off, leaving ring-like scars at the base of the shoot.2 The position of the resin-canals is easily seen on examining a thin section with a lens ; and can often be made out bysqueezing the leaf, after it is cut across, when the resin will be observed exuding from the two canals.3 In A . pectinata, A. ccphalonica, and A . Nordmanniana, the resin-canals are marginal in the leaves of lateral branches,and are median in the leaves of cone-bearing branches. Cf. Guinier and Maire, in Bull. Soc. Bot, France, Iv. 189 (1908).IV 713 B

714 The Trees of Great Britain and Irelandwhen situated in the substance of the leaf about equidistant between its upper andlower surfaces, or marginal or sub-epidermal, when placed in the lower part of theleaf close to the epidermis; fibro-vascular bundle simple in some species, dividedinto two parts in other species.Flowers monoecious, the two sexes on separate branchlets; male flowers usuallyabundant and on the lower side of the branchlets over the upper half of the tree ;female cones on the upper side of the branchlets, usually only near the top of thetree, but in some species borne all over the upper half of the tree. Staminateflowers, 1 solitary in the axils of the leaves of the preceding year's shoot; stamensspirally crowded on a central axis, anthers surmounted by a knob-like projectionand dehiscing transversely. Female cones, 1 arising as short shoots, composed ofnumerous imbricated fan-shaped ovuliferous scales, and an equal number of muchlonger mucronate bracts ; ovules inverted, two on each scale.Mature cones erect on the branchlets, composed of closely imbricated woodyscales, more or less fan-shaped with short stalks. Bracts adnate to the outer surfaceof the scales at the base ; either concealed between the scales or with their tipsexserted and then often reflexed over the margin of the scale next below ; dilated atthe apex, entire or two-lobed, prolonged into a triangular mucro. Seeds two on theinner surface of each scale, winged, and with resin-vesicles. The cones ripen in oneseason ; and the scales, bracts, and seeds fall away from the central spindle-like axisof the cone, which persists for a long time on the tree. The seedling has four to tencotyledons, stomatiferous on their upper surface.The species of Abies are distinguishable from all other conifers by the circularbase of the leaves, which on falling leave circular scars on the branchlets.The species of Abies have been variously divided into sections by differentauthors, but no satisfactory arrangement has yet been made out. Mayr proposedthree sections based on the colour of the cones ; but, as Sargent 2 points out, colouris not a constant character in several species. The cones are of value in the discrimination of the species, by taking into account their age, general appearance, andcharacters as a whole ; but the scales are often very variable in shape in the samespecies, and the bracts, while more constant in form, often show considerablevariation in their length. It is most convenient, in practice, especially as cones arein most cases not available for examination, to group the species, according to thecharacters of the buds, branchlets, and foliage, which are, as a rule, very constant inthe same species. Hickel 3 proposes three sections, based on the characters of thebranchlets and buds; but his division is artificial, as it separates species closelyallied by the characters of their cones.Some notes on the genus Abies, for which we are indebted to Mr. J. D.Crozier, forester to H. R. Baird, Esq. of Durris, Kincardineshire, are inserted.Mr. Crozier's long experience in the east of Scotland gives a special value to hisopinion on their respective qualities for planting in Scotland, which our own1 Both the Staminate flowers and the young female cones are surrounded at the base by involucres of bud-scales.2 Sitva N. Amer. xii. 97, adnot. ( 1898). Sargent proposes three sections, based on the characters of the leaves.3 Bull. Soc. Dendr. France, 1 907, p. II.Abies 715could not have, though in almost every case he confirms the conclusions at whichwe had already arrived.About thirty species are known, of which twenty-six have been introduced andare distinguished below. The silver firs are natives of the temperate parts of thenorthern hemisphere, usually occurring in mountainous regions; attaining highelevations towards the south, as in Guatemala, Algeria, Himalayas, and Formosa;and descending to low levels in the extreme north, as Alaska, Labrador, and Siberia.The following table is based upon characters taken from the foliage, buds, andshoots of lateral branches, occurring on the lower part of the tree. As regards theleaves, their arrangement upon the branchlets, the position of the resin-canals, andwhether the apex is entire or bifid must be noted. The presence of stomata on theupper surface of the leaf is peculiar to certain species. The young shoots are eithersmooth or deeply grooved with prominent pulvini; and are glabrous in some species,pubescent in others, the pubescence when present being either confined to thegrooves or spread over the whole branchlet. The buds vary in size and shape andalso in the quantity of resin, which in some cases is so slight that they may bedescribed as non-resinous ; whilst in other species the scales are covered with ordeeply immersed in resin.Certain species are distinguishable at a glance by some prominent character.A. bracteata has a bud entirely different from that of any other species. A . Pinsapo,with its short, thick, rigid leaves, standing out radially from the shoot, is unmistakable. A . cephalonica, with a more imperfect radial arrangement, is distinguishedby its long flattened leaves ending in a single sharp cartilaginous point. A . firma ispeculiar in its remarkably broad very coriaceous leaves, which end in two sharpunequal points. A . grandis has the leaves quite pectinate in the horizontal plane,those of the upper rank about half the size of those below. A. Mariesii is distinguished by the shoot being densely covered with a ferruginous tomentum. A .brachyphylla and A . Webbiana have deeply-furrowed shoots with prominent pulvini,which become more marked in the second year; and the bark begins to scalevery early on the branches and trunk of the tree. A . nobilis and A . magnifiedare peculiar in the upper median leaves curving up from the shoot after beingappressed to it for some distance. A . Pindrow has long pale green leaves veryirregularly arranged.I. Leaves radially arranged on the branchlets ; apex of the leaf not bifid.1. Abies Pinsapo, Boissier. Spain. See p. 732.Leaves rigid, short, less than f inch long, thick, acute at the apex ; resincanalsmedian. Shoots glabrous. Buds resinous.2. Abies cephalonica, Loudon. Greece. See p. 739.Leaves thin, flattened, about i inch long, ending in a sharp cartilaginouspoint; resin-canals marginal. Shoots glabrous. Buds resinous.In van Apollinis, the radial arrangement is imperfect, and the leaves end ina short point.

716 The Trees of Great Britain and IrelandII. Leaves on the lateral branches pectinate in arrangement; the t^vo lateral setseither in one plane, or with their upper ranks directed upwards as wellas outwards, showing a V-shaped depression, as seen from above, between thetwo sets.* Resin-canals marginal?3. Abies bracteata, N uttall. California. See p. 796.Leaves long, 2 inches or more, rigid, ending in a spine-like point. Shootsglabrous. Buds peculiar in the genus, elongated, fusiform, membranous,non-resinous.4. Abies grandis, Lindley. Western N. America. See p. 773.Leaves all in one plane, those in the upper rank about half the lengthof those below, up to 2 inches long, bifid at the apex; upper surfacegrooved and without stomata. Shoots minutely pubescent. Buds small,resinous.5. Abies Lowiana, Murray. California. See p. 779.Leaves in a V-shaped arrangement, IJ to 2^ inches long, bifid at the apex ;upper surface grooved and with eight lines of stomata. Shoots and budsas in A . grandis.6. Abies firma, S iebold and Zuccarini. Japan. See p. 762.Leaves in a V-shaped arrangement, rigid, very coriaceous, broad, up to \\ inchlong, ending in two sharp cartilaginous points. Shoots pubescent in thefurrows between the slightly raised pulvini. Buds small, ovoid, onlyslightly resinous.7. Abies homolepis, Siebold and Zuccarini. Japan. See p. 764-Leaves in arrangement and appearance like A . firma ; but shorter, lesscoriaceous, narrower, and whiter beneath. Shoots with prominent pulvini,glabrous. Buds ovoid, resinous, larger than in A . firma.8. Abiespectinata, De Candolle. Europe. See p. 720.Leaves pectinate in one plane or tending to a V-shaped arrangement, aboutan inch long, slightly bifid at the apex. Shoot grey, with short pubescence.Buds ovoid, non-resinous.9. Abies Webbiana, Lindley. Himalayas. See p. 750.Leaves V-shaped in arrangement, up to 2^ inches long, bifid, silvery whitebeneath. Shoots with prominent pulvini and deep grooves, with a reddishpubescence confined to the grooves. Buds large, globose, resinous.** Resin-canals median?10. Abies balsamea, M iller. Eastern N. America. See p. 803.Leaves slender, scarcely i inch long, bifid at the apex, with six to eight linesof stomata in each band on the lower surface. Shoots, smooth, grey, withscattered short erect grey pubescence. Buds globose, resinous.11. Abies Fraseri, Poiret. Alleghany Mountains. See p. 806.Leaves as in A . balsamea, but shorter and whiter beneath, with eight to1 A. cilicica and A. numidica, with weak shoots, come in this section. See Nos. 22 and 23.2 Abies lasiocarfa, Nuttall, often has the leaves more or less pectinate, and might be sought for here. See No. 26.Abies 717twelve lines of stomata in each band beneath. Shoots smooth, yellowish,with dense reddish curved or twisted pubescence. Buds globose,resinous.12. Abies brachyphylla, Maximowicz. 1 Japan. Seep. 765.Leaves in a V-shaped arrangement, short, scarcely exceeding f inch, slightlybifid, white beneath. Shoots glabrous, with prominent pulvini and deepgrooves. Buds conical, resinous.III. Leaves on lateral branches not pectinate above, but densely crowded, those in themiddle hne directed forwards in imbricated ranks, their bases not beingappressed to the branchlet. On the lower side of the shoot the leaves are intwo lateral sets.* Resin-canals marginal?13. Abies Nordmanniana, Spach. 3 Caucasus, Northern Asia Minor. See p. 746.Leaves up to \\ inch long, with rounded bifid apex. Shoots smooth, withshort scattered erect pubescence. Buds ovoid, brown, non-resinous.14. Abies amabilis, Forbes. Western N. America. See p. 782.Leaves in arrangement and size like those of A . Nordmanniana, but muchdarker shining green, and with a truncate bifid apex ; they emit a fragrantodour when bruised. Shoots smooth, with short wavy pubescence. Budssmall, globose, resinous.15. Abies religiosa, Schlechtendal. Mexico, Guatemala. See p. 808.Leaves about i inch long, gradually narrowing from the middle to theusually entire apex, which is occasionally slightly emarginate. Shootswith prominent pulvini and dense minute erect pubescence. Buds shortlycylindrical, resinous.The median upper leaves are much less numerous than in the two precedingspecies.16. Abies Mariesii, Masters. Japan, Formosa. See p. 771.Leaves shorter and broader than in Abies Veitchii, widest in their upperthird, with a rounded and bifid apex. Shoot densely covered with aferruginous tomentum. Buds small, globose, resinous.** Resin-canals median.17. A bies Veitchii, Lindley. Japan. Seep. 768.Leaves up to i inch long, truncate and bifid at the apex, uniform in width,very white beneath, with nine to ten lines of stomata in each band. Shootssmooth, covered with dense short erect pubescence. Buds small, globose,resinous.The upper median leaves, pointing forwards, stand off from the shoot at awider angle than in A . Nordmanniana.1 Abies umbellata, Mayr, is said to be very similar in foliage to this species. See the description of this species, p. 768.2 A. numidica with strong shoots, is distinguished from all these species by the leaves of the upper side being directedbackwards. See No. 23.3 A. cilicica, with strong shoots, resembles a weak A. Nordmanniana. See No. 22.

718 The Trees of Great Britain and Ireland18. Abies sachalinensis, Masters. Saghalien, Yezo, Kurile Isles. See p. 760.Leaves long and slender, up to if inch, uniform in width, with a rounded andbifid apex, white beneath, seven to eight lines in each stomatic band.Shoots with prominent pulvini, and a dense short pubescence confined tothe grooves. Buds small, globose, resinous.19. Abies sibirica, Ledebour. N. E. Russia, Siberia, Turkestan. See p. 758.Leaves long and slender, up to i J inch, uniform in width ; apex rounded andeither slightly bifid or entire; four to five lines in each stomatic bandbeneath. Shoots ashy grey, quite smooth, with a scattered minutepubescence. Buds small, globose, resinous.IV. Leaves on lateral branches not pectinate above; those in the middle linecovering the branchlet, and curving Jtpwards after being appressed to theshoot for some distance at their base. The leaves are in two lateral sets onthe lower side of the branchlet. Resin-canals marginal.20. Abies nobilis, Lindley. Washington, Oregon, California. See p. 786.Leaves above closely appressed by their bases to the branchlet, which theycompletely conceal; about i inch long, entire at the apex, flattened, groovedon the upper surface in the middle line; stomata usually present on bothsurfaces. Shoots with a dense, short brown pubescence. Terminal budsgirt at the base by a ring of acute or subulately-pointed pubescent scales.21. Abies magnijica, Murray. Oregon, California. Seep. 792.Leaves above appressed at their bases, for a short distance only, to thebranchlet, which they do not completely conceal; longer than in A. nobilis,up to if inch, entire at the apex, quadrangular in section, not grooved onthe upper surface; stomata always present on both surfaces. Shoots andbuds as in A. nobilis.V. Leaves on lateral branches arranged in two ways, which are often observable onthe same tree, and depend ^^pon the vigour of the shoots.22. A bies cilicica, Carriere. Asia Minor. See p. 744.Leaves either (A) pectinate above with a V-shaped depression between thelateral sets, or (B) with the median leaves above crowded and covering thebranchlet, as in A . Nordmanniana. The leaves are slender, up to \\ inchlong, not conspicuously white below, slightly bifid at the rounded or acuteapex; resin-canals marginal. Shoots smooth, with scattered short erectpubescence. Buds small, ovoid, non-resinous.Vigorous shoots of this species resemble a weak A . Nordmanniana ; but withthe leaves shorter, more slender, and less white beneath, the buds beingmuch smaller.23. A . numidica, De Lannoy. Algeria. See p. 737.Leaves either (A) pectinate above with a V-shaped depression; or (B)crowded and covering the upper side of the branchlet, but different fromAbies 719all other species in the median leaves above, in that case, being directedbackwards and not forwards. Leaves short, up to f inch long, broad,rounded at the entire or slightly bifid apex ; in most cases with four to sixbroken lines of stomata on their upper surface near the tip; resin-canalsmarginal. Shoots brown, shining, glabrous. Buds large, ovoid, nonresmous.VI. Leaves irregularly arranged; those on the lower side of the branches not trulypectinate.24. Abies Pindroiv, Spach. W. Himalayas. See p. 755.Leaves all directed more or less forwards ; those above irregularly and imperfectly covering the branchlet; those below mostly pectinate, but with somedirected downwards and forwards. Leaves soft, pale green, up to z\ incheslong, bifid at the apex with two sharp cartilaginous points; resin-canalsmarginal. Shoots grey, glabrous. Buds large, globose, resinous.25. A bies concolor, Lindley and Gordon. Colorado, Utah, Arizona, New Mexico,Northern Mexico, Southern California. See p. 777.Leaves imperfectly pectinate both above and below, some in the middle linebeing always directed forwards and not laterally outwards; up to 2 to 3inches long; apex entire; upper surface convex and not grooved, bearingfifteen to sixteen lines of stomata ; resin-canals marginal. Shoots smooth,olive-green, glabrous. Bud large, conical, resinous.26. Abies lasiocarpa, N uttall. Western N. America. See p. 800.Leaves either (A) in an imperfect pectinate arrangement, or (B) with most ofthe leaves directed upwards, those in the middle line above crowded, andstanding edgeways; i^ inches long, narrow, usually entire, with conspicuous lines of stomata on the upper surface, especially in its anteriorhalf. Resin-canals median. Shoots smooth, with a moderately dense,short wavy pubescence. Buds small, conical, resinous.Four species, A . Delavayi, Franchet j 1 A . Fargesii, Franchet; 2 A . squamata,Masters; 8 and A. recurvata, Masters; * occur in the mountains of western Chinaand are not included in the above list. The two first species are reported byMasters to have been introduced by Wilson ; but, on inquiry, we find that only onespecies of Abies from China is now growing in the Coombe Wood nursery. It isprobably A . Fargesii; but, as the plants are still very young, we are uncertain ofthis identification, and think it best to leave this species undescribed for the present.(A. H.)^ Journ. dc Bot. 1 899, p. 255; Masters, Card. Chron. xxxix. 212, fig. 82(1906).2 Journ. dt Bot. 1 899, p. 256 ; Masters, Card. Chron. xxxix. 212, fig. 83 (1906).3 Card. Chron. xxxix. 299, fig. 121 (1906), and Journ. Linn. Soc. (Bot.), xxxvii. 423 (1906).4 Journ. Linn. Soc. (Bot.), x xxvii. 423 (1906).

72,0 The Trees of Great Britain and IrelandABIES PECTIN AT A, COMMON SILVER FIRAbiespectinala, De Candolle, in Lamarck, More Franf. iii. 276 (1805); Willkomm, Forstliche Flora,112 (1887); Mathieu, Flore Forestiere, 5 25 (1897) ; Kent, Veitch's Man. Coniferce, 530 (1900).Abies alba, 1 Miller, Diet. ed. 8, No. i (1768); Kirchner, Lebengesch. Bliitenpfl. Mittekuropas, i. 78(1904).Abies vulgaris, Poiret, in Lamarck, Diet. vi. 514 (1804).Abies Picea, Lindley, Penny Cycl. i. 29 (not Miller) (1833).Pinus Picea, Linnaeus, Sf. PI. 1 001 (1753).Pinns Abies, Du Roi, Obs. Bot. 3 9 (1771).Pinuspectinata, Lamarck, Fl. Franc, ii. 202 (1778).Picea pectinata, Loudon, Arb. et Frut. Brit. iv. 2329 (1838).A tree attaining under favourable conditions about 150 feet in height and 20feet or more in girth. Bark on young trees, smooth, greyish; ultimately fissuringand becoming rough and scaly. Buds small, ovoid, non-resinous; scales few,brownish, rounded at the apex. Young shoots grey, smooth, with a scattered shorterect pubescence, which is retained in the second year.Leaves on lateral branches pectinately arranged in two lateral sets; those belowthe longest and directed outwards and slightly forwards in the horizontal plane;those above directed upwards and outwards, forming between the two sets a shallowV-shaped depression. Leaves about i inch long, -fa inch broad, linear, flattened,narrowed at the base, tapering slightly to the rounded, bifid apex; upper surfacedark green, shining, with a continuous median groove and without stomata; lowersurface with two white bands of stomata, each of seven to eight lines; resin-canalsmarginal.On leading shoots the leaves are radially arranged, and differ considerably fromthose on lateral branches; they are thicker, with median resin-canals, acute and notbifid at the apex, and often show lines of stomata on their upper surface towards thetip. Leaves on cone-bearing branches are nearly all directed upwards, very sharppointed,and almost tetragonal in section.Trees, standing in an isolated position, usually begin to flower at about thirtyyears old; when crowded in dense forests, much later, usually not before sixtyyears old.Staminate flowers, surrounded at the base by numerous imbricated scales,cylindrical, about i inch long, with greenish - yellow stamens. Female cones,appearing in August of the previous year as large rounded buds, enclosed in brownscales, and situated just behind the apex of the shoot; in spring, when developed,erect, cone-shaped, about i inch long, surrounded at the base by fringed scales ;bracts numerous, imbricated, denticulate, ending in long, acuminate points, andcompletely concealing the much smaller ovate, rounded ovuliferous scales.1 Abies alba, the oldest name under the correct genus, was never in use until lately, when it has been resuscitated bySargent and some continental botanists. This is one of the cases where adhesion to strict priority would lead to great confusion ; and hence we have adopted the name Abies pectinata, by which the tree is generally known.Abies 72,1Cones on short stout stalks, cylindrical, slightly narrowed at both ends, obtuseat the apex, about 6 inches long, 2 inches in diameter, greenish when growing, dullbrown when mature, with the points of the bracts exserted and reflexed. Scalestomentose externally, fan - shaped, about i inch broad and long; upper marginslightly uneven; lateral margins denticulate, each usually with a sinus, below theslight wings on the outer side of the scale; claw clavate. Bract with an oblongclaw, extending up three-quarters the height of the scale, and expanding above intoa lozenge-shaped denticulate lamina, which ends in a sharp long triangular mucro.Seed with wing about an inch long; wing about twice as long as the body of theseed.SEEDLINGSeed sown in spring germinates in three or four weeks. The cotyledons,usually five in number, are at first enveloped, as with a cap, by the albumen of theseed ; but speedily casting this off, they spread radially in a whorl at the summit ofthe short caulicle, and remain green on the plant for several years; about an inch inlength, linear, obtuse at the apex, flat beneath, and slightly ridged on the uppersurface, which shows two whitish bands of stomata. In the first year only a singlewhorl of true leaves, arising immediately above the cotyledons and alternating withthem, is produced. Primary leaves short, acute, or obtuse, but not emarginate atthe apex, and with the stomatic bands on the lower surface. A terminal bud closesthe first season's growth, the plant scarcely attaining two inches high. In the secondyear ordinary leaves, arranged spirally on the stem, are produced. The growth ofthe plant in the first two or three years is mainly concentrated in the root, whichdescends deep into the soil, the increase in height of the stem above ground beingtrifling. The stem branches in the third or fourth year, and produces annually forsome years one or two lateral branches, making no great growth in height, reachingin the ninth year an average of two feet. About the tenth year normal verticillatebranching begins; and from this onwards the plant makes rapid growth.VARIETIESDr. Klein gives in Vegetationsbilder illustrations of some remarkable forms 1which the silver fir assumes at high elevations in Central Europe, and which he calls" Wettertanne " or " Schirmtanne." These trees have lost their main leader throughlightning, wind, or otherwise, and have developed immense side branches whichspread and then ascend, sometimes forming a candelabra-like shape. The finest ofthis type known to him is at St. Cerques in Switzerland, and measures at breastheight no less than 7.40 metres in girth, about the same as the largest of theRoseneath 2 trees.Other varieties, distinguished by their peculiar habit, occur in the wild state.1 These forms are also described by Dr. Christ in Garden and Forest, ix. 273 (1896).8 One of the trees at Roseneath, Dumbartonshire, has a similar growth of erect branches, like leaders from some of thehorizontal liinbs. This is figured, from a photograph by Vernon Heath, in Card. Chron. xxii. 8, fig. I (1884). At Powerscourtthere is also a large tree, 13 feet 3 inches in girth, with branches prostrate on the ground and sending up several uprightstems.IV c

72,2, The Trees of Great Britain and IrelandVar. pendula? with weeping branches, has been found in the Vosges and inEast Friesland.Var. virgata? found in Alsace and Bohemia, has long pendulous branches, onlygiving off branchlets near their apices, and densely covered with leaves.Var. pyramidalis? This form, which in habit resembles the cypress or aLombardy poplar, was found growing wild in the department of Isere in France. Avery fine example, about 35 feet high in 1904, is growing in the arboretum ofSegrez.Var. columnaris? very slender in habit, with numerous short branches, all ofequal length, and with leaves shorter and broader than in the type.Var. tortiiosa, a dwarf form, with twisted branches, and bent, irregularly-arrangedleaves.Var. brevifolia, another dwarf form, distinguished by its short broad leaves.Remarkable variations in the cones have also been observed. A tree, discoveredby Purkyne 5 in Bohemia, bore cones, umbonate at the apex, and with short and nonreflexedbracts. Beissner 6 mentions a tree, growing in the park at Worlitz nearDessau, which produced cones a foot in length.DISTRIBUTIONThe common silver fir is a native of the mountainous regions of central andsouthern Europe. The northern limit of its area of distribution begins in thewestern Pyrenees about lat. 43 in the neighbourhood of Roncesvalles in Navarre;and crossing the chain it extends along its northern slope as far as St. Beat; fromhere it bends northwards to the mountains of Auvergne, whence it is continued in anorth-easterly direction through Burgundy and French Lorraine, crossing the easternslope of the Vosges about the latitude of Strasburg. From here it curves for somedistance westward, and reaching Luxemburg, is continued through Trier and Bonnto southern Westphalia. Across the rest of Germany, according to Drude, whogives a map of the distribution of the species, the northern limit extends as anirregular line about lat. 51 , which touches Hersfeld, Eisenach, the northern edge ofthe Thuringian forest, Glauchau, Rochlitz, Dresden, Bautzen, and Gb'rlitz; and endsin the southern point of the province of Posen. Around Spremberg to the northof the limit just traced, it is found wild in a small isolated territory.The eastern limit, beginning in Posen, extends through Poland along the RiverWartha to Kolo, crosses to Warsaw, and descending through Galicia west ofLemberg, reaches the Carpathians in Bukowina; and is continued along themountains of Transylvania to Orsova on the Danube.The southern limit is not clearly known as regards the Balkan peninsula, asthe silver fir, which occurs in the mountains of Roumelia, Macedonia, and Thrace,1 Kottmeier found peculiar weeping silver firs in the Friedeberg forest, near Wittmund in East Friesland, in 1882.Cf. Wittmack's Garlenzeitung, 1 882, p. 406, and Conwentz, Seltene Waldbdume in Westfreussen, 1 61 (1895).2 Caspary, in IlempeFs Oesterr. Forstzeitung, 1 883, p. 43.3 Carriere, Conif. 280. 4 Carriere, Rev. Hort. 1 859, p. 39.6 Willkomm, Forstliche Flora, 1 18 (1887). 8 Nadelhohkunde, 433 (1891).Abies 7*3supposed to be A . pectinata, is more probably a form of A . Apollinis. In Italy thecommon silver fir reaches its most southerly point on the Nebroden and MadoniaMountains of Sicily at lat. 38 . From here the limit follows the Apennines upthrough Italy, crosses into Corsica, and from there passes into Spain, where itextends from Monseny, near the Mediterranean coast in lat. 41 25', parallel to thePyrenees, through the mountains of Catalonia and northern Aragon to Navarre.In Spain the silver fir also occurs westwards on a few points of the northern littoralin the Basque provinces and Asturias.Within the extensive territory just delimited, the silver fir is very irregularlydistributed, being totally absent in many parts, as on the plains and lower mountainsof southern Europe. In the eastern part of its area it occurs only as isolated treesor in small groups in the beech and spruce forests; whereas, in the western part, asin France and in parts of Germany, it forms forests of great extent, either pure or inwhich it is the dominant species.In France the largest forests of the silver fir are in the Vosges and in the Jura.Important forests also occur in the eastern parts of the Pyrenees, the Cevennes, themountains of Auvergne, and the Alps of Dauphine. It is rare on the hills ofBurgundy, and does not occur in the Ardennes. There are small woods of thisspecies on some of the hills in Normandy, which are, however, supposed to beplanted and not indigenous. The great forest of the Vosges* is about 50 miles longby 5 to i o miles in width, and contains about 200,000 acres, situated mainly between1100 and 3300 feet elevation. This forest consists chiefly of silver fir, though, insome parts, there is a considerable mixture of beech, spruce, and common pine.The most productive woods are on siliceous soil, and only contain 10 per cent ofbeech and pine; their mean annual production being about 100 cubic feet per acre,the volume of timber standing on each acre averaging 4500 cubic feet.In the Jura there are even richer and more homogeneous forests than in theVosges, being according to Huffel the finest in Europe. Here the soil is limestone.One of these forests, which covers Mount La Joux, between 2100 and 3000 feetaltitude, contains 10,600 acres, and consists of about 90 per cent silver fir and 10 percent of spruce. The annual yield per acre is 170 cubic feet of timber. The totalvolume of standing timber, including only trees over 2 feet in girth, is 6000 cubicfeet per acre. The net revenue is thirty-two shillings an acre. There are severalother forests equally valuable in this region.One of the finest silver firs 1 in France, a tree called " Le President," is growingin the forest of La Joux. It is 163 feet high, with a clean stem of 93 feet, and a girthof 15 feet; and contains 1600 cubic feet of timber. In the forest 2 of Gerardmer,in the Vosges, there are two fine trees. One, the Beau Sapin, has a height of144 feet and a girth of 13 feet 8 inches; it contains 777 cubic feet of timber, and isvalued at ^16. The other, the Gdant Sapin, has a height of 157 feet and a girthof 14 feet 5 inches; it contains 1095 cubic feet of timber, and is valued at ,£27. Inthe Pyrenees the silver fir occurs between 4500 and 6500 feet elevation, and trees1 See Huffel, Economic Forestiire, i, 349, 350, 353 (1904).2 Cf. Trans. A Scot. Arh. Soc. xviii. 131 (1905).

724 The Trees of Great Britain and Irelandof great age, about 800 years old, are said* to have existed there at the beginning ofthe i gth century.In Corsica the silver fir occurs in the great forests of Pinus Laricio, butis not abundant, as it only grows, as a rule, in scattered groups in the gullies,where the soil is deeper and richer than elsewhere; and at Valdoniello I onlysaw a few trees, none of which were of large size. M. Rotges, of the ForestService, informed me that it occurs in greatest quantity in the forest of Pietropiano,near Corte.In Italy the silver fir is unquestionably wild on the Apennines, and considerableforests exist at Vallombrosa and Camaldoli, which are now owned by the government. That at Camaldoli is particularly fine, the total area covered by the silverfir being about 1600 acres. The trees are dense on the ground and very vigorousin growth; and this is easily explained by the heavy rainfall, which, as measured atSt Eremo, in the middle of the forest, at 3600 feet altitude, averages about 80inches annually. I saw, when I visited Camaldoli, in December 1906, no treesof great size ; but one was cut down in 1884, and a log of it shown at the NationalExhibition at Turin in that year, which measured 140 feet in height and 17 feet ingirth.The silver fir also occurs in Sicily in small quantity, on the higher mountains,and specimens without cones, which I saw in the museum at Florence, are peculiarin the foliage, and form possibly a connecting link between A . pectinata and A .numidica.In Germany, towards the northern part of its area of distribution, the silver firis met with growing wild on the plains, as in Saxony, Silesia, and Thuringia.Towards the south it is entirely a tree of the mountains, occupying a definite zone ofaltitude, which, in the Bavarian forest, lies between 950 and 400x3 feet. The largestforests, which are nearly pure, occur in the Black Forest and in Franconia; those inBavaria, Bohemia, Thuringia, and Saxony being smaller in extent.In Switzerland small forests occur at Zurich, Payerne, and on Mount Torat;the silver fir ascending in the Swiss Alps to 530x3 feet altitude. (A. H.)As to the size 2 which the silver fir attains in its native forests, many particularsare given by French and German foresters, some of which have been quoted above.None exceed, however, what I have seen in the virgin forests of Bosnia, where Imeasured near Han Semec, at an elevation of about 3000 feet, a fallen tree over 180feet long, whose decayed top must have been at least 15 to 20 feet more. Loudonstates that he saw, in the museum at Strasburg, a section of a tree of the estimatedage of 360 years, cut in 1816 at Barr, in the Hochwald, which was 8 feet in diameterat the base and 150 feet high.The virgin forests of Silesia and Bohemia contain silver firs of immense size,of which very interesting particulars are given by Gb'ppert,3 who states that, inPrince Schwarzenberg's forest of Krummau, there existed many silver firs of from1 Willkomm, Forstliche Flora, 1 16, note (1887).2 Kerner, Nat. Hist. Plants, Eng. trans. i. 722, gives the " certified height" of Abies pectinata as 75 metres, or 250feet; but this is not confirmed hy other authorities.3 H. R. Goppert, Skizzen zur Kenntniss der Urwalder Schlesiens und Bohmens, 1 8 (1868).rPAbies 7*5120 to 200 feet high, free from branches up to 80 to 120 feet, and as muchas 6 to 8 feet in diameter. He quotes Hochstetter, 1 who measured in theGreinerwald, near Unter-Waldau, at an elevation of 2563 feet, a silver fir blowndown by a storm, which was 9^- feet in diameter at breast height and 200 feet long,and produced 30 klafter of firewood.The silver fir is planted outside the area of its natural distribution in most partsof France, in Belgium, and in western and northern Germany, but not beyond lat.51 in eastern Prussia. It is occasionally planted in Norway, and at Christianahas attained 68 feet in length by 3^ feet in girth. At Thlebjergene, near Trondhjem,where, on the side of a hill, sloping down to the sea, with an easterly exposure,a fine plantation, 2 mainly of spruce and Scots pine, was made in 1872 and subsequentyears, there are some splendid groups of silver fir, 30 to 40 feet in height, apparently exceeding in rapidity of growth the native spruce beside it. It is met with ingardens in the Baltic provinces of Russia, as in Lithuania where there is a smallwood near Grodno, and in Courland and Livonia; here, however, it always remainsa small tree, never bears cones, and is much injured by severe winters.One of the most remarkable plantations in Europe is the one made by theHanoverian Oberforster, J. G. von Langen, in the Royal Park of Jaegersborg, nearCopenhagen, about 1765. I visited this place in 1908, and measured some of thetrees. I found that the largest now standing near the entrance at Klampenborgwas 125 feet by 12 feet 10 inches. This tree is figured in a work 8 kindly sent meby Skovrider H. Mundt. There are, however, many taller trees on the southside of the main drive, two of which I found to be 140 feet by 9 feet, and 140 feetby 8 feet in girth, respectively. I measured the girth of twenty trees out of sixtytwowhich are growing on an area of 100 by 30 paces, and believe them toaverage over 130 feet high, with an average girth of 7^ feet. In Lutken's work fulldetails are given of the measurements of these trees taken in 1893, and confirmedin 1898 by Oppermann, who found 432 trees, averaging 38-9 metres in height andcontaining 1400 cubic metres per hectare; which is equal to 20,0x30 cubic feet peracre in the round, or 15,700 feet English quarter-girth measure. My own hastyestimate on the spot was about 12,000 feet English quarter-girth measure per acre.These wonderful silver firs grow on a deep, sandy loam, on level ground near thesea, and seem to have passed their prime. Some of their timber has been used asrafters in the Secretariat hall of the new Raadhus at Copenhagen.A.pectinata* was brought to the eastern United States early in the nineteenthcentury; but it is not hardy even in the middle states.Witches' brooms and cankered swellings, due to the fungus Êcidiumelatinum, De Bary, are common on the silver fir in the continental forests; andare often seen in Ireland and the south-west of Scotland, 5 though apparently rarein England, where they have been noticed in Norfolk 6 and at Haslemere. 71 Hochstetter, Aits dan Bohmerwalde, Allg. Augsb. Zeit. 1 855, N - l g2- Cf- Sendtner, Die Vegetations-Verhaltnissedes Bayerischen Waldes ( 1860). 3 Lutken, Den Langenske Forstordning, p. 286, fig. 5 (Copenhagen, 1899).2 Seen by Henry in 1908.4 Sargent, Silva N. Amer. xii. 100, adnot. ( 1898). 6 Somerville, in Hartig, Diseases of Trees, Eng. trans. 179 (1894).8 Trans. Norfolk and Norwich Naturalist? Soc. vii. p. 255. i Specimens at Kew.

72,6 The Trees of Great Britain and IrelandThe swellings which affect the trunk or branches are due to the irritation ofthe fungus mycelium, which is perennial and stimulates the wood and bark toabnormal growth. These swellings become fissured and are entered by thespores of other fungi, which rot the wood ; and the tree, if the stem is affected,is often broken off at the weakened spot by storms or falls of snow. Thewitches' brooms begin as young shoots, bearing small yellowish leaves, on theunder surface of which two rows of tecidia are developed in August. Theseshed their spores at the end of that month and the leaves soon afterwards die andfall off. The affected shoots keep on growing, and develop into peculiar growths,set upright generally on the branches, and consisting of numerous twigs anastomosedtogether. The fungus passes one stage of its life on various species of Stellaria,Cerastium, and their allies, and Fischer 1 recommends the extirpation of these plantsfrom nurseries in which the silver fir is raised.The silver fir is very liable in its native forests to be attacked by the mistletoe.Modified roots, the so-called sinkers of the parasite, have been found in the woodenclosed in forty annual rings and as much as 4 inches long, showing that mistletoemay live on the tree for forty years. When the mistletoe dies the rootlets andsinkers survive for a time, but finally moulder and fall to pieces. The affected partsof the wood show numerous perforations, and exactly resemble the wood of atarget that has been penetrated by shot or small bullets."The bark of the silver fir remains alive on the surface to an advanced age; and,on this account, when branches, stems, or roots of adjoining trees get into contact,they often become grafted together. This is the explanation of the curiousphenomenon of the vitality of the stumps of certain trees in forests. After thestem is cut down, these stumps continue to increase in size and produce a callosity,which eventually covers the stump in the form of a hemispherical cap. Such astump procures its nourishment from an adjoining tree, with which its roots havebecome grafted.8CULTIVATIONThe silver fir 4 was introduced into England about the beginning of the seventeenthcentury; but the exact date is uncertain. The earliest trees recorded are twomentioned by Evelyn, 8 which were planted in 1603 by Serjeant Newdigate inHarefield Park in Middlesex. These had attained about 80 feet high in 1679, butfrom inquiries made by the late Dr. Masters, there is no doubt that they have longsince been cut down.Though in its own country the silver fir is a tree of the mountains, yet itattains its greatest perfection in the south and west of England, Scotland, and1 Abstract of Fischer's paper injburn. Roy. Hort. Soc. xxvii. 272 (1902).1 See Kerner, Nat. Hist. Plants, Eng. trans. i. 210, fig. 48 (1898). We have never seen or heard of mistletoe on thesilver fir in this country. 3 gee Mathieu, Flore Foresttire, 529 (1897).4 Staves were found, in 1900, lining the ancient wells in the Roman city of Silchester, Hants; and the wood wasidentified by Marshall Ward with A . pectinata. The casks, from which the staves had been taken, were probably importedfrom the region of the Pyrenees, and had either contained wine or Samian ware. Cf. Clement Reid, in Archaeologia, Ivii.253. 256 (ôi)- 6 Sylva, 106 (1679).Abies 72,7Ireland, under conditions of soil and climate very unlike those of its native forests.Though it will endure the severest winter frosts without injury, yet unless under thecover of other trees, or in very sheltered situations, it is often injured by spring frost,on account of its tendency to grow early. As regards soil it is somewhat critical,for though Boutcher 1 says that he has seen the largest and most flourishing silver firson sour, heavy, obstinate clay, yet I have never myself seen fine trees on any butdeep, moist, sandy soils, or on hillsides where the subsoil was deep and fertile. Healso says it is vain to plant them in hot, dry, rocky situations, and this is my ownexperience on oolite formations, where I have never seen a large or well-developedsilver fir. In the east and midland counties they usually become ragged at the topbefore attaining maturity, and in this country rarely attain a great age withoutsuffering from drought and wind.Though foresters of continental experience recommend this tree for underplanting,on account of its ability to grow under dense shade, yet from an economicalpoint of view it cannot be recommended here; and I do not know of any place inEngland where the financial results of planting the silver fir are, or seem likely tobe, such as would justify growing it on a large scale; partly because of its veryslow growth when young, and partly because its timber is not valued as it is inFrance and Germany. Mr. Crozier's experience 2 is very noteworthy.The silver fir seeds itself very freely in some parts of England, Scotland, andIreland, 3 but the seedlings are so slow in growth and so delicate for the first fewyears, that few survive the risk of frost, rabbits, and smothering. Sir CharlesStrickland tells me that in a wood of silver firs at Boynton, Yorkshire, which weremostly blown down in 1839, he remembers that a few years afterwards the growth ofyoung seedlings was in places so dense that he could hardly force his way throughthem. Some of these self-sown trees are now 6 feet in girth and 60 to 70 feet high,but many are stunted from want of space. Their parents are rough and branchy,dying at the top, and 10 to 12 feet in girth.REMARKABLE TREESThough the silver fir will probably be in time surpassed in height and girth bysome of the conifers of the Pacific coast of America, yet at present it has no rival insize among coniferous trees in Great Britain. Perhaps the tallest which I have seenin England is the magnificent tree (Plate 208) which grows in Gates Wood, at thetop of Cowdray Park, Sussex, at an elevation of 500 to 600 feet, and now owing toits being deprived of the shelter of the surrounding trees, likely to be blown down1 Boutcher, A Treatise on Forest Trees, 146 (1775).2 Formerly one of our most reliable trees, but now hopelessly unreliable as a timber crop, owing to its susceptibility toattack by Chermes. Like the larch, our old trees are practically immune to attack, but the difficulty in getting up youngstock experienced throughout the greater part of the country is likely to lead to its extinction altogether as an economicspecies. Has been much recommended by continental trained foresters even of late years for the purpose of underplantingin our Scotch woods, and some of those experiments I saw lately. The result is a hopeless failure in all of them.(J. D. CROZIER.)3 At Auchendrane, near Ayr, according to Mr. J. A. Campbell, there are several acres of self-sown seedlings; and inCounty Wexford I have also seen great numbers.

72,8 The Trees of Great Britain and Irelandby the first severe gale. I measured this tree in 1906 in company with Mr. Roberts,forester to the Earl of Egmont, as carefully as the nature of the ground would allow,and believe it to be still over 130 feet in height; when I first saw it in 1903 it wastaller. It is clear of branches to at least 90 feet and 10 feet 2 inches in girth. Inthe background some spruce which are even taller may be seen in our illustration.I am informed by Mr. F. H. Jervoise, of Herriard Park, Hants, that there wasa silver fir there which probably exceeded this height before its top was broken offabout sixteen years ago. A photograph, taken in 1851, shows the height to havebeen then at least double what it now is, namely 70 feet, and another tree standingnot far off measures approximately 140 feet.In the Shrubbery at Knole Park, Kent, a very large silver fir is now about110 feet high, with a clean bole about 80 feet by 12 feet; but its top is broken off,and it looks as if it might have been much taller.At Longleat there are a great number of very fine silver firs near the Gardens; andalso in the valley at Shearwater, the largest of which I measured in 1903, and foundto be about 130 feet by 16 feet 5 inches in girth. 1 Mr. A. C. Forbes estimated thecontents of this tree at 550 feet, and in the Trans. Eng. Art. Soc. v. 399, givesthe measurements of a group of twenty-seven trees, 120 years old, growing on anarea of ^ of an acre at the same place as follows: Average height, 130 feet;average girth at 5 feet, 9 feet; average contents, 180 cubic feet. Total, 5000 cubicfeet. I doubt whether any similar area of ground in England carries so muchtimber, except, perhaps, a group of chestnut and oak in Lord Clinton's park atBicton. Silver fir requires unusually good soil to attain these dimensions. Plate209 shows a part of this grove which stands at an elevation of about 500 feet ona greensand formation.There is a row of very fine silver firs by the road on Breakneck hill in WindsorPark, one of which I measured as 130 feet by 11 feet, and no doubt many as large,or nearly so, can be found in other parts of the south and west of England; but,as a rule, when the tree attains about 100 to no feet its top ceases to grow andbecomes ragged.Near the great cedar at Stratton Straw less (see Plate 133) there are some tallsilver firs, one of which in 1907 was 131 feet by 9 feet 7 inches; and Mr. Birkbeckinformed me that another, believed to be the tallest tree in Norfolk, and measuring135 feet, had been blown down in 1895 at tne same place.There are some very fine silver firs still standing at Eslington Park, Northumberland, which were planted about 1760, though Mr. Wightman, the gardener, informsme that the largest, which could be seen standing above all the other trees, wasblown down in a gale in December 1894. It measured 122 feet by 21 feet at fivefeet from the ground, and at fifty feet from the ground was still 9 feet ingirth.Almost equal to these are the trees in the Ladieswell Drive, near AlnwickCastle, Northumberland, which I saw in 1907; though not much exceeding 100 feet1 Loudon states that the tallest silver fir known in England in his time was believed to be at Longleat, and measured138 feet high by 17 feet in girth; but this tree cannot now be identified.Abies 729in height, they measure from 14 feet to 16 feet in girth, the largest being estimatedby Mr. A. T. Gillanders, forester to the Duke of Northumberland, to contain about600 cubic feet each.At Rydal Park, Cumberland, Mr. W. F. Rawnsley informs me that a silver firwas felled which contained 420 cubic feet, and doubtless there are others in thenorth-west of England as large. 1In Wales, however, I have seen none remarkable for size, though there aremany places which seem as suitable as those I have mentioned.In Scotland the silver fir attains its maximum of size in the south-west, and ina district where the climate is most unlike that of central Europe; being muchwarmer in winter, cooler in summer, and with a rainfall of 60 to 80 inches and evenmore in exceptional years.On the Duke of Argyll's property at Roseneath are the champion silver firs ofGreat Britain, both as regards age and girth. Strutt figures them in Silva Scotica(plate 6), and states that the largest was then about 90 feet by 17 feet 5 inches.Loudon, twenty years later, gave the height as 124 feet, the age as 138 years, andthe diameter of the trunk as 6 feet; but this height is almost certainly an error, aswhen I visited Roseneath in September 1906, a careful measurement made thelargest about no feet by 22 feet 7 inches, and the other, which stands close by it,105 feet by 22 feet i inch. 2 Plate 210, from a negative for which I have to thankMr. Renwick, is the best I have been able to obtain of these noble trees, whichgrow close to sea-level in deep sandy soil. The Duke of Argyll believes them tohave been planted about 1620 or 1630.Near Inveraray Castle, on the lower slopes of Dun-y-Cuagh, Mr. D. Campbell,the Duke's forester, showed me some splendid silver firs, over 120 feet high and15 feet in girth, and assured me that in his younger days he had helped to measuresome which were much larger; one he believed to have been 24 feet in girth,containing over 800 feet of timber. On the Dalmally road, a little above thestables at Inveraray, are the tallest trees of the species that I have seen in Scotland;one measures 135 feet, or perhaps as much as 140 feet, by 16^ feet; another about135 feet by 14 feet 3 inches; and there may be even taller ones here which I couldnot measure. These splendid trees were, as the Duke of Argyll informs me,probably planted by Duke Archibald in 1750, but their timber is so coarse that itis of little value, and is principally used by Glasgow shipbuilders for keel blocks.Some of the most remarkable silver firs which I have seen in any country areat Ardkinglas, now the property of Sir Andrew Noble, near the head of Loch Fyne.They are described by J. Wilkie, and well illustrated in the Trans. Scot. Arb. Soc.ix. 174, and show a tendency, which I cannot explain, to throw out immensebranches, which, after growing horizontally 10 to 15 feet from the main trunk, turnup and form an erect secondary stem. The largest of these (op. cit. plate 11), according to Wilkie's careful measurement in 1881, was 114 feet high by 18 feet in girth at1 Sir Richard Graham of Netherby Hall, Cumberland, showed me a very remarkable tree in a wood called HogKnowe, which has large spreading branches, 80 paces in circumference, and measures 98 feet by 14^ feet. Mr. Watt ofCarlisle has been good enough to send me a photograph of this tree, taken by his sister." See Card. Chron. xxii. 8, fig. I (1884), and xxvii. 166, fig. 39 (1887), where good illustrations of these trees are given.IV D

73 The Trees of Great Britain and Ireland2\ feet. I made it in 1905 about 21 feet at the same height and 14 yards round theroots. Wilkie computed that the main stem contained 557 cubic feet and thebranches 692 cubic feet, including bark, which exceeds the largest tree of thespecies recorded in this country. I certainly have never seen anything surpassingit in bulk, even in the virgin forests of Bosnia, though I have measured a fallensilver fir there which was at least 200 feet high. Another of these trees figured onplate 12 of the same volume, was estimated at 437 feet in the stem, and 449 feetin the ten principal limbs. At the same place is a very fine tree which Mrs. HenryCallender, who showed it to me, called " The Three Sisters," 115 feet high accordingto Wilkie, I made it, twenty-four years later, 120 feet, with a bole only 8 feet long,where it divides into three tall stems nearly equal in height and measuring just abovewhere they separate, 8 feet 4 inches, 8 feet 5^ inches, and 8 feet 7 inches respectively.The Union trees, 1 in the avenue at Auchendrane, Ayrshire, planted in 1707,are six in number, the largest being, in 1902, 97 feet high and 16 feet i inch ingirth. Another tree in the flower garden here, planted at the same time, was 110feet by 16 feet in 1902.In the island of Bute, James Kay describes, in Trans. Scot. Arb. Soc. ix. p. 75,some fine silver firs which grew in a clump north-east of the circle walk in the woodsof Mountstuart, the seat of the Marquess of Bute. They were of immense height(120 feet), and could be seen for miles standing out like an island among this forest ofsylvan beauty. There were nineteen silver firs, five spruce, one Scots pine, and twobirches, all standing on a space of 60 yards square, where they were healthy and notovercrowded. They were very uniform in size, and ran from 10 to 12 feet in girth,ten being straight to the top and nine forked at 30 feet to 60 feet up. 2In other parts of Scotland the silver fir usually attains smaller dimensions, thelargest that I have seen being on the banks of the Tay, near Dunkeld, and atDupplin Castle, where I measured a tree over 100 feet high by 17^ feet in girth.But Mr. W. J. Bean, in Kew Bulletin, 1 906, p. 266, mentions an immense tree,which was blown down on November 17, 1893, near Drummond Castle, when 210years old. The stump of this tree was 6^ feet in diameter, and the cubic contentsare said to have been 1010 cubic feetAt Dawyck, near Peebles, in a cold situation at about 500 feet above the sea,Mr. F. R. S. Balfour showed me some large silver firs which far surpass the larchesgrowing near them, which are believed to have been planted about 1730. Thelargest of the firs is 112 feet by 15^ feet.In most parts of Ireland the silver fir is a thriving tree wherever planted, and seemsto be well suited to the climate. It was probably introduced early in the eighteenthcentury, as, according to Hayes, there were trees 100 feet high and 12 feet in girthin 1794 at Mount Usher, in Co. Wicklow. The largest silver fir in Ireland that weknow of is at Tullymore Park, Co. Down, the seat of the Earl of Roden, growingin a sheltered valley below the house. Col. the Hon. R. Jocelyn, who showed me1 Cf. Renwick, in Trans. Nat. Hist. Soc. Glasgow, vii. 265 (1905).2 Mr. Kay informs me that many of the trees described by him thirty years ago have since been blown down, and I couldnot identify these silver firs when I visited Bute recently.Abiesthis tree in 1908, informed me that it was marked on a plan about 200 years old,and though still vigorous in appearance, it seems to be hollow for some way up. Itmeasures from 115 to 120 feet high, with a girth of 18 feet 10 inches; and at about20 feet from the ground throws out four large branches, which become erect, andform a tree of the candelabra type. (Plate 211.) At Carton, the seat of the Dukeof Leinster, a tree was 16 feet i inch in girth in 1904, but the top had been blownoff by the great gale of 1903. The finest silver firs in Ireland are probably thosegrowing at Woodstock, Co. Kilkenny, where the biggest tree was in 1904 over 120feet high by 15 feet 4 inches in girth. There are also here four trees standing soclose together that they can be encircled by a tape of 30 feet; one of these is 133feet high by 10 feet 10 inches in girth. At Avondale, Co. Wicklow, Mr. A. C.Forbes measured a tree in 1908, 125 feet in height and 15 feet 4 inches in girth.At Tykillen, Co. Wexford, the silver fir grows well and seeds itself freely, but doesnot attain anything like the dimensions above noted. There are fine trees atCastlemartyr, Co. Cork, one of which measures 114 feet by 14 feet 8 inches.TIMBERThough on the Continent the wood of the silver fir is in some districts, andfor purposes where strength combined with lightness is required,1 valued morehighly than that of the spruce or pine, yet in England it is little appreciated,because it seldom comes to market in any quantity, and the trees are rarely cleanenough to make good boards. But I am assured by Dr. Watney that, when slowlyand closely grown, it is distinctly superior in quality to that of the spruce, and that heuses it in preference on his own property for estate building; and Mr. H. E. Asprey,agent to the Earl of Portsmouth at Eggesford, Devonshire, where this tree growsvery well, tells me that he finds the timber quite equal to that of spruce for all estatepurposes. The Marquess of Bath informs me that a lot of 22 trees, averaging140 feet each, were sold privately at 5^d. per foot, and used at Trowbridge formaking tin-plate boxes ; but most of his silver fir timber goes to the Radstock coalpits, where it is used underground.Laslett says 2 that " the pinkish white and scarcely resinous wood works up well,with a bright silky lustre, and is of excellent quality for carpentry and ship-work.It is light and stiff, and like spruce takes glue well. Nevertheless it is as yet farless in request than the latter, though it is employed in the making of paper pulp,as well as for boards, rafters, etc." 8 So little is it known, however, to the Englishtimber merchant that the author of English Timber does not even mention it, and Iam not aware that it is imported to England as an article of commerce.Strasburg turpentine, which was formerly extracted from the resinous glandsfound on its bark and largely used for the preparation of clear varnishes and atone time used as medicine, is now apparently superseded by other resins, though,according to Fluckiger and Hanbury,4 it was still collected to a small extent in theVosges in 1873. (H. J. E.)1 Cf. Mouillefert, Essences Forestieres, 338(1903). 3 Timber and Timber Trees, 343 (1896).3 Christ, Flore de la Suisse, 255 (1907), says that its white wood is delicate and not so much in request as the moreresinous wood of the spruce. 4 Pharmacographia, 6 15 (1879).

732 The Trees of Great Britain and IrelandABIES PINSAPO, SPANISH FIRAlies Pinsapo, Boissier, Biblioth. Univ. Getifrve, xiii. 167 (1838), and Voyage Espagne, ii. 584, tt.167-169 (1845); Masters, Card. Chron. xxiv. 468, f. 99 (1885), xxvi. 8, f. i (1886), and iii.140, f. 22 (1888); Kent, Veitch's Man. Coniferce, 5 34 (1900).Pinus Pinsapo, Antoine, Conif. 6 5, t. 26, f. 2 (1842-1847).Picea Pinsapo, Loudon, Etuycl. Trees, 1 041 (1842).A tree attaining about 100 feet in height and 15 feet in girth. Bark smooth inyoung trees, becoming rugged and fissured on old trunks. Buds ovoid, obtuse at theapex, resinous. Young shoots glabrous, brownish, with slightly raised pulvini.Leaves on lateral branchlets radially arranged, linear, flattened, but thick, rigid,short, £ to f- inch long by about ^ inch wide, gradually narrowing in the upper thirdto the acute apex ; upper surface convex without a median furrow and with eight tofourteen lines of stomata; lower surface with two bands of stomata, each of six orseven lines ; resin-canals usually median. 1 In young plants the leaves are longer andend in sharp cartilaginous points. On cone-bearing branches the leaves are shortand thick, lozenge-shaped in section, with twenty or more lines of stomata on theupper surface, and two bands of stomata of about ten lines each on the lower surface,which has a prominent keeled midrib.Staminate flowers crimson, cylindrical, J inch long, surrounded at the base bytwo series of broadly ovate obtuse scales.Cones sessile or subsessile, brownish when mature, pubescent, cylindrical,tapering to an obtuse apex; 4 to 5 inches long by i^ to if inches in diameter.Scales: lamina three-sided, i inch wide by f inch long, upper margin almost entire,lateral margins nearly straight, laciniate; claw short, obcuneate. Bract minute,situated at the base of the scale, ovate, orbicular or rectangular, denticulate,emarginate with a short mucro. Seed with wing i£ inch long; wing two to threetimes as long as the body of the seed. In cultivated specimens the cones and scalesare usually considerably smaller than in wild trees.Cotyledons z six, convex and stomatiferous on the upper surface, flattish andgreen on the lower surface.HYBRIDSA series of hybrids have been obtained between A . Pinsapo and two otherspecies, A . cephalonica and A . Nordmanniana, of which a full account is given byDr. Masters in his valuable paper on hybrid conifers. 3i. Abies Vilmorini, Masters. 4 This is a tree growing at Verrieres near Paris,which has the following history. In 1867, M. de Vilmorin placed some pollen ofA. cephalonica on the female flowers of a tree of A. Pinsapo. A single fertile seedwas produced, which was sown in the following year; germination ensued and the1 The resin-canals in this species are variable in position. Cf. Guinier and Maire, in Bull. Sec. Bot. France, Iv. 190 (1908).2 Masters, in lift. * Jcurn. Key. Hort. Soc. xxvi. 99 seq. ( 1901). 4 Ibid. 109.Abies 733seedling was planted out in 1868. M. Phillipe L. de Vilmorin 1 states that the treewas in 1905, 50 feet high by 5 feet in girth; and has three main stems, one ofwhich, however, was broken by a storm two years ago. In its habit and foliage itresembles A. Pinsapo more than the other parent. The leaves, however, are longerand less rigid than in A . Pinsapo, and bear stomata only on their lower surface ;moreover their radial arrangement on the branchlets is imperfect. The cones, whichare produced in abundance and contain fertile seeds, resemble those of A . cephalomca,being fusiform in shape ; they have longer bracts than in A . Pinsapo, in some yearsexserted, in other years shorter and concealed between the scales. Seedlings raisedfrom this tree, now four years old, have acuminate sharp leaves like those ofA. cephalonica.2. Abies insignis, Carriere, Rev. Hort. 1 890, p. 230. This hybrid was obtainedin i or 1849 in the nursery of M. Renault at BulgneVille in the Vosges. Abranch of A . Pinsapo was grafted on a stock of the common silver fir (A. pectinata) \and after some years the grafted plant produced cones. Seeds from these weresown ; and of the seedlings raised one-half were like A . Pinsapo, the remainder beingintermediate in character, it was supposed, between A . Pinsapo and A . pectinata ;and the variation was considered to be the result of graft hybridisation. However,at no great distance there was growing a tree of A . Nordmanniana ; and it is moreprobable that the hybrid character of the seedlings was the result of a cross fromA. Pinsapo fertilised by the pollen of A . Nordmanniana. A complete account ofthese seedlings is given by M. Bailly. 23. Abies Nordmanniana sfieciosa, Hort. 2 This hybrid was raised in 1871-1872by M. Croux in his nurseries near Sceaux, the cross being effected by placing pollenfrom A . Pinsapo on female flowers of A . Nordmanniana. A full account of thishybrid is given by M. Bailly. 24. Mosers hybrids. Four different forms, all raised from A. Pinsapo, fertilisedby the pollen of A . Nordmanniana, which were obtained in 1878 by M. Moser atVersailles. Full details are given in Dr. Master's paper, to which we refer ourreaders.DISTRIBUTIONA. Pinsapo has a restricted distribution, being confined to the Serrania deRonda, a name given to the mountainous region around Ronda in the south of Spain.The late Lord Lilford informed Bunbury 3 in 1870 that he had seen it growing onthe Sierra d'Estrella in Portugal ; but we have not been able to confirm thestatement.There are three main forests of this species, none of considerable extent, occurringin localities at considerable distances apart. I visited these forests in December1906, and explain the rare occurrence of the tree as due to the fact, that in the dryclimate of the south of Spain, it can only exist on the northern slopes of mountain> fortus Vilmorinianus, 69, plate xii. (1906). 1902, p. 162, fig. 66.See also Card. Chron. 1 878, p. 438; Rev. Hort. 1889, p. 115, andRev. Hort. 1890, pp. 230, 231. 8 Arboretum Notes, 147 (1889).

734 The Trees of Great Britain and Irelandchains running due east and west; and these are seldom met with. In suchsituations the soil is never exposed to the direct rays of the noonday sun, andpreserves in consequence a great deal of moisture. The tree never grows evenon north-west or north-east slopes, and is strictly limited to aspects looking duenorth.The most important forest is in the Sierra de la Nieve, a few miles to the eastof Ronda. Here the tree extends for several miles in scattered groves on the northslope of the range, growing on dolomitic limestone soil, usually in gullies or underthe shade of the cliffs. It occurs mainly at elevations of 4000 to 5900 feet, thoughit occasionally descends to 3600 feet. In shaded situations and where the soil isdeep, there are dense groves of thriving trees, without any admixture of otherspecies; but at the lower elevations, where there is more sun, the trees are scatteredand mixed with oak and juniper. In exposed situations, at high elevations, the treesare windswept, stunted, and more or less broken. Seedlings are numerous in manyplaces. The largest trees, seen by me, were a group, on the road across themountain from Ronda to Tolox, at a spot called Puerto, de las animas. One ofthese (Plate 212) was 106 feet in height and 13 feet 8 inches in girth; and anotherwith a double stem, not so tall, girthed 16 feet 3 inches. This group is overhung bya precipice, and is at 4700 feet altitude. The stump of a tree, which had been cutdown, showed 240 annual rings and was 32 inches in diameter.The second forest, and by far the most picturesque, lies to the west of Ronda,on the northern slope of the precipitous peak, Cerro S. Cristoval or Sierra delPinar, close to the mediaeval town of Grazalema. The fir grows here on atalus, composed of sharp angular white limestone stones ; and the contrast betweenthe dense mass of green foliage of the tree and the pure white ground from which itsprings, is remarkably beautiful. The stones and pebbles are loosely aggregated ;and beneath the surface they are mixed with a mass of black mould, in which theroots of the tree freely spread. The fir extends along the precipitous side ofthe mountain for about two miles, forming a band of continuous forest, which reachesnearly to the summit of the peak, attaining about 5800 feet altitude, and descendinggenerally to 4000 feet, reaching in one gully to 3600 feet. Seedlings are numerous.There is no undergrowth, except an occasional daphne; but climbers like ivy andclematis are common. None of the trees are so tall as those in the Sierra de laNieve; but many have gigantic short trunks, in one case girthing 25 feet, and areextremely old. In this forest, trees with glaucous foliage, not seen elsewhere, are notat all uncommon.The third wood of A. Pinsapo occurs on the Sierra de Bermeja, which overhangsthe town of Estepona and the Mediterranean coast. This wood, which covers onlya small area, is most accessible from Gaucin, a station on the railway betweenGibraltar and Ronda. Here the soil is disintegrated serpentine rock, and thetree grows on the northern slope, between 4100 and 4900 feet, though stuntedspecimens occur up to 5400 feet. The fir is pure on the precipitous upper partof the mountain ; but lower down is mixed with Pinus Pinaster. The largest tree,which I measured, was 90 feet high by 13 feet 5 inches in girth.Abies 735Isolated groups of a few trees, the remains of former forests, are reported to begrowing on the Sierra de Alcaparain, near Carratraca, north-east of Ronda, and atZahara and Ubrique, not far from Grazalema. Mr. Mosley of Gibraltar, who gaveme valuable help and information, saw A . Pinsapo also growing on the SierraBlanca de Ojen near Marbella.HISTORY AND CULTIVATIONThis species was discovered by Edmond Boissier in 1837. He sent abouthalf-a-dozen seeds to M. de Vilmorin in the same year, and from one of these wasraised the very fine tree, which is now growing at Verrieres 1 near Paris, and whichis certainly the oldest cultivated specimen. This tree was in 1905, 70 feet high by7 feet 3 inches in girth. Abies Pinsapo was introduced into England in 1839 byCaptain Widdrington, 2 who was the first to obtain information about the existenceof a new species of Abies in Spain, though he was anticipated in its discovery byBoissier. (A. H.)In cultivation this has proved to be, all over the southern, midland, and easterncounties, one of the most ornamental of its genus, and is perfectly hardy on drysoils throughout Britain, ripening seed at least as far north as Yorkshire. It is oneof the few silver firs that seems to require lime to bring it to perfection, and thoughit will grow fairly well on sandy soils, it will not thrive without perfect drainage, oron heavy clay. It seems to have a great tendency to divide into several leaders andoften forms a bushy rather than a clean trunk, unless carefully pruned. It is notoften injured by spring frost, and, though not likely to have any economic value, isa tree that should be planted in all pleasure grounds on well-drained soil, andin a sunny situation.The seedlings which I have raised grow at least as fast as those of A . pectinata,and are hardier when young, but require five or six years' nursery cultivation beforethey are fit to plant out.The wood is soft and knotty like that of most of the silver firs when grownsingly in cultivation.REMARKABLE TREESThough specimens of this tree of from 50 to 60 feet high are found in manyplaces all over England, we have not measured any which are specially remarkable.The largest recorded at the Conifer Conference in 1891 was a tree reported to be62 feet high by 9 feet in girth, at Pampisford in Cambridgeshire; but thesemeasurements were erroneous, as it now is only 56 feet high by 7 feet 3 inches ingirth. Here there is a remarkable dwarf form 8 of this species, which is only afoot in height, with branches prostrate on the ground for 6 or 7 feet.The largest tree we know of is growing in a sheltered position in moist soil, atCoed Coch, near Abergele in North Wales, the residence of the Hon. Mrs. Brodrick.1 Hortus Vilmorinianus, 69, pi. 7 (1906).a Sketches in Spain, ii. 239.3 This is var. Hammondi, Veitch, Conifers, ed. i. p. 105.

w736 The Trees of Great Britain and IrelandThe gardener, Mr. Hunter, informs us that it is 82 feet high by 10 feet 2 inches ingirth ; but has never coned. There is a fine tree at Oakly Park, Ludlow, measuring70 feet by 5 feet 8 inches in 1908.At Hardwicke, Bury St. Edmunds, Sir Hugh Beevor measured a tree in 1904,which was 63 feet high and 8 feet 11 inches in girth. At Fornham Park, also inSuffolk, he found a tree, which was planted in 1866, 50 feet by 6 feet 7 inches ;and says that its growth kept pace with that of an Atlantic Cedar close by.Col. Thynne has taken a photograph of a narrow, pyramidal, symmetrical treeat Longleat, which was 65 feet high by 7 feet 9 inches in 1906 (Plate 213). AtDogmersfield Park, Hants, the seat of Sir H. Mildmay, I measured a well-shapedtree, 65 feet high by 6 feet 10 inches.There are several good trees at Lilford Park, Oundle, growing on oolitic limestone ; but Lord Lilford informed Henry that these were not raised from seedbrought home by his father, and could give no confirmation of Bunbury's statementthat the latter found the tree growing wild in Portugal.At Essendon Place, Hertford, a slender tree was 68 feet high by 5 feet i inchin 1907. At Merton Hall, near Thetford, Norfolk, there is a tall tree, 75 feet by5 feet 10 inches, the stem being bare of branches for 30 feet.At Highnam, Gloucestershire, Major Gambier Parry reports a fine specimen,growing in the pinetum, which measured 60 feet by 6 feet 8 inches in 1906.At the Rookery, Down, Kent, the gardener, Mr. E. S. Wiles, reported in 1906a fine specimen, 70 feet by 9 feet, which is growing on stiff yellow loam, intermingledwith flint and clay, resting on chalk.In Wales the best that I have seen is a tree at Bodorgan, Anglesey, the seat ofSir G. E. Meyrick, which in 1906 was about 70 feet high, and had some largewitches' brooms growing on it.In Scotland, we have seen none of more than average size, a tree at Sconebeing about the best, and, generally speaking, the climate seems too cold for this tree.Sir Archibald Buchan-Hepburn, however, reports one at Smeaton-Hepburn, EastLothian, which was 60 feet by 7 feet 9 inches in 1908.In Ireland, there is a tree at Curraghmore, Co. Waterford, which the gardener,Mr. D. Crombie, reported in 1905 to be 65 feet high by 8 feet in girth. AtCarton, the seat of the Duke of Leinster, there is a good tree, 54 feet high by8 feet in girth in 1903. At Coollattin, Wicklow, another was 55 feet by 4 feet10 inches in 1906.Prof. Hansen states 1 that fine trees of 50 feet high or more may be seen inseveral Danish gardens, where it has produced cones; and that the tree exists inthe south of Sweden and Norway.In the eastern United States it 2 never really flourishes, although it is possibleto keep it alive for many years in favourable situations, even as far north as easternMassachusetts.(H. J. E.)1 /. X. Hort. Soc. xiv. 476 (1892).2 Sargent, Silva N. Amer. xii. 100, adaot. ( 1898).Abies 737ABIES NUMIDICA, ALGERIAN FIRAbies numtdica, De Lannoy, ex Carrifcre, Rev. Hort. 1 866, pp. 106, 203 ; Van Houtte, flore des Serres,xvii. 9, t. 1717 (1867); Masters, Card. Chron. iii. 140 (in part and excluding figures) (1888);Trabut, Rev. Gen. Bot. i. 405, ff. 17, 18 (1889); Kent, Veitch's Man. Conifera, 5 29 (1900).Abies Pinsapo, Boissier, var. baborensis, Cosson, Bull. Soc. Bot. France, viii. 607 (1861).Abies baborensis, Letourneux, Cat. Arb. et Arbust. d'Algerie ( 1888).Pimis Pinsapo, Parlatore, DC. Prod. xvi. 2, p. 423 (in part) (1868).Picea numidica, Gordon, Pinet. 2 20 (1875).A tree attaining 70 feet in height and 8 feet in girth. Bark grey, smooth inyoung trees, becoming scaly and fissured on old trunks. Buds large, ovoid, acute atthe apex, non-resinous; scales ovate, acute, with white scarious margins, usually freeat the apex. Young shoots brown, shining, glabrous, with slightly raised pulvini butwithout grooves.Leaves on lateral branches pectinate below, the two lateral sets directed outwardsin the horizontal plane; those above shorter, crowded, directed upwards, and either,as on weak shoots, forming a narrow V-shaped pectinate arrangement, or, on strongshoots, with the median leaves directed backwards (not seen in any other species)and covering the upper side of the branchlet. Leaves short, ^ to f inch long, -fainch broad, linear, flattened, gradually tapering to the base, broadest above themiddle or uniform in width in the upper three-fourths, rounded at the apex, which issometimes entire but usually slightly bifid ; upper surface dark green, shining, withthe median groove often faint and rarely continued to the apex, in many leaves withfour to six broken lines of stomata in the middle line near the apex; under surfacewith two white bands of stomata, each of about eight to nine lines; resin-canalsmarginal. Leaves on cone-bearing branches all more or less upturned, those of themiddle ranks also directed slightly backwards, short, rigid, rounded and entire at theapex.Cones on short stout stalks, brownish, cylindrical with an obtuse apex andtapering base, about 5 inches long by \\ inch in diameter, with the bracts entirelyconcealed. Scales ; lamina fan-shaped, \\ inch wide, f inch long, upper marginalmost entire, lateral margins denticulate and either straight or with a wing on eachside above ; claw short, obcuneate. Bracts, scarcely reaching half the height of thescales, with a broad oblong claw and an expanded ovate denticulate lamina, whichis acuminate or cuspidate at the mucronate apex. Seed with wing about an inchlong; wing about i^ times as long as the body of seed. Cones of cultivated treeshave smaller scales with more developed lateral wings; and shorter bracts, scarcelyreaching % the height of the scale.The seedlings of this species have been fully described by Fliche. 1IVIn Bull. Soc. Forest. Franche-Contt et Belfort, 1 903, p. 168.E

738 The Trees of Great Britain and IrelandIDENTIFICATIONThe short broad leaves, which have usually four to six broken lines of stomataon their upper surface near the apex in the middle line, are a good mark of thisspecies. On strong shoots the backward direction of the median leaves, whichdensely cover the upper side of the branchlet, is also very characteristic.DISTRIBUTIONAbies numidica is very restricted in its distribution, being, so far as is known,confined to a small area towards the summit of the northern slope of Mount Babor, inthe Kabylie range in Algeria. It grows between 5000 and 6600 feet altitude in aclimate where snow lies upon the ground from December to April. In January,1907, I visited Kerrata, at the head of the famous gorge of Chaba-el-Akra; andfound that the ascent of the mountain, only a few miles distant, was impracticable.M. Bernard, Inspector of Forests at Bougie, who has charge of the forest of Mt.Babor, informed me that the northern slope contains an area of 4000 acres, and isclothed with a dense forest, composed mainly of cedar and Quercus Mirbeckii inthe upper zone between 4700 and 6600 feet, and of Q. Mirbeckii, Q. castanecefolia,and Acer obtusatum, in the lower zone below 4700 feet. The total number of treesof Abies numidica scarcely exceeds 3000; and they only grow towards the summit,where they occur scattered amongst the cedars and oaks. None of the trees aremore than 70 feet high, and the largest is only 8 feet in girth. The small size is dueto their exposed position, and possibly to the destruction of larger trees by thenatives in former times. Seedlings are rare ; and according to M. Bernard, this isaccounted for by the poor germinating quality of the seed, as only 4 to 15 per centof it produced plants with him. The soil on which the tree grows is limestone,its surface being composed of stones and pebbles, underneath which there is aconsiderable mixture of mould. 1Abies marocana, Trabut, 2 discovered in January 1906 by M. Joly, in themountains south of Tetuan, in Morocco, is intermediate in the characters of thefoliage between A . numidica and A . Pinsapo. M . Trabut showed me a branchlet,when I was in Algiers in 1907 ; but in the absence of cones, it is impossible todecide whether it deserves to rank as a new species. Seeds of this should bereadily procurable; and the attention of travellers is directed to the possibility ofintroducing a new silver fir. (A. H.)HISTORY AND CULTIVATIONThe Algerian fir was discovered in 1861 by Captain de Guibert. The firstseeds were sent to France in 1862 by M. Davout, a forest officer; and anothersupply and six young plants were forwarded in 1864 by M- de Lannoy.M. Maurice de Vilmorin, in Arbres Forestlers Strangers, 3 3 (1900) gives an account of Abies numidica on MountBabor. He noticed that many of the trees had short stout trunks, free of branches to 10 or 12 feet, occasionally more or lesstwisted, and often dividing into several stems." In Bull. Soc. Bot. France, liii. 154, t. 3 (1906). In the plate, the name Picea marocana, Trabut, appears by mistake.Abies 739The tree is rare in cultivation in England. There are two or three youngspecimens at Kew; and Kent, writing in 1900, mentions small trees, about 20 feethigh, growing at Bicton, and Streatham Hall in Devonshire.At Pampisford, Cambridgeshire, there are two trees with fine healthy foliage,the larger of which, 37 feet high and 3 feet 2 inches in girth, bore cones in 1907.There is also a specimen at Highnam 35 feet by 3 feet 2 inches. Though we havenot identified any specimens in Scotland Mr. Crozier speaks of it as a handsome andfree-growing tree which bore cones in 1906 and seems quite at home at Durris.In Ireland the finest we know of is at Fota, where a tree 39 feet by 6 feet wasbearing cones in 1908. Lord Barrymore informs us that it was planted in 1878.There is a good specimen at Glasnevin, 38 feet by 3 feet 7 inches in 1906; and oneat Castlewellan measured, in the same year, 25 feet by 3 feet.At Verrieres 1 near Paris, two trees, dating from the original introduction in 1862,were, in 1905, 46 feet in height by 4 feet 3 inches in girth. (H. J. E.)ABIES CEPHALONICA, GREEK FIRAbies cephalonica, Loudon, Arb. et Fntt. Brit. iv. 2325 (1838); Masters, Card. Chron. xxii. 592,f. 105 (1884); Kent, Veitch's Man. Conifera, 498(1900); Halacsy, Consp. Fl. Grcecce, iii.450 (1904).Abies Apollinis, Link, Linnaa, xv. 528 (1841).Abies Regina Amalia, Heldreich, Gartenflora, ix. 313 (1860).Abies Panachaica, Heldreich, Gartenflora, x. 286 (1861).Picea cephalonica, Loudon, Card. Mag. xv. 238 (1839).Pinus cephalonica, Endlicher, Cat. Hort. Vindob. 1 218(1842).A tree attaining about 100 feet in height. Bark greyish brown, smooth inyoung trees, in old trees fissuring into small oblong plates. Buds conical or ovoid,obtuse at the apex, composed of thick ovate acute keeled scales, with prominent tips,and covered with a layer of resin. Young shoots smooth, light brown, shining,glabrous.Leaves on lateral branches radially arranged, but not so regularly as inA. Pinsapo, their apices pointing outwards and slightly forwards, those of theupper ranks shorter than those beneath. Leaves linear, flattened, curved, abouti inch long, TV to ^ inch broad, abruptly tapering at the base, narrowing graduallyin the anterior two-thirds, and ending in a long cartilaginous point; upper surfacedark green, shining, with the median furrow not continued to the apex, and usuallywith several broken lines of stomata ; lower surface with two white bands of stomata,each of seven or eight lines; resin-canals marginal. Leaves on cone-bearingbranches all upturned, curved, rigid, broad, with the apex simply acute and notprolonged into a fine cartilaginous point.Cones, on short stout stalks, about 6 inches long by \\ inch in diameter,cylindrical, slightly tapering at both ends, brownish, with the bracts golden brown,exposed, and reflexed. Scales: lamina narrowly fan-shaped, almost triangular;1 Hortus Vilmorinianus, 69, pi. I (1906).

I740 The Trees of Great Britain and Irelandupper margin convex, undulate or entire ; lateral margins with two short denticulatewings ; base curving but not auricled on each side of the oblong claw. Bracts : clawoblong, ^ inch wide, extending f the length of the scale; lamina lozenge-shaped, %inch wide, denticulate, ending in a triangular mucro, exserted and reflexed over theedge of the scale next below. Seed-wing about twice as long as the seed; seedwith wing about i inch long.Seedling ; x caulicle tapering upwards, reddish brown, erect, stout. Cotyledons,five or six, acute not mucronate, about i£ inch long ; upper surface dotted irregularlywith stomata and grooved in the middle line. Primary leaves half the length of thecotyledons, not mucronate ; lower surface with stomata.Var. Apollinis, Beissner, Nadelholzkunde, 440 (1891).Abies Apollinis, Link, Linnaea, xv. 5 28 (1841).This variety differs from the type in the arrangement and shape of the leaves.On lateral branchlets the radial arrangement is imperfect, most of the leaves standingcrowded on the upper side of the branchlet, with their apices directed upwards, thosein the middle line straight and vertical, those on the sides curved and bendingupwards ; on the lower side of the branchlet a few leaves are directed downwards andforwards. Leaves thicker and broader than in the type, about i£ inch long by ^ to înch broad, ending in a short acute point, bevelled off from behind ; upper surfacewith a continuous median groove and two to three short lines of stomata near thetip; lower surface with two bands of stomata, each of ten lines.The cones do not differ in any essential characters from those of the type; andthe differences noted by Murray 2 in the broader bract and expanded wing of theseed are trifling and inconstant.Halacsy considers A bies Regince Amalifs and Abies panachaiaca to be meresynonyms of A bies cephalonica; and only allows the variety Apollinis, distinguished,according to him, by its acute leaves, those in the type ending in an acuminate orvery sharp spine-like point. According to other authorities, A . Regince Amalice ismore akin to var. Apollinis than to the type. In all probability there is a seriesof intermediate forms connecting the type and var. Apollinis:1'DISTRIBUTIONAccording to Halacsy this species occurs in the sub-alpine region of almost allthe higher mountains of Greece, between 2700 and 5700 feet elevation. The typeis met with in the island of Cephalonia on Mount Enos ; and on the mainlandin Doris on Mount Kiona, in Attica on Mount Parnes, and in Arcadia onMasters, m MS., who states that in var. Apollinis the cotyledons are seven in number, sub-acute at the apex, andabout I inch long; primary leaves shorter and more pointed than the cotyledons.* Proc. Roy. Hort. Sac. iii. 141 (1863).Guim« and Maire, in Bull. Soc. Bot. France, Iv. 187, figs. 2 and 3 (1908), describe a variety, with leaves like thoseof A . cilicua, which grows on Mount Pjndus in Thessaly.AbiesMounts Msenalus, Madara, Thaumasion, and Rhudia. Var. Apollinis occurs, inEpirus on Mounts Tsumerka, Strungula, Peristeri; in Thessaly, on Pindus andOlympus; in Eubcea on Mount Dirphys. It has also been found in Hellas onMounts CEta, Tymphrestus, Parnassus, Helicon, Cithseron, Pateras, and Parnes;and in Peloponnesus on Mounts Chelmos, Olenos, Malevo, Taygetos.As Halacsy considers A bies Regince Amalice to be the same as the type, and notthe var. Apollinis, his account of the distribution differs from that generally adopted,in which the view taken is that the type is confined to the island of Cephalonia, andthat all the continental forms are referable to the var. Apollinis?In Cephalonia the forest of this species occurs on Mount Enos, along a ridge4000 to 5000 feet above sea-level and about 12 to 15 miles in length. It was 36miles in circumference in 1793 ; but its area was considerably reduced by disastrousfires in 1798. No recent account of this forest, of which full details were given byGeneral Napier in 1833, nas come under our notice.The form which occurs in the mountains of Arcadia, distinguished as var.Regince Amalice'2' by some authors, is remarkable for its capacity of producing coppiceshoots, when the trunk is felled; and the main stem, even when untouched, is saidoften to produce secondary stems and branches from the old wood. (A. H.)CULTIVATIONSeeds 3 were first sent from Cephalonia to England by General Sir CharlesNapier in 1824; and the first plants, few in number, were raised by Mr. C. Hoareof Luscombe Castle, who distributed them to various places. 4 Some time afterwards Mr. Charlwood 5 sold seeds to the public, having received a cask of conesfrom General Napier.The form Regince Amalice was first noticed in 1856 by Schmidt of Athens,who found a forest of this tree near Tripolitza in Central Arcadia; its seeds haverecently been introduced abundantly.A. cephalonica seems to be quite hardy over the greater part of Great Britain,but it is rather more susceptible to spring frosts than A . Pinsapo, because it startsearlier into growth, and on this account should not be planted in low, damp, orexposed places. It seems to grow on limestone, but not to be so distinctly a limelovingtree as A. Pinsapo. It ripens seeds in good years in the south of England,but the seedlings which I have raised do not grow so fast as those of A . Pinsapo?1 With regard to the occurrence of this variety in Roumelia, Macedonia, and Thrace, see our remarks on p. 722concerning the distribution of A. pectinata in the Balkan peninsula.1 See Regel, Gartenflora, ix. 299, fig. (1860); and Seemann, Card. Chron. 1861, p. 755 fig.3 London, loc. cit.4 A list of these places is given in Loudon, Card. Mag. 1838, p. 31, and in Pinetum Britannicum, ii. p. 179." Loudon, Card. Mag. 1 839, p. 238.6 Owing to its susceptibility to late frosts and to attack by Chermes, it is now nearly impossible to grow this tree up to aplanting size. Its timber, when closely grown and of some age, is, in my opinion, the best of the European silver firs, beinghard, close in texture, and heavier in a dry state than any I have yet handled. Var. Apollinis is less subject to injury byfrosts and attack by Chermes than the above, and seems well adapted for planting in the north of Scotland. In cultivationit maintains a more conical outline, and is easily distinguished from the type. (J. D. CROZIER.)

The Trees of Great Britain and IrelandREMARKABLE TREESProbably the finest tree in the British Isles is the one growing at Barton, BurySt. Edmunds, which in 1908 was 95 feet in height by 13 feet 3 inches in girth. Thistree (Plate 214) is very symmetrical, branched to the ground, and in full vigour,though probably it has nearly attained its limit of height, as the top of the crown offoliage is flattened. This is one of the original plants raised at Luscombe, andwas planted at Barton in 1838, being then about thirteen or fourteen years old.According to Bunbury, 1 it did not suffer in the slightest degree from the severewinter of 1860. In 1857, it was 35 feet in height; and in 1858 began to bear cones,which are confined to the topmost of the lateral branchlets. In 1867, the heighthad increased to 58 feet, and the trunk at three feet from the ground was 7 feet ingirth. Seedlings have been frequently raised from its seed. One of these seedlings,which was sent many years ago to Lord Rayleigh, is now growing at Terling Place,Essex, and measures 53 feet high and 3 feet 3 inches in girth.Another of the original trees is now growing at Luscombe Castle, near Dawlish,in a rather exposed place, about 200 feet above sea level; when I saw it, in April1908, it was a healthy and well-shaped tree, 75 feet by n feet.There is a very fine healthy tree at Blount's Court, Oxfordshire, which Henrymeasured in 1907, as 87 feet in height by 10 feet 8 inches in girth. Another plantedat the Coppice, Henley, in 1860, measured in 1905, 62 feet high by 8^- feet in girth.At Pampisford, Cambridgeshire, there are two trees, the larger of which was, in 1908,55 feet by 6 feet i inch. The Cephalonian fir has been largely planted on LordWalsingham's estate at Merton, Norfolk, the largest specimen, 52 feet by 9 feet7 inches, dating from 1852. On the Thetford road there is an avenue of these trees,growing in loose, shallow sand, which have attained at forty-eight years old anaverage girth of 8^- feet. The growth of the tap-root is stopped by the compactchalk sub-soil, wide-spreading horizontal roots being formed, which have no greathold in the shifting sand ; and several trees have been uprooted by storms.At Heron Court, near Christchurch, I measured in 1906 a very large tree withragged top, 82 feet by 10 feet 8 inches. At Beauport, Sussex, there is a good tree,which in 1904 was about 80 feet high by 10 feet 3 inches in girth.At Powderham there is a very large and spreading, but ill-shaped tree, whichappears as though in the mild, soft climate of south Devon it would not be longlived. In 1892 it was recorded as the largest in Great Britain, being then 77 feetby 11 feet at 3 feet from the ground.At Killerton there is a large tree which measured in 1903 80 feet byii feet 9 inches. It forks at about 25 feet. At Highclere another, in the sameyear, measured 75 feet by n feet. At High Canons, Herts, Mr. H. Clinton Bakermeasured a tree in 1908, which was 58 feet by 8 feet 3 inches. At Bayfordbury,a tree planted in 1847 was 7 feet by 6 feet n inches in 1905.At Castle Kennedy there is a very wide-spreading tree, which in 1904 measuredAbies 74359 feet by 9 feet 8 inches. Around it were several natural seedlings, from i foot to5 feet in height. At Smeaton-Hepburn another measured, in 1905, 53 feet high by10 feet in girth. A number of Cephalonian firs were planted at Blairadam, the seat ofSir Charles Adam, Bart, in Kinross-shire, by his ancestor Sir Frederick Adam, whowas governor of the Ionian Islands in 1824, and who was censured by GeneralNapier for not sufficiently protecting the forests in Cephalonia. Several of thesetrees still survive at Blairadam, the largest in the garden near the entrance gate being49 feet high, and 8 feet 2 inches in girth at 4 feet. It divides into several stems atabout 25 feet. Another measures 42 by 5^ feet, and there are several smaller ones,but the tops in most cases have been at various times injured by wind and frost. Inother parts of Scotland the tree grows fairly well, but not so fast as in the south, thebest I have heard of being at Abercairney, where Mr. Bean 1 records one 75 feet highin 1906. As this, however, was in 1892 only reported as 50 feet high there may bea mistake. Other good trees are growing at Whittingehame, East Lothian, at HaddoHouse, Aberdeenshire, and at Ochtertyre, 1 Perthshire.In Ireland, the largest Cephalonian Fir known to us, is growing at Adare Manor,Co. Limerick, the seat of the Earl of Dunraven; and, in 1903, was 86 feet high by9 feet 4 inches in girth.At Powerscourt, Co. Wicklow, a tree measured, in 1903, 55 feet by 8 feet 9inches; and at Hamwood, Co. Meath, there is a fair specimen which in 1904 was50 feet by 9 feet 6 inches.At Cahir Park, Co. Tipperary, there are four trees of nearly equal size, onemeasuring 46 feet by 6 feet 2 inches. Specimens sent in 1906 by Mr. AustinMackenzie show that these trees belong to var. Apollinis.In the Botanic Garden at Bergielund, near Stockholm, a tree, planted in 1890,was, when seen by Henry in August 1908, 30 feet in height and i foot in diameter,and exceeds in rapidity of growth all the other conifers in the garden. In theBotanic Garden, at Christiania, there is a tree, about 25 feet in height, which is,however, not quite hardy, being slightly browned by frost. Hansen 2 says that thisspecies had attained in 1891 a height of 44 feet and a girth of 6 feet, at 40 yearsold, in the gardens at Carlsberg, near Copenhagen.A. cephalonica has proved hardy 8 in eastern Massachusetts, where it hasalready borne cones.Though General Napier stated that the wood of this tree in Cephalonia isvery hard and durable, yet as grown in this country it is not likely to have anyeconomic value, as it is too knotty and coarse for any but the commonest purposes.(H. J. E.)' Ktw Bulletin, 1 906, pp. 266, 267. 2 Journ. Roy. Hort. Sac. xiv. 463 (1892).3 Sargent, Sihia N. Amer. xii. 99, adnot. ( 1898). Sargent, however, states in his account of the Pinetum at Wellesleyin 1 905, p. 12, that the tree here, which is 51 feet by 6 feet, was considerably injured in the severe winter of 1903-4.1 Arboretum Notes, 1 44.

744 The Trees of Great Britain and IrelandABIES CILICICA, CILICIAN FIRAbies cilicica, Carrifcre, Conif. 2 29 (1855), and flore des Serres, xi. 67, t. 1108 (1856) ; Tchihatcheff,Asie Mineure, ii. 494 (1860); Heuzd in Rev. Hort. 1 856, p. 81, f. 14; Kent, Veitch's Man.ConifercR, 500 (1900); Hickel, in Bull. Soc. Dend. France, 1 908, p. 183.Abies selinusia, Carrifcre, Flore des Serres, xi. 69 (1856).Pinus cilicica, Kotschy, Oestr. Bot. Wochenbl. iii. 409 (1853).Picea cilicica, Gordon, Pin. Suppl. 50 (1862).A tree attaining in Asia Minor 100 feet in height and 7 feet in girth. Barkashy-grey in colour, smooth in young trees, deeply fissured and scaly in old trees.Buds * small, non-resinous, ovoid, acute at the apex ; scales few, keeled, with theirtips more or less free and not appressed. Young shoots smooth, greyish-brown, withscattered short erect pubescence ; bark fissuring slightly on the second year'sshoot.Leaves on lateral branches usually pectinately arranged, the upper ranks pointingoutwards and upwards, thus forming a V-shaped depression above between the twolateral sets ; on vigorous shoots, the median leaves on the upper side are directedforwards and upwards, and cover the branchlet, the V-shaped depression beingobliterated. Leaves thin and slender, i to \\ inch long, ^ inch wide, linear,flattened, uniform in width except at the tapering base, apex rounded or acute andslightly bifid ; upper surface light green with a continuous median groove andusually without stomata, rarely with two to three short lines in the groove near theapex ; under surface with two narrow greyish bands of stomata, each of six to sevenlines ; resin-canals marginal. Leaves on cone-bearing branches, upturned, curved,more rigid and broader than those on barren branches, minutely bifid at the truncateor obtuse apex.Cones of wild trees subsessile or on short stout stalks, cylindrical,tapering to an acute apex, 6 to 9 inches long by 2 to 2^ inches in diameter,brownish when ripe. Scales 2 larger than in any other species ; lamina if inchwide, -f inch long, fan-shaped, upper margin thin and entire, lateral margins convex,denticulate, with a sinus on each side ; claw short, obcuneate. Bract with anoblong claw, expanding above into an ovate or quadrangular denticulate lamina,tipped with a short mucro, extending to £ or ^ the height of the scale. Seed-wingabout i^ times as long as the seed ; seed with wing about i|- inch long. Incultivated specimens, scales smaller, i^ inch wide by f inch long ; bracts with avery short claw and a lamina not reaching more than \ the height of the scale ; seedwith wing about i inch long.DISTRIBUTIONThis species is confined to Asia Minor and northern Syria, occurring on theLebanon and the Antitaurus, and forming, in company with the cedar, great forests1 The buds are characteristic ; and, as Hickel points out, distinguish this species from all the others.a The peculiar hook-like processes of the scales which occur in some specimens are probably abnormal.Abies 745in the Cilician Taurus. It was first discovered by Kotschy 1 in the Cilician Taurusin August 1853, in the valley of Agatsch Kisse, at an elevation of 4000 to 5000 feet.It is known to the Turks as Ak Illeden, white fir, and grows in thick forestssometimes unmixed with other trees, sometimes in company with oaks, cedars, andjunipers. Yew and Pinus Laricio also occur in these forests, which are protectedfrom woodcutters by their inaccessibility. The climate of these mountains isextremely hot in summer, and cold in winter, with much snow in the upper region.Post 2 says that it is found in alpine and subalpine Lebanon, and in the AmanusMountains in the extreme north of Syria, but does not give any details of its size orthe elevation at which it grows.CULTIVATIONThe first seeds, received by the museum at Paris in 1854 from M. Blanch,French Consul at Saida, failed to germinate. Balansa sent a good supply of seed in1855. From these or from Kotschy's seed the few trees which we have found inEngland were probably raised.The Cilician fir is extremely rare in cultivation in this country. The best specimen we have found is a tree, growing at Welford Park, Newbury, which in 1908 was51 feet by 4 feet 4 inches. Mr. Ross, the gardener, informs us that he found this treeas a small plant in a pot, when he came to Welford Park in May, 1860. The treehas been considered by many people to be Abies homolepis, and was figured in theGarden, for 1904, under that name. It is unquestionably, however, Abies cilicica,of which it has the foliage, and only differs slightly from wild specimens in thesmaller size of the cones and scales. Mr. A. B. Jackson has identified two atBicton, 48 feet by 4 feet and 47 feet by 3^ feet respectively; and another at theHeath, Leighton Buzzard, which is 48 feet by 3 feet 10 inches.The finest in Scotland is a tree at Durris, which Mr. Crozier reports to be55 feet high and 5 feet 8 inches in girth. It was incorrectly labelled A . amabilis.Another good specimen is growing at Castle Kennedy, which measured in 1904,48 feet in height by 5 feet i inch in girth. A second tree here, not so tall, is verythriving. Kent mentions a tree at Rossdhu in Dumbartonshire.A tree at Glasnevin was 34 feet by 3 feet 2 inches in 1907; and a specimen, atPowerscourt, 37 feet by 2 feet 8 inches in 1906, did not seem to be very thriving.There is a good specimen at Verrieres, near Paris, of which a figure is given byM. Philippe L. de Vilmorin in Hortus Vilmorinianus (plate i). This tree is about60 feet high. Another and slightly taller tree is growing in the Pare de Cheverny,in the department of Loir et Cher. Parde says that at Harcourt (Eure) itreproduces itself naturally.According to Sargent, 8 Abies cilicica, with the exception of Abies concolor, is themost beautiful of those silver firs, which are perfectly hardy and satisfactory inthe north-eastern states of the U.S. Some trees are 40 feet in height, notably at1 Seise in den Cilicischcn Taurus ( 1858).IVa Flora of Syria, p. 751. 3 Silva N. Amer. xii. 99 adnot. ( 1898).F

746 The Trees of Great Britain and IrelandMr. Hunnewell's pinetum, Wellesley, Mass. ; l Mr. Hall's garden, near Bristol, RhodeIsland; and Mr. Hoope's pinetum, West Chester, Pennsylvania. Sargent statesthat the tree does not thrive in western Europe, as the young shoots, which appearearly in the spring, are killed by late frosts; and in consequence it is not propagatedby nurserymen. Seeds from wild trees are difficult to procure. (A. H.)ABIES NORDMANNIANA, CAUCASIAN FIRAbies Nordmanniana, Spach, Hist. Vfg. xi. 418 (1842) ; Regel, in Gartenflora, xx. 259, t. 699 (1871) ;J. D. Hooker, Bot. Mag. t. 6992 (1888); Masters, Card. Chron. xxv. 142, f. 30 (1886); Kent,Veitch's Man. Coniferce, 5 26 (1900).Pinus Nordmanniana, Steven, Bull. Soc. Nat. Moscow, xi. 45, t. 2 (1838); Loudon, Card. Mag.xv. 225 (1839).Picea Nordmanniana, Loudon, Encycl. Trees, 1 042 (1842).A tree attaining in the Caucasus over 200 feet in height and 15 feet in girth.Bark in cultivated trees greyish brown, smooth when young, becoming slightlyfissured in older trunks. Buds ovoid, acute at the apex, brown, non-resinous, withovate, acute, slightly keeled scales. Young shoots grey, smooth, with very scatteredshort erect pubescence.Leaves on lateral branchlets, pectinately arranged below, the two lateral setsspreading more or less in the horizontal plane ; those above shorter, directed forwardsand densely covering the branchlet in imbricated ranks. Leaves linear, flattened,about i to \\ inch long, ^ to ^ inch wide, uniform in width except at the graduallytapering base; apex rounded and bifid; upper surface dark green, shining, with acontinuous median groove and without stomata; lower surface with two conspicuously white bands of stomata, each of eight or nine lines ; resin-canals marginal.Leaves on cone-bearing branches all curved and upturned.Staminate flowers ovoid-cylindric, ^ inch long, each with three series ofinvolucral bracts.Cones sub-sessile, cylindrical, tapering at both ends, about 6 inches long by 2inches in diameter, brown in colour, with the bracts exserted and reflexed. Scales :lamina, about \\ inch wide by f inch long, either with a denticulate wing on eachside or with straight lateral margins; claw obcuneate. Bract with oblong claw,expanding above into an almost orbicular lamina, which is denticulate and tippedwith a long triangular mucro. Seed with wing about an inch long, the wing beingtwice the length of the body of the seed.VARIETIES AND HYBRIDSSeveral varieties are mentioned by Beissner, which are said to differ from thetype in foliage, the leaves being shorter, glaucous, or yellow in colour. None ofthese appear to be in cultivation in England.1 Elwes saw this tree in May 1905, and remarked that it was very similar in growth to A . Nordmanniana, which hasshorter darker leaves and denser habit. It had not suffered from the severe frost of the preceding winter which in some placeshad injured the Caucasian fir. According to Sargent, The Pinetum at Wellesley in 1905, p. 12, this tree is 49 feet high and5 feet in girth.Abies 747Var. equi-Trojam, Guinier and Maire. 1 A peculiar form, discovered bySintenis on Mount Ida in north-west Anatolia. It has reddish-brown glabrousshoots, leaves acute at the apex and only slightly emarginate, and cones with bractsmuch exserted and almost concealing the scales.The hybrids, which have been obtained between A. Nordmanniana and A.Pinsapo are dealt with in our article on the latter species.DISTRIBUTIONThis species is a native of the mountains in the southern and south-easternshores of the Black Sea, including the western spurs of the Caucasus. According toRadde, 2 it is entirely absent from the eastern parts of the Caucasus and Talysch, itseasterly limit being longitude 42 . It usually grows between 3000 and 6600 feetelevation, and either forms pure forests or is associated with Picea orientalis, beingoccasionally mixed both with that species and Pinus sylvestris. I t is said to prefercalcareous soil and to be dominant on the limestone formations, which are not sofavourable to the growth of the oriental spruce and the common pine. In pureforests, the trees stand very close together; and in their deep shade underwood isabsent and no light reaches the ground, which is very dry and covered with a thicklayer of brown needles. Such forests are the last hiding-place of the European bisonin a truly wild condition.The largest tree mentioned by Radde, the age of which is not given, grew in thevalley of the Labba in the district of Kuban, and measured 213 feet in height and 15feet in girth at breast height, and the stem alone had a volume of 1236 cubic feet.On an area of about 2^ acres in this forest fifteen trees nearly as large weregrowing. It thrives best and attains its largest size at high elevations, 5000 to6000 feet; where stems 150 to 170 feet in height, with a girth of 10 feet, are quitecommon. The oldest tree, which is recorded by Radde, was 370 years old, andmeasured 170 feet high by 10 feet in girth.Abies Nordmanniana was also found by Balansa 8 in Lazistan, and by Sintenis 8at Kostambul in Paphlagonia. Guinier and Maire 4 in 1904 found it growing onMount Olympus in Bithynia, where, on the northern slope between 3700 and6000 feet, it forms extensive forests, either pure or mixed with Pinus Laricio, beech,oak, and chestnut, and constituting the timber line at 6300 feet. These botanistsstate that on Olympus, as well as in the Caucasus, it is a light-demanding tree, aleast in the young stage, as the seeds everywhere germinate in open and unshadedplaces. The discovery by these authors of A . Nordmanniana on Mount Olympusand of the van equi-Trojani on Mount Ida extends the distribution of this specieswestwards through northern Asia Minor to the borders of the îgean Sea.1 In Bull. Soc. Bot. France, Iv. 186, fig. I (1908). This variety was referred to A. pectinata by Boissier, in Fl. Orientalis,v. 701 (1881).2 Pflanzenverb. Kaukasusland, 1 84, 222, 244, etc. (1899). 3 Specimens in Kew herbarium.4 In Bull. Soc. Bot. France, Iv. 185, fig. I (1908). The silver fir on Mount Olympus was erroneously identified withA. pectinata by Boissier in Flora Orientalis, v. 701 (1881).

748 The Trees of Great Britain and IrelandAbies Nordmanniana was first recognised as a distinct species by the Finnishbotanist, Nordmann, Professor at Odessa, whose name it bears. He found it in 1836in the Caucasian province of Imeretia. Pallas and other early botanists had referredthe Caucasian silver fir to Abies pectinata. I t was introduced 1 into Europe in 1848,when Alexander von Humboldt obtained seeds from the Caucasus, which were sownin the Berlin Botanic Garden. (A. H.)CULTIVATIONNo other silver fir found in the Old World is more thoroughly at home inGreat Britain, for it grows luxuriantly on soils where the common silver fir willnot thrive; is absolutely uninjured by spring frost, even in a young state, and ripensseed as far north as Perthshire and County Down. It seems equally at home onrich loam in the south-east of England, on oolite gravel in the Cotswold Hills, andin the peaty soil and wet climate of Argyllshire. Out of 102 returns sent to theConifer Conference from all parts of Great Britain, 78 mention this tree and nearlyall speak well of it, though it is said to fail at about thirty years old on strong loamin Worcestershire, and to be liable to aphis at Durris in Kincardineshire. 2Sir Herbert Maxwell 3 states that the Crimean silver fir (a misleading name, asit does not occur wild in the Crimea), after it attains twenty to thirty years of age,frequently succumbs to the attacks of aphis, and gives as an instance in proof of this,that at Benmore, where large numbers were planted thirty to forty years ago, veryfew now remain. But I do not think that this is a fair example, as the climate ofBenmore is very wet, and the soil in many places very shallow. In the warmerand drier parts of Scotland I have seen many flourishing specimens, though not sofine as in England.Wilkie 4 says that at Tyninghame, in East Lothian, it is later in starting growththan the common silver fir, grows more freely when young, and either for use orornament is certainly the more valuable of the two. Webster, also, whose experiencewas gained in Ireland, North Wales, and Kent, says, 5 " If A . nobilis be the best ofthe Californian silver firs, this is without doubt the finest of the European or Asiaticspecies." He expected that at no distant date it would supplant the silver fir forforest planting, the timber being of excellent quality, the tree more ornamental, andless exacting as regards soil. He says that it succeeds well on reclaimed peat bog,stiff loam, decomposed vegetable matter, and light gravelly soils.1 Hansen, \i\fourn. Roy. Horl. Sac. xiv. 471 (1892). In the catalogue of the Pinetum at Beernem, in western Flanders,Baron Serret says that he received his specimen in April 1847, from Lawson and Son, Edinburgh ; and the earliest introduction would seem from this to have been prior to that stated by Hansen.2 A. Nordmanniana, the most susceptible of all silvers to attack by Chermes either in a seedling or older state. Forgeneral purposes this tree is doomed, and it is only by repeated spraying with insecticide that it will be possible to preserveeven the largest specimens. In growth, it has proved itself much slower than A . pectinata, and being densely branched and ofa shade-bearing nature, its timber when cut up has generally been coarse and knotty. In Scotland it has never beenregarded by foresters as of economic importance. (J. D. CROZIER.)3 Green's Encyclopedia of Agriculture, ii. 112 (1908). The erroneous statement that this fir occurs wild in the Crimeaappears to have been first made in Veitch's Man. Coniferie, 1st ed. 102 (1881), and has been repeated by Masters, Hansen,the Kew Handlist of Conifers, etc. No species of Abies grows wild in the Crimea. Cf. Demidoff, Voyage RussieMeridionale et la Crimie, ii. 231, 232, 375, 646 (1842).4 Trans. Royal Scot. Arb. Sac. xii. 211 (1889). 6 Ibid. 257.Abies 749As the seed can now be procured in quantity and at a cheap rate, even whenhome-grown seed is not available, there seems to be no reason why this beautifultree should not be raised at the same rate as the common silver fir and planted inpreference to the latter, for though it has not yet had time to attain its full size inthis country it grows quite as fast, and from what little we know of its timber islikely to be at least as valuable. Its average rate of growth is from i to 2 feetannually when once established ; and though we have as yet no evidence that itwill endure dense shade as well as the silver fir, yet the accounts of its growthin the Caucasus lead one to expect that it will do so.REMARKABLE TREESAmong the numerous specimens that we have measured in various places inEngland, I have seen none to surpass a very healthy and vigorous tree which growsin a wood facing east on the banks of the river at Eggesford, the property of theEarl of Portsmouth in Devonshire, which in April 1904 measured 84 feet by 5 feet 7inches, and had produced cones. But a tree growing in a wood called Hook's Groveat Bayfordbury is perhaps taller; it was about 85 feet by 6 feet 10 inches in 1907.At Strathfieldsaye, in the same year, I measured one as 78 feet by 6 feet 7inches, and at Hemsted, in Kent, there is a tall but very slender specimen, not overforty years planted, which bids fair to become a very large tree. In 1905 it was 68feet by only 3 feet 7 inches. At Lynhales, Herefordshire, the seat of S. Robinson,Esq., another is 70 feet by 5^ feet and growing freely.In Wales it is thriving at Penrhyn; where there are two trees, one with its topbroken being about 75 feet by 10 feet; the other even taller measures 6 feet 10inches in girth; and at Hafodunos, where it does well in plantations, Henrymeasured one 60 feet by 6 feet 7 inches in 1904.In Scotland the largest recorded in 1891 was at Poltalloch, and then was saidto measure 70 feet by 6 feet, but when measured by Mr. Melville in 1906 he madeit only 73^ feet by 7 feet 4 inches.The finest I have seen myself is one at Moncreiffe which, in 1907, I made to beno less than 79 feet by 6^ feet; a healthy tree from which many seedlings have beenraised. This is stated by Hunter to have been planted about 1856, and in 1888 wasonly 30 feet by 2 feet 2 inches. It is said to have been hybridised by the silver fir,but I could not see anything in the seedlings to distinguish them.In Ireland it also grows very well. A tree at Carton, the seat of the Duke ofLeinster, was 74 feet by 5 feet 4 inches, and one at Fota 68 feet by about 6 feetin 1903. Another at Mount Shannon, Limerick, measured, in 1905, 75 feet by 8feet 9 inches. A good specimen at Ballykilcavan, Queen's County, measured 68 feetby 5 feet 2 inches in 1907. There are many fine healthy specimens at Dereen inCo. Kerry.In the University Botanic Garden at Upsala, in Sweden, a tree was seen byHenry in 1908 which was about 40 feet high and branched into three stems near theground, the result evidently of injury to the leader by severe frost in early youth.

75 The Trees of Great Britain and IrelandAccording to Hansen, 1 it is said to thrive at Trondhjem in Norway, but Henry sawno specimens at Trondhjem or Christiania. It is often planted in Danish gardensand forests, and is quite hardy in Denmark.According to Sargent, 2 it is very hardy in the eastern United States, as farnorth, at least, as eastern Massachusetts, but although handsome when young, is aptto become thin and shabby here at an early stage. (H. J. E.)ABIES WEBBIANA, HIMALAYAN FIRAbies Webbiana, Lindley, Penny Cyclop, i. 30 (1833); Griffith, Icon. As. PI. t. 371 (1854); Masters,Card. Chron. xxii. 467, f. 86 (1884), and x. 395, f. 47 (1891); Hooker, Card. Chron. xxv.788, ff. 174, 175 (1886), and Flora Brit. India, \. 654 (1888); Kent, Veitch's Man. Coniferee,543 (1900); Gamble, Indian Timbers, 7 18 (1902).Abies spectabilis, Spach, Hist. Vlg. xi. 422 (1842).Abies Mariesii, Masters, Bot. Mag. t. 8098 (1906) (not Masters, Card. Chron. xii. 788 (1879)).Pinus Webbiana, Wallich, ex Lambert, Genus Pinus, 7 7, t. 44 (1828).Finns spectabilis, D. Don, Prod. fl. Nepal. 5 5 (1825).Picea Webbiana, London, Art. et Frut. Brit. iv. 2344 (1838).A tree, attaining in the Himalayas 150 feet or more in height and 35 feet 8 ingirth, with thick spreading horizontal branches ; bearing a flattened crown of foliage.Bark speedily scaling on young stems ; on old trunks, greyish brown, rough,irregularly fissured and very scaly. Buds large, globose, brownish, covered withresin, which conceals the keeled obtuse scales. Young shoots reddish brown withprominent pulvini, separated by deep grooves ; pubescence short, erect, reddish,confined to the grooves and not spreading over the pulvini. In the second year'sshoot the pulvini and grooves are more marked, the pubescence being retained.Leaves on lateral branchlets pectinately arranged, in two lateral sets, each ofseveral apparent ranks; the lower ranks on each side extending outwards in thehorizontal plane ; the upper ranks, with leaves becoming gradually shorter, directedoutwards and upwards, and forming a V-shaped depression, in the bottom of whichthe upper side of the branchlet is plainly visible. Leaves i to z\ inches long, £$inch wide or more, linear, flattened, uniform in width except at the shortly taperingbase, rounded and bifid at the apex ; upper surface dark green, shining, with acontinuous median groove and without stomata; lower surface with two broadconspicuously white bands of stomata; resin-canals marginal. Leaves on conebearingbranchlets similar to those on barren branchlets.Cones on short stout stalks, resembling in shape and colour those of A . Pindrow;in native specimens both from Sikkim and Kumaon, smaller than those on cultivatedtrees; scales fan-shaped, about % inch wide and | inch long (not including theshort obcuneate claw); bracts extending to near the upper edge of the scale, with anoblong claw, expanding above into a suborbicular denticulate lamina, tipped with a1 Injourn. Roy. Hort. Soc. xiv. 472 (1892).z Silva N. Amer. xii. 98, adnot. ( 1898). The trees, however, at Wellesley, one of which is 59 feet by 5 feet, wereslightly injured during the severe winter of 1903-4. Cf. Sargent, The Pinetum at Wellesley in 1 905, p. 12.3 Hooker, Him. Journ. ii. p. 108.Abies 75 1short triangular cusp and without any emargination. In cultivated specimens, conesvery large, 6 to 8 inches long, 2 to 3 inches in diameter, bluish when growing,brownish when mature, cylindrical, slightly tapering to an obtuse apex ; scales muchbroader than in wild specimens ( \\ inch) ; bracts only extending to about half theheight of the scale, with a broad rectangular claw, only slightly narrower than thebroadly ovate denticulate lamina, which is tipped with a short triangular cusp :seed with wing about an inch long; wing broadly trapezoidal, shining brown, andabout i \ times as long as the seed.The cones of Abies Pindrow are very similar, the main difference being that inthe latter the expanded portion of the bract is situated close to the lower edge of thescale, and is oval, less finely denticulate, and emarginate above with a minute mucroin the emargination.VARIETIESThe above description, which, as regards the leaves and branchlets, applies toordinary cultivated specimens of A bies Webbiana, also fits exactly the form of thatspecies which occurs in Sikkim, and does not differ from the original descriptionwhich was founded on specimens from Nepal. The high-level silver fir, however, which occurs in the western Himalayas appears to be a much shorterleavedtree than that which is common in Sikkim ; and has been supposed by someto be a form of Abies Pindrow. This form, which is apparently the same asspecimens collected on the Chor mountain near Simla by Sir George Watt, is metwith occasionally in cultivation, and may be distinguished as follows :Var. brevifolia? a tree with smooth bark on the stem and branches. Youngbranchlets grey, with only slightly prominent pulvini; pubescence short, erect,brown, confined to the indistinct fine grooves between the pulvini. Leaves muchshorter than in the type, not exceeding i|- inch in length, greyish beneath with twoinconspicuous stomatic bands.This variety differs in appearance from the type, which has longer leaves, verysilvery white beneath; but agrees with it in the arrangement, texture, and shape ofthe leaves. The grey colour and comparative smoothness of the branchlets, and thesmooth bark on the stem and branches, suggest some affinity with A . Pindrow ; butthe long, slender, narrow leaves of the latter species, differently arranged on theglabrous branchlets, are entirely different.I first received cultivated specimens of this variety from Glasnevin, Kilmacurragh,and Batsford Park, where there are young trees, which have not yetproduced cones. The Glasnevin and Kilmacurragh trees were raised from seed,sent from the Himalayas in 1879, but without any record of the precise locality;and they resemble the type in habit. The origin of the Batsford tree is obscure.i Brandis, in Indian Trees, 692 (1906), distinguishes two forms of A. Webbiana, viz. :(a) "A. Webbiana, Lindley. High Level Silver Fir of N.W. Himalaya." This is identical with our var. brtvifolia,and is not the same as A. Webbiana, Lindley.(b\ " A. densa, Griffith. East Himalayan Silver Fir." This, from a comparison of type specimens in the Kew Herbarium, is identical with A. Webbiana, Lindley, which was founded on Pinus Webbiana, Wallich, described from Nepalspecimens.

75 2 The Trees of Great Britain and IrelandA tree of this variety, with smooth bark on the trunk and branches, is growing atPowerscourt, and measured in 1906, 58 feet high by 5 feet 5 inches in girth. Itbore numerous cones, similar to those of the type, but smaller in size and not soblue in colour. Another tree, about 40 feet high, also with very smooth bark,is growing at Holker Hall, Cark-in-Cartmel, Lancashire. (A. H.)DISTRIBUTIONAbies Webbiana occurs in the inner Himalayas from Afghanistan toBhutan at elevations of 10,000 to 14,000 feet, but rare 1 below 11,000 feet.In its western area, i.e. in the north-west Himalaya, it usually commences togrow at 1000 to 2000 feet above the line where Abies Pindrow disappears; andGamble has never seen the two species growing together. It is here usuallystunted and gnarled, with very short leaves and short thick cones, and occurscommonly with Betula utilis and Rhododendron campanulatum. Both it and thebirch are the last trees to be seen before the treeless snowy wastes begin in thewestern Himalayas. In Nepal, according to Don, it occurs on Gosainthan.In its eastern area, Bhotan and Sikkim, it is apparently a larger and finer tree.Griffith 2 mentions it, under the temporary name of Abies densa, as constituting vastforests in Bhotan, remarkable for their sombre appearance, at 12,000 feet, being rareunder 9500 feet.It is slow in growth, the average rate in Sikkim being about 12 rings perinch of radius, and is of much less economic importance than Abies Pindrowis in the north-west. Large quantities of planks, however, are exported fromLachoong to Tibet, and their preparation is an important native industry; butHooker 8 says that the timber of Sikkim conifers is generally soft and inferior to thatof European species.In Sikkim this is the most abundant conifer in the interior; extending from a littleabove 8000 to 13,000 feet or more; at its lower limits scattered or isolated amongother trees at 9000 to 11,000 feet, and forming forests which are sometimes almostpure, or in the Lachen and Lachoong valleys mixed with Tsuga Brunoniana.Higher up on drier slopes it occurs scattered among Larix Griffithi and manyspecies of shrubs and rhododendrons. On the Singalelah range which dividesSikkim from Nepal it begins to appear shortly before reaching Sandukpho, andon the boundary ridge north of that hill assumes a very wind-swept and oftengnarled habit; the tops being often broken and covered with a dense mass offerns, orchids, Ribes, begonias, and climbing plants of many species, and sometimessupporting shrubs and trees, which, favoured by the extremely moist summer climate,and from June to October almost constantly bathed in mist, become epiphytic.One tree which I specially noticed on this ridge at about 11,000 feet, bore on itsdecaying crown no less than four good-sized shrubs of different species, a Pyrus, an1 As in the Chor Hills, south of Simla.2 Notula, iv. 19 (1854), and Itin. Notes, ii. 141 (1848). Griffith subsequently abandoned his name A. densa andadopted that of A. Webbiana. 3 Himalayan Journals, ii. 45, note.Abies 753Aralia, a Rhododendron, and a birch, some of which had stems as thick as my leg.Plate 215 from photographs taken by my late friend, Mr. C. B. Clarke, at this spot,very well represents the trees I saw. The largest trees which I have found of thisspecies were on the track from Lachoong to the Tunkra Pass, leading into theChumbi valley, some of which must have been nearly 200 feet high, with stemsclean up to 40 feet, and Sir Joseph Hooker measured a tree here no less than 35 feetin girth. Higher up, however, it assumes a stunted form and grows mixed withjunipers.HISTORY AND CULTIVATIONAbies Webbiana was discovered early in the nineteenth century by CaptainWebb. Seeds were repeatedly sent to England by Dr. Wallich, which probablycame from Nepal, but none appear to have germinated till 1822, when some plantswere raised in the Fulham Nursery. It is remarkable that most of the trees ofA. Webbiana seen in this country resemble more nearly the Sikkim form, than theshort-leaved Western form. It is probable that none of the original trees now exist,as they were planted in the vicinity of London, where the tree does not thrive, as itis very liable to be cut by spring frosts.Though this tree is one of the most beautiful of its genus in the few parts ofEngland where it really succeeds ; and will resist severe winter frosts without injurywhen on well-drained soil, yet its tendency to start into growth before the danger ofspring frosts has passed, has caused its death in very many places. If seeds could beprocured from the more alpine regions of Kashmir or the trans-Indus mountains, theymight endure our climate better, but most of the trees now growing in Englandwere probably raised from seed collected by Sir J. D. Hooker in Sikkim.It ripens seed, however, in some parts of England and Scotland, and I haveraised seedlings in 1901 from cones grown near Exeter, of which a few havesurvived though now not more than a foot high. A shady, elevated, and yetsheltered situation, is best for this species, and as regards soil a deep sandy loam.REMARKABLE TREESThe largest specimen of A. Webbiana recorded at the Conifer Conference in1891 was at Howick Hall, Northumberland, the seat of Earl Grey, and was thensaid to be 51 feet by 8 feet. I am informed by Mr. Lambert that it has lost its leaderseveral times since this date, and now measures about 50 feet by 8^ feet.The largest we have measured is a double-stemmed tree at Beauport, Sussex,64 feet by 84- feet in 1904; but Mr. A. B. Jackson found a tree at Tregothnan inCornwall which was 74 feet by 8^ feet, and another tree at the same place 66 feetby 9 feet. Both of these bore cones in 1908.At Menabilly, Cornwall, there is a healthy tree of no great size, which bore largecones in 1907, and these remained in perfect condition on the tree in April 1908,when I visited the place; and at Pencarrow there is one 64 feet by 6j feetIVG

754 The Trees of Great Britain and IrelandAt Fulmodestone, Norfolk, in a wood where the Earl of Leicester has plantedon a deep moist soil a large number of conifers, close enough to shelter each other,a tree was measured by Henry in 1905 as 69 feet by 6| feet. It has since beenmuch damaged by snow, but has produced cones from which Capt. R. Coke hasraised a few seedlings, and now measures only about 60 feet high.At Enville Hall, Stourbridge, there is a tree, which now looks as if it weresuffering from drought. It measured, in 1904, 68 feet by 5 feet 4 inches.A large tree, said to have been about 75 feet high, died and was cut down atPenrhyn in North Wales in 1902. The stump, which I saw, was about 7^ feetround. At Hafodunos, in North Wales, Colonel Sandbach states that this species isalways nipped by the frost and forms new leaders when the old ones are killed, thegrowth being quite checked.At Castle Kennedy there is a short avenue of trees of this species, averagingabout 40 feet high by 6 feet in girth; but the tops had been cut off, as they hadbecome bare and unsightly from exposure to wind. Here A . Webbiana begins toproduce cones at an early age; and there is a seedling 20 feet high, with manysmaller ones near it.At Poltalloch, Argyllshire, I measured a fine healthy tree 61 feet by 5 feet3 inches which in 1906 bore no cones.At Keir, Perthshire, there are two trees, one with a broken top, the other 57feet by 4 feet 10 inches, and more narrowly pyramidal than is usually the case.Seedlings were raised from the seed of this tree about ten years ago; but as a ruleit bears very small cones (only 3 inches in length) with unfertile seed. A tree atDunphail, Morayshire, has also produced small cones, which slightly resemble thoseof Abies Mariesii. 1 This tree was probably planted in 1856, and is now only 33 feetin height; but has a double leader. 2In Ireland, A . Webbiana thrives well, and there is a good number of fair-sizedtrees. It is said, however, to be slightly touched by frost at Fota, in the south ofIreland, where the temperature fell to 14" Fahr. during four nights in the winterof 1901-1902. A tree at Fota was in 1903, 47 feet high by 3 feet 7 inches in girth.At Churchhill, Armagh, the seat of Mr. Harry Verner, there was growing in1904 a tree laden with cones, even on the lowermost branches ; it measured 53 feetby 4 feet 10 inches.At Courtown, Co. Wexford, a tree was recorded at the Conifer Conference of1891, as being 52 feet by 6 feet 3 inches. (H. J. E.)1 The Dunphail tree has been described and figured by Masters as Abies Mariesii, in Bot. Mag. t. 8098 ; but there is nodoubt that this is erroneous. Specimens which I have seen show ordinary foliage and branches of A. Webbiana; and some ofthe cones are as large as those usually produced by this species. (A. H.)2 Thrives at Durris only in partial shade, when exposed it suffers much from late spring frosts, both top and branch shootsbecome clubbed and unsightly. Cones at a comparatively early age about 25 years. Is of no economic value.(J. D. CROZIER.)Abies 755ABIES PINDROW, PINDROW FIRAbies Pindrow, Spach, Hist. Vlg. xi. 423 (1842); Masters, Card. Chron. xxv. 6 91, f. 154 (1886);Kent, Veitch's Man. Conifera, 5 33 (1900); Gamble, Indian Timbers, 7 19 (1902); Brandis,Indian Trees, 6 92, 720 (1906).Abies Webbiana, Lindley, var. Pindrow, Brandis, Forest Flora Brit. India, 5 28(1874); Hooker,Flora Brit. India, \. 6 55 (1888).Pinus Pindrow, Royle, Illust. Bot. Himalaya, 3 54, t. 86 (1839).Picea Pindrow, Loudon, Arb. et Frut. Brit. iv. 2346 (1838).A tree attaining in the Himalayas over 200 feet in height, with a girth of 25feet. Narrowly pyramidal in habit, with the branches small and short. Barksmooth and silvery grey when young; greyish brown, deeply and longitudinallyfissured on old trunks.Buds large, globose, covered with white resin. Young shoots quite smooth,grey, glabrous, the bark assuring slightly in the second year. Leaves on lateralbranches mostly pectinate below, pointing forwards and outwards in the horizontalplane, some of the median leaves, however, being directed downwards and forwards;above, covering the shoot, those in the middle line much shorter and directedforwards and slightly upwards. Leaves, soft in texture, up to 2^ inches long, verynarrow (^ inch wide), linear, flattened, shortly tapering at the base and narrowinggradually in the anterior third to the acute apex, which is bifid with sharp unequalcartilaginous points; upper surface dark green, shining, with a continuous mediangroove and without stomata; lower surface paler with two greyish bands of stomata,each of about eight lines ; resin-canals marginal. Leaves on cone-bearing shoots allupturned and more or less directed forwards, covering the shoot in the middle lineabove, shorter than on barren branches and only slightly bifid at the apex.Cones on short stout stalks, bluish when growing, brown when mature, cylindrical,about 6 inches long by 3 inches in diameter. Scales ; lamina about \\ inch wide byf inch long, fan-shaped, variable in form, with two slight wings in cultivated specimens, not winged and with the lateral edges straight or curved in wild specimens,base auricled. Bracts with the expanded portion situated on the scale just abovethe claw, oval, denticulate, emarginate above with a minute mucro. Seed with wingabout i inch long, the wing narrowly trapezoidal and about \\ times as long asthe body of the seed.IDENTIFICATIONAbies Pindrow is remarkably different in most characters from Abies Webbiana,with which it has been united by many authors. The trees are very distinct inhabit, A . Pindrow forming, both in the Himalayas and in cultivation in England, anarrow pyramid with short branches; while A . Webbiana is a broader tree withwide-spreading branches. The bark of the former is smooth, that of the latterscaly. The former has smooth, glabrous, grey shoots; the latter has shoots with

756 The Trees of Great Britain and Irelandprominent pulvini, separated by pubescent furrows. The cones are similar in sizeand colour; but differ in the shape and position of the bracts. The arrangementand character of the foliage are entirely different.Var. intermedia.Pinus (Abies) sp. nova (?), M'Nab, in Proc. Roy. Irish Acad. ii. 692, f. 19 (1876).A tree at Eastnor Castle, planted thirty-seven years ago, and about 60 feet inheight with a girth of 3 feet 4 inches, is apparently identical with the form describedby M'Nab, who mentions two specimens, one collected in the western Himalayasby Hooker and Thomson, and another from a tree, which formerly grew at CastleKennedy. Mr. Mullins, the gardener at Eastnor Castle, informs me that the tree isnarrowly pyramidal in habit, with dark green foliage, and smooth bark on the stemand branches.Specimens, which I have received, show the following characters: Branchlets,buds, bark, and habit, as in A . Pindrow. Leaves more pectinate than in thatspecies, and arranged on the branchlets as in A . Webbiana ; about 2^ inch inmaximum length, dark shining green above; gradually tapering in the upper third,as in A. Pindrow; thicker than in this species ; lower surface convex ; resin-canalsmedian, in which respect this variety differs from both A . Pindrow and A . Webbiana.Cones about 4 inches long and 2 to 3 inches in diameter, resembling those of A .Webbiana in the position and shape of the bracts.This variety is intermediate in many respects between A . Pindrow and A .Webbiana, and is possibly a hybrid. (A. H.)DISTRIBUTIONAbies Pindrow is more restricted in distribution than A . Webbiana and occursat a lower elevation. It is met with in the outer Himalayas from Chitral to Nepal,at elevations of 7000 to 9000 feet, occasionally ascending to 10,000 feet; andcommonly grows in ravines with a northerly or westerly aspect. It is oftenassociated with Picea Morinda, Quercus dilatata, the deodar, and Pinus excelsa ; butmore often is accompanied by broad-leaved trees, such as the walnut, maples, birdcherry, and Indian horse-chestnut. Madden says that it forms dense forests onall the great spurs of Kumaon and occurs in Kashmir. According to Gamble, ithas the same narrowly pyramidal habit with short branches which it assumes incultivation in England. It grows very tall, but does not attain so great a height asthe deodar. The largest trees correctly noted were measured in the Mundali forestin Jaunsar, and had heights varying from 188 to 206 feet with girths of 19 to 25feet. The rate of growth averages 13 rings per inch of radius. It bears intenseshade and its natural reproduction is excellent. The timber is employed indiscriminately with that of Picea Morinda, though not quite so good. It is used forplanking, tea-boxes, packing-cases, and makes excellent shingles, and would besuitable for railway sleepers if creosoted.Abies 757REMARKABLE TREESThe finest trees in Britain are probably two, which I saw in 1906, growing inthe grounds of Mr. Victor Marshall at Monk Coniston, the largest of which is now69 feet high and 4 feet 9 inches in girth. The climate here is mild and damp,but the soil dry and slaty. Mr. S. A. Marshall wrote to Kew that one of these treesconed for the first time in 1902, and in 1904 produced many cones. 1There is a well-shaped tree at The Coppice, Henley, the seat of Sir WalterPhillimore, Bart. It came from Dropmore as a young plant in 1858, and grew veryslowly at first, sustaining some damage from the frost of May 1867. It is nowhealthy and thriving, and measured, in 1907, 62 feet high by 3 feet 8 inches in girth.At Bury Hill, Dorking, the seat of R. Barclay, Esq., there is a well-grown specimen,which he informed me had been raised from a tree at Denbies in the sameneighbourhood, and which in 1908 measured 58 feet by 3 feet 2 inches. Itgrows, like so many of the best conifers in this country, on greensand. AtDropmore, a tree 48 feet by 3 feet 10 inches in 1905, coned for the first timein 1907. At High Canons, Herts, Mr. H. Clinton Baker measured a tree, whichwas bearing cones in February 1908, as 48 feet high by 2 feet 8 inches in girth ;and in a wood, on his own property at Bayfordbury, there is a thriving young tree,32 feet high and i foot 5 inches in girth. At Cuffnells, near Lyndhurst, a good-sizedbut badly-grown tree was bearing cones in 1907. At Leighton Hall, near Welshpool,the seat of Mrs. Naylor, there are three, at a considerable elevation, the bestof which in 1908 was about 50 feet high. In the pinetum at Lyndon Hall, Oakham,Rutland, Henry measured in 1908 a fine specimen, 58 feet by 3 feet 3 inches, whichwas planted, according to Mr. E. L. P. Conant, by his father in 1864.In Scotland I measured a tree at Conon House in Ross-shire, in 1907, whichwas about 50 feet by only 3 feet 4 inches. At Smeaton-Hepburn, East Lothian, atree planted in 1844, according to Sir Archibald Buchan-Hepburn, Bart., was, in1908, 56 feet high by 4 feet i inch in girth. Seedlings have been raised from it.At Durris, Mr. Crozier says that it suffers from the same causes as A . Webbiana.In Wigtownshire, this species is tender, and a tree at Galloway House, 48 feet by3 feet 10 inches, as measured by Henry in 1908, has been much injured by frost.Sir Herbert Maxwell reports a good specimen at Stonefield, Loch Fyne ; and another,in the Quarry garden at Gordon Castle, which measures 69 feet by 4 feet 9 inches.In Ireland, the tallest tree is at Charleville, Co. Wicklow, the seat of ViscountMonck. In 1904 it was 60 feet by 5 feet i inch. It produced cones in 1903, andsince then has been in an unhealthy state. This tree has sent forth from the stemepicormic branches. At Kilmacurragh, Co. Wicklow, the seat of Mr. ThomasActon, there is a healthy tree, which, in 1904, was 51 feet high by 5 feet 5 inches.At Powerscourt, there are two good trees, one 55 feet by 5 feet in 1904, which borecones in 1902, 1903, and 1904; the other is 42 feet by 6 feet 3 inches in girth. AtBrockley Park, Queen's County, a tree, planted thirty-five years ago, measured in1907, 51 feet by 5 feet 6 inches. (H. J. E.)1 A tree at Kenfield Hall, near Canterbury, produced cones in 1886. Cf. Card. Chron. xxvi. 85 (1886)."

75 8 The Trees of Great Britain and IrelandABIES SIBIRICA, SIBERIAN FIRAbies sibirica, Ledebour, Fl. Alt. iv. 202 (1833); Masters, Journ. Linn. Soc. (Sot.) xviii. 519(1881); Kent, Veitch's Man. Conifera, 5 39 (1900).Abies Pichta, Forbes, Pin. Woburn. 1 13, t. 39 (1840).Abies Semenovii, Fedtschenko, Bot. Centralblatt, Ixiii. 210 (1898), and Bull. Herb. Boissier, vii. 191(1899).Pinus sibirica, Turczaninow, Bull. Soc. Nat. Mosc. xl. 101 (1838).Pinus Pichta, Endlicher, Syn. Conif. 1 08 (1847).Picea Pichta, Loudon, Arb. et Frut. Brit. iv. 2338 (1838).A tree attaining about 100 feet in height. Bark smooth, greyish, and coveredwith resin-blisters, even in old trees. Buds small, globose, brownish, smooth, andcovered with resin. Young shoots ashy-grey, with a scattered minute erect pubescence, quite smooth, the pulvini not being at all prominent; in the second year,the bark fissures slightly, and the pubescence is retained.Leaves on lateral branches resembling in arrangement those of A . Veitchii,but more irregular; the lower ones pectinate, and directed outwards and forwards,a few, however, in the middle line with their apices directed forwards anddownwards; on the upper side the leaves cover the branchlet and are directedforwards and upwards in the middle line, being about three-fourths the length of thelower leaves. Leaves linear, flattened, slender, up to i^ inch long, -£G inch wide,uniform in width except at the slightly narrowed base; apex rounded, slightly bifidor entire; upper surface light green, shining, with a continuous median groove andrarely two to three short lines of stomata near the apex in the middle line; lowersurface greyish in colour, with two narrow bands of stomata, each of four to fivelines ; resin-canals median. Leaves on cone-bearing branches all upturned, curved,thick, short (f inch long), acute at the apex.Cones sessile, cylindrical, obtuse at the apex, 2 to 3 inches long, i£ inch indiameter, bluish when growing, brown when mature, with the bracts concealed.Scales; lamina fan-shaped, thin, f to f inch wide, J inch long; upper and lateralmargins denticulate; base with a sinus on each side of the obcuneate claw. Bract,at the base of the scale, rectangular or reniform, coarsely denticulate, j^ inchbroad, with a short triangular mucro. Seed with wing about f inch long; wingbroad, purplish, about twice as long as the body of the seed.The form x described by Fedtschenko as a new species (A. Semenovii) occurs inTurkestan. Specimens show longer leaves, more pubescent branchlets, and slightlydifferent cone-scales and bracts. Korshinsky, however, in a note in the Kewherbarium, states that the Turkestan tree is identical with A . sibirica fromthe Ural and Altai; and the differences noted would probably disappear if therewere more material to examine.A weeping variety of this species was seen by Conwentz 2 in 1881 in Regeland Kesselring's nursery at St. Petersburg.This species, with long slender leaves covering the branchlet above, is best1 Cf. Guinier and Maire's remarks on this form in Bull. Soc. Bot. France, Iv. 184 (1908).2 Seltene fValdtdume in Westpreussen, 161 (1895).Abies 759distinguished by its ashy-grey smooth shoots which are minutely pubescent, and itssmall globose resinous buds. It can only be confused with A . sachalinensis, whichhas shoots with prominent pulvini, and leaves with broader and whiter bands ofstomata below. A . lasiocarpa, which has, like this species, median resin-canals, isdistinguished by the peculiar arrangement of the leaves, which have conspicuouslines of stomata on their upper surface.DISTRIBUTION AND CULTIVATIONThis species is the most widely distributed of all the silver firs, occupying largeareas of both the plains and mountains of north-eastern Russia and Siberia. InEuropean Russia it forms forests in company with spruce and larch, or rarely withbirch and aspen; and occurs through the governments of Archangel, Vologda,Kostroma, Perm, Ufa, Kazan, and Orenburg. On the mountains it does not go ashigh as the timber line, and does not extend so far south in European Russia as thespruce. It is common in the Ural range, and attains perhaps its maximum development in the Altai, 1 where it forms vast forests between 2000 and 4500 feet elevation.In Turkestan it is found in the Thianshan mountains, and is reported by Korshinskyto form forests at low altitudes in the province of Ferghana, where it grows inmixture with Picea Schrenkiana. Its distribution in Siberia is not clearly known,but it appears to be widely spread from west to east, its northern limit on theYenisei being 66 lat. and on the Lena 60 lat. It occurs on the high lands ofDahuria, and, according to Komarov, reaches its most easterly point in the Yablonoimountains, being replaced in Kamtschatka, by A . gracilis, Komarov ; and on theborders of the sea of Okhotsk and in Manchuria by A . nephrolepis, Maximowicz.According to Loudon this species was introduced from the Altai into Englandin 1820. It is very rare in cultivation, and does not grow for any length of time inthe south of England, where the climate is unsuitable to it. Even at Durris,Mr. Crozier describes it as "a slow-growing, many-headed, and evidently shortlived tree." There is an unhealthy specimen, about 15 feet high, at Ochtertyre inPerthshire. At Pampisford, Cambridgeshire, there is a small tree, 30 feet highby 13 inches in girth in 1907, narrowly pyramidal in habit, with the lower brancheslayering and producing five independent stems about six feet in height. This treehas been badly damaged by the snowstorm of April 1908. Another at Bictonmeasured 28 feet by i foot 8 inches in 1908.In August 1908 I saw a fine specimen in the University Botanic Garden,Upsala, Sweden, which was 70 feet high and i foot in diameter, forming a verynarrow pyramid, closely resembling the habit of A . Pindrow. Hansen 2 says thatspecimens, forty years old, have attained a height of about 40 feet in Denmark; andthat there are beautiful examples in the Botanic Gardens at Helsingfors, Finland(lat. 60 ), where many seedlings have sprung up around the old trees. (A. H.)1 I saw this tree in the forests on the north slopes of the Altai, where the climate was damp, but it did not strike me as afine or large tree, and was not seen in the drier valleys towards Mongolia. (H. J. E.)2 Journ. Roy. Hort. Soc. xiv. 477 (1892).

760 The Trees of Great Britain and IrelandABIES SACHALINENSIS, SAGHALIEN FIRAbies sachalinensis, Masters, Card. Chron. xii. 588, f. 97 (1879), ar>d Journ. Linn. Soc. (£ot.) xviii.517 (1881); Mayr, Abiet. jap. Reiches, 42, t. 3, f. 6 (1890); Sargent, forest Flora of Japan,83 (1894); Kent, Veitch's Man. Coniferce, 5 37 (1900).Abies Veitchii, Lindley, var. saclialinensis, F. Schmidt, Mdm. Acad. St. Petersbourg, ser. 7, xii. 175,t. 4, ff- 13-17(1868).A tree attaining in Yezo 130 feet in height. Bark smooth, grey in colour.Buds small, ovoid - globose, rounded at the apex, covered with white resin.Young shoots grey, with prominent pulvini and grooves; pubescence, short, dense,and confined to the grooves. In the second year the pulvini, grooves, andpubescence are well-marked.Leaves on lateral shoots arranged similarly to those of A . Nordmanniana;those below longest, pectinate in two sets in the horizontal plane, directedoutwards and forwards ; those above covering the branchlet in imbricatedranks, the median leaves shortest, directed forwards and appressed to theshoot. Leaves linear, flattened, slender, very thin, about i|- inch long, ¥\j inchwide, uniform in width except at the shortly tapering base; apex rounded andshortly bifid; upper surface grass-green, shining, with a continuous median grooveand without stomata; lower surface with two narrow bands of stomata, each ofseven to eight lines ; resin - canals median. Leaves on cone - bearing shoots upturned, acute or rounded at the minutely bifid apex.Cones, 3^r inches long, i£ inch in diameter, cylindrical, tapering to an obtuse orslightly acute apex, conspicuously marked externally by the reflexed greenish bracts,which leave little of the surface of the scales visible. Scales crescentic, small ;lamina, |- to f- inch wide, nearly \ inch long, deeply auricled by two basal sinuses,the denticulate wings ending in a sharp point on each side of the sinus; uppermargin entire; outer surface densely tomentose. Bract with a broad cuneate claw,expanding above into an almost orbicular lamina, emarginate and mucronate on itsupper margin. Seed with wing f inch long; wing broader than long and shorterthan the body of the seed.VARIETIESVar. nemorensis, Mayr, loc. cit. This variety is met with in north-easternYezo and the Kurile Isles, and is distinguished by its smaller cones, about i\ inchesin length, with minute concealed bracts. In the cones this variety resemblesA. Veitchii.Sargent mentions a curious variety found by Miyabe in central Yezo, in whichthe bark, wood, and bracts of the cone are red in colour.IDENTIFICATIONAbies sachalinensis agrees in many technical characters with A . Veitchii, but owingto its longer and more slender leaves looks different from that species, and wouldAbies 761not in practice be confused with it. The prominent pulvini of the young branchlets,which are only pubescent in the grooves, will distinguish it at once from Abiessibirica, which it resembles in general appearance. The latter species has quitesmooth branchlets, provided with a scattered minute pubescence. (A. H.)DISTRIBUTION AND CULTIVATIONThis species was discovered in the Island of Saghalien by Schmidt in 1866,and was subsequently found, in 1878, in Yezo by Maries, who sent home seeds inthe following year. 1 The tree is known in Japan as Todo-matsu.It is a native of the Kurile Islands, Saghalien, and the northern island ofJapan. 2 In Saghalien it either forms pure woods or is mixed with one or bothof the spruces (Picea ajanensis and Picea Glehnii) which occur in that island.This is the common, and perhaps the only, silver fir of Hokkaido, where itextends from nearly sea-level up to 4000 or 5000 feet altitude, and all over the islandin suitable places; in the south usually as a scattered tree in mixed forests ofdeciduous trees; in the north and some parts of the west central districts in densepure forests, or with a mixture of birch and poplar. The finest areas of this speciesare in the Imperial domains at Tarunai, Uryu, Kushiro, and in the State forestsat Shari, and Kunajiri. I endeavoured to visit some of these under theguidance of Mr. Shirasawa, but owing to the torrential rains which flooded thecountry in the middle of July and broke the railway in many places, I was unable todo so. The country where these forests occur is much like parts of eastern Siberia,having a hot, moist summer, a warm autumn, and a very heavy snowfall which liesfor four to five months; the climatic conditions, therefore, are such that the tree isnot likely to be a success in Great Britain, and, so far as I could see, it has nospecial beauty to recommend it. The largest that I saw were about 100 feet by9 feet, but it grows taller in some places.The timber is of fair quality, and is used in house- and ship-building, also forfurniture and paper-making ; and is worth at Tokyo about lod. per cubic foot.The Saghalien fir is rare in cultivation, the largest specimen we have seenbeing one at Fota, in the south of Ireland, which was about 25 feet high in1907. It looks healthy, but begins to grow early in the season, and is saidto be frequently hurt by spring frost. We have measured no specimens inScotland, but one at Murthly Castle, about 16 feet high, is reported by Mr. Bean 8as not looking healthy. According to Kent, the tree is, like most of the coniferscoming from similar climates, unable to thrive in England. In New England,however, it grows much better, and I saw healthy young trees at Mr. Hunnewell'splace at Wellesley, Massachusetts.(H. J. E.)1 Hortus Veitchii, 337 (1906).2 It was reported by Matsumura ( Tokyo Bot. Mag. xv. (1901), p. 141), to occur in Formosa, on Mt. Morrison; butthis was a mistaken identification, as the silver fir in this locality is A. Marian, according to Hayata, in Tokyo Bot. Mag.xix. (1905), p. 45. 3 Gard- Chron. xii. 117 (1907).IVII

762 The Trees of Great Britain and IrelandABIES FIRMA, JAPANESE FIRAbies firma, Siebold et Zuccarini, ft. Jap. ii. 15, t. 107 (1844); Masters, Gard. Chron. xii. 198, 199(1879), and Journ. Linn. Soc. (Bot.) xviii. 514 (1881); Mayr, Abiet. Jap. Reiches, 3 1, t. i, f. i(1890); Shirasawa, Icon. Ess. Forest. Japan, text 17, t. 6, ff. 1-21 (1900) ; Kent, Veitch'sMan. Conifercs, 5 06 (1900).Abies bifida, Siebold et Zuccarini, Fl. Jap. ii. 18, t. 109 (1844).Abies Momi? Siebold, Verhand. Batav. Gen. xii. 101 (1830) (nomen nuditni).Pinus firma and Pinus bifida, Antoine, Conif. 7 0, 79 (1846).Picea firma, Gordon, Pinet. 1 47 (1858).A tree, attaining 150 feet in height and 16 feet in girth. Bark of branchesand trunk early becoming scaly, in old trees fissuring into small plates. Buds small,ovoid, obtuse at the apex, brown, glabrous, slightly resinous. Young shootsbrownish grey, with the pulvini slightly raised and separated by grooves ; pubescenceshort, erect, scattered, confined to the grooves. Older shoots retaining the pubescenceand fissuring between the pulvini, which are not very prominent.Leaves on lateral branches pectinately arranged; those below extendinglaterally outwards in the horizontal plane; those on the upper side graduallyshortening to nearly one-third of the length of the lower leaves, and directed in twosets laterally outwards and slightly upwards, forming a shallow V-shaped arrangement. 2 Leaves up to i^ inch long, linear, flattened, very coriaceous, shortly taperingto the base, broadest about the middle (^ inch or more), gradually narrowing to theacute apex, which ends in two sharp cartilaginous points, unequal in size; uppersurface dark green, shining, with a continuous median groove and without stomata ;lower surface with broad greyish bands of stomata, each of about ten to twelve lines ;resin-canals close to the epidermis of the lower surface. Leaves on cone-bearingbranches upturned, rounded and entire or only minutely bifid at the apex.Staminate flowers, \ inch long, ovoid-conic, surrounded at the base by two tothree series of broadly ovate scales.Cones on stout short stalks, cylindrical, tapering shortly at the base, and obtuseor flattened at the slightly narrowed apex, yellowish-green before ripening, brownwhen mature, 4 to 5 inches long by i|- to if inches in diameter, with the tips of thebracts exserted between the scales but not reflexed. Scales: lamina i^ incheswide by f inch long, broadly trapezoidal; upper margin thin, minutely denticulate;lateral margins convex, denticulate ; base broad with a sinus on each side of theobcuneate claw. Bract extending either nearly up to the edge of the scale orbeyond it, always visible externally between the scales, oblong in the lower half,expanding above into an oval lamina, which ends in a triangular cusp. Seed-wingbroadly trapezoidal, about twice the length of the body of the seed. Seed withwing nearly f inch long. Cotyledons four.1 This name, which has been adopted by Sargent, Sitoa N. Amer. xii. IOI, adnol., was published without any description, and cannot be maintained. Cf. Masters, Gard. Chron. loc. cit.2 On vigorous shoots, the leaves are directed more upwards so that the V-shaped depression is very acute.AbiesThe broad coriaceous leaves, ending in two cartilaginous points of unequalsize, and pectinately arranged, are characteristic of this species, which when onceseen, can scarcely be confused with any other.Abies holopkylla, Maximowicz, 1 which has been identified by Dr. Masters withA. firma, is considered by Komarov, 2 the latest observer, to be a distinct species.It differs in the leaves not being bifid, and also in the bracts of the cone, which areshort, scarcely extending more than one-third the length of the scale. This species,according to Komarov, attains 150 feet high, and grows in mountain woods atelevations not exceeding 1800 feet above sea-level, in the Manchurian provinces ofUssuri, Kirin, and Mukden, and also in northern Korea. It was introduced intocultivation in Russia by Komarov, who sent seeds in 1898.Other specimens of Abies from the Chinese provinces of Hupeh, Shensi, andYunnan s have also been considered by Dr. Masters to be A . firma; but thisidentification is doubtful. (A. H.)DISTRIBUTIONThis, the best known fir in Japan, is widely distributed in the south, and,according to Mayr, does not extend north of lat. 40 , and attains perfection in thewarm sub-tropical provinces of Kii, Shikoku, and Kiusiu. It is very commonlyplanted in temple grounds and parks, but few of these specimens looked as if theisolated situation agreed with them ; and wherever I saw the tree growing naturally,it was scattered among deciduous trees and other conifers in more or less shadyplaces in the forest. It grows to a great size in the sheltered valleys and moist, deepsoils of the central and southern provinces. I measured one at Myanohara, on theNakasendo road near Wada, which was 135 feet by 16 feet, but this tree was dyingat the top, and may have been planted or have been a natural seedling in a templegrove. Another in the forest near the entrance to Koyasan was about 120 feet highby 15 feet 9 inches in girth, but the average size of the mature trees that I saw wasnot over 100 feet by 9 feet. A third, growing close to a temple at Narai (Plate 216),measured 125 feet by nj feet. As the timber is of little value except for packingcases,tea-boxes, and pulp-wood, the tree is not much planted at the present timeexcept for ornament. It reproduces itself freely from seed whenever the conditionsare suitable, and its large greenish-yellow cones are fully formed in August.According to Rein 4 its natural habitat is from 1000 to 1500 metres, but thoughthis may be the case in the southern island, I should say it was too high for thecentral provinces, as in Kisogawa I saw it much lower, and I do not think it therereaches 4000 feet.This species was introduced into Europe in 1861 by J. Gould Veitch, 5 but hasnever become common in cultivation, though it seems to be hardy even in someparts of Scotland. It undoubtedly requires a warm, moist climate to bring it to1 MM. Biol. vi. 22 (1866). 2 Flora Manshurice, 204(1901).3 Journ. Linn. Soc. (Bot.) xxvi. 557 (1902); axAJourn. Bot. 1 903, p. 270.4 Industries of Japan, 2 35 (1889). Mayr says that it ascends to 700 feet in the north and to nearly 7000 feet in the south.6 Hortus Veitchii, 335 (1906), where it is stated that it was also sent in 1878 by Maries.

764 The Trees of Great Britain and Irelandperfection, and seems to be in most places a slow grower. The seedlings which Ihave raised from Japanese seed will not grow on my soil, from which I infer thatlime is distasteful to them.REMARKABLE TREESThe largest tree that we know of is at Carclew in Cornwall. This was reportedin 1891 to be 45 feet by 2 feet 8 inches, and when I measured it in 1902 hadincreased to about 60 feet by 4 feet. Another at Pencarrow, in the same county, wasabout 59 feet by 6 feet 5 inches in 1908. At High Canons in Herts, Mr. ClintonBaker showed me a specimen which bore cones in 1907 and measured 47 feet by3i feet.There is a good-sized tree at Grayswood, with longer and less sharp-pointedleaves than usual, and another at Tortworth which in 1905 was 30 feet by 3 feet9 inches. A tree planted at Bagshot Park by the late Emperor of Germany onJuly 10, 1880, was, when I saw it in 1907, 36 feet by 3 feet n inches.In Scotland 1 the best that we have seen is at Castle Kennedy, which, in 1904,Henry found to be 44 feet by 5 feet 5 inches. Another in a wood at Munches,Dalbeattie, was 30 feet by 2^ feet. Trees were reported to be growing in i 8gi"at Balmoral, and at Haddo House in Aberdeenshire, but we have not identifiedthem.In Ireland there were thriving trees at Fota 25 feet high, and bearing cones in1907 ; at Hamwood, Co. Meath, 36 feet by 2 feet 10 inches in 1904 ; and at Powerscourt,which in 1906 was bearing cones, and measured 39 feet by 3 feet n inches.(H. J. E.)ABIES HOMOLEPISAbies homolepis, Siebold et Zuccarini, Fl. Jap. ii. 17, t. 108 (1844); Masters, Card. Chron. 1 879,p. 823, an&Journ. Linn. Soc. (Bot.) xviii. 518 (1881).Abies Tschonoskiana, Regel, in Index Sem. Hort. Petrop. 3 2 (1865).Pinus Harryana, M'Nab, Proc. R. Irish Acad. ii. 689, PI. 47, f. 16 (1876).This species, imperfectly described by Siebold and Zuccarini, is considered byMayr to be a form of A . brachyphylla. I t is different in the pulvini of the branchlets,in the shape and arrangement of the leaves, and in the position of the resin-canals inthe latter. Specimens in cultivation, described below, agree with the type of Sieboldand Zuccarini's species in the Leyden Museum. The cones are unknown; and it ispossible that it may be a juvenile form or variety of A . brachyphylla; but in thepresent state of our knowledge, it is best kept distinct.1 Has been tried at Durris repeatedly, but does not live beyond a year or two. Quite unfitted for our climate. (J. D. CROZIER.)3 Card. Chron. x. 458 (1891).Abies 765As seen in cultivation at Kew, it is a small tree, resembling in bark and habitA. brachyphylla* The foliage, however, is rather like that of A.firma, and the treeis occasionally cultivated under that name.1Buds ovoid, obtuse at the apex, whitened with resin, much larger than those ofA. firma. Young shoots grey, glabrous, with prominent pulvini and grooves, whichbecome less marked in the second year.Leaves on lateral branches arranged as in A . firma, those of the upper rankabout half the length of those of the lower rank, linear, flattened, rigid and slightlycoriaceous, up to about i^ inch long, ^ inch wide (much narrower than in A.firma),tapering gradually to the base, and narrowing near the rounded or acute apex, whichis bifid with two short unequal cartilaginous points; upper surface dark green,shining, with a continuous median groove and without stomata; lower surface withtwo raised narrow white bands of stomata, each of about eight lines ; resin-canalsmarginal.Though this species has been in cultivation 2 since 1876 or earlier, we have seenno large specimens; and are ignorant as to whether it changes in character as itgrows older or is short-lived. Its distribution in Japan is not known.(A. H.)ABIES BRACHYPHYLLA, NIKKO FIRAbies brachyphylla? Maximowicz, Mil. Biol. vi. 23 (1866); Masters, Card. Chron. xii. 556 (1879),an&Journ. Linn. Soc. (Bot.) xviii. 515 (1881); Hooker, Bot. Mag. t. 7114 (1890).Abies homolepis? Mayr, Abiet. Jap. Reiches, 3 5, t. 2 f. 3 (1890); Shirasawa, Icon. Ess. Forest. Japon,text 14, t. 3, ff. 1-12 (1900); and Kent, Veitch's Man. Conifera, 5 13 (1900) (not Siebold etZuccarini, Fl.Jap. ii. 17, t. 108 (1844)).Pinus brachyphylla^ Parlatore, in D.C. Prod. xvi. 2, p. 424 (1868).Picea brachyphylla^ Gordon, Pinetum, 2 01 (1875).A tree attaining in Japan over 100 feet in height and 16 feet in girth. Barkfissuring and scaly on young branches and on the stems of young trees, becominglike that of a spruce on old trees. Buds ellipsoid or broadly conical, obtuse at theapex, smooth, brownish, resinous. Young shoots greyish, glabrous, with prominentpulvini, separated by deep grooves, the pulvini and grooves becoming more markedin older shoots.Leaves on lateral branches pectinate ; those below extending laterally outwardsin the horizontal plane, with a few in the middle line directed forwards and downwards ; those on the upper side of the branchlet directed upwards and outwards, in1 It is readily distinguishable from A.firma by its glabrous shoots, larger buds, and much less coriaceous and narrowerleaves.2 M'Nab, foe. cit. mentions plants of this species, which were growing under the name of A. Veitchii in several nurseries.3 The following description applies to the tree, described by Maximowicz, which is, in my opinion and that of Dr.Masters, very different from Abies homolepis, S. et Z., which is treated by us as a distinct species.

The Trees of Great Britain and Irelandtwo lateral sets, separated by an acute V-shaped depression. Lower ranks with thelongest leaves (about f inch), those in the other ranks gradually diminishing in sizeas they approach nearer the middle line above. Leaves linear, flattened, uniform inwidth except at the gradually tapering base, about -fa inch, rounded and slightlybifid at the apex; upper surface dark green, shining, with a continuous mediangroove and without stomata; lower surface with two broad conspicuously whitebands of stomata, each of ten to twelve lines ; resin-canals median.Leaves on cone-bearing branches shorter than on barren branches, those onthe upper side of the shoot crowded and directed upwards, so that the V-shapeddepression between the lateral sets is scarcely visible.Cones on short stalks, cylindrical, slightly narrowed at the base and apex,4 inches long, i| inch in diameter, purple when growing, brown when mature.Scales very thin and flat; lamina fan-shaped, i| inch long by f inch wide, uppermargin entire, lateral margins with denticulate short wings ; claw short and obcuneate.Bract short, not extending half-way up the scale; with a sub-orbicular finelydenticulate lamina, tipped by a minute mucro, and a short obcuneate claw. Seedwith wing about f inch long, the wing about i| times as long as the body of theseed. In cultivated specimens, the scales of the cone and the seeds are smallerthan in wild specimens. The very thin flat scales with the short minutelydenticulate bract distinguish well this species.IDENTIFICATIONThis species has short leaves very white underneath, with an acuteV-shaped depression between the lateral sets on the upper side of the branchlet,and is best distinguished by the very prominent pulvini and grooves on the branchlets.The bark of the branches and young stems begins to scale very early, anunusual character in the silver firs, and conspicuous in this species, in A . homolepis,and in A . Webbiana. The resinous buds, glabrous shoots, and leaves with medianresin-canals are additional points in the discrimination of A , brachyphylla.DISTRIBUTION(A. H.)According to Mayr this tree occurs on the main island of Japan, between lat.36 and 38 , in the interior of the mountainous provinces; where it attains itsmaximum development in the zone of the beech forests, some trees attaining asmuch as 130 feet in height. The Japanese informed Mayr that it was also presenton the highest peaks of Shikoku ; but Shirasawa limits its distribution to the centralchain of Honshu above 3000 feet elevation, and says that it grows in mixture withbroad-leaved trees.Abies 767scattered through the forest of deciduous trees and attains a height of 100 feet ormore, the largest that I measured being 105 feet high by 16 feet in girth, and 95 feetby ii feet. Higher up the mountains it becomes mixed with Abies Veitchii anddiminishes in size. Its range of distribution is not accurately known, for thoughJapanese botanists distinguish it from the other species, as Dake-momi, the forestersand woodmen, who call all silver firs momi, do not seem, so far as I could learn,to distinguish it from A. Veitchii and A . Mariesii.CULTIVATIONThe date of introduction of this tree is not certainly known. Kent gives it asabout 1870, and Mr. H. J. Veitch tells me that he believes that the first seeds weresent by Dr. Regel from St. Petersburg, but it was at first grown under other names.It seems to thrive in most parts of England as well as or better than theother Japanese firs, but neither the trees I have planted, nor the seedlings I haveraised from Japanese seed, will live long on the calcareous soil at Colesborne;a moist climate in summer, and a deep sandy soil free from lime being apparentlythe most favourable conditions for its existence. At Kew it seems to grow fasterthan other firs.At Pampisford, Cambridge, a narrow conical tree measured in 1907, 44 feet by3 feet 6 inches. At Grayswood, Haslemere, a tree, obtained from Messrs. Veitchin 1882, measured in 1906, 41 feet by 3 feet 3 inches. Both these trees bear conesfreely. A specimen at Bicton is about 47 feet high, and bore cones in 1902. Thereis a very thriving one on the lawn at Bridge Park, Kent, planted by the Duke ofManchester in 1885, which now measures 30 feet high by 3 feet. At Dropmorethere is one which in 1908 was 32 by 2 feet. At Kew, where there are severalthriving trees, this species first produced 1 cones in this country in 1887. AtPencarrow, 2 a tree measured 40 feet by 3 feet 10 inches in 1907.In Scotland we have seen no specimen of any size. 8Kent figures a handsome tree at Castlewellan, Co. Down, which was about 35feet high and coning freely in 1907. Henry measured one at Fota, which was, in1903, 40 feet in height and 2 feet 8 inches in girth. Another at Glasnevin was, in1906, 38 feet by 2 feet 6 inches.This species 4 is very hardy in eastern Massachusetts, U.S., where it has alreadyproduced cones. A tree in Mr. Hunnewell's pinetum at Wellesley was 35 feet highin 1905. . (H. J. E.)l Card. Chron. ii. 248 (1887).1 This tree is figured under the erroneous name of A . Veitchii, in Hortus Veitchii, plate opposite p. 83 (1906).3 A . brachyphylla seems in every way adapted for cultivation in the north of Scotland, but too little is yet known of rateof growth to enable an opinion to be formed of its economic value. (J. D. CROZIER.)4 Sargent, in Silva, N. Amer. xii. 102 adnot. ( 1898), and The Pinetum at Wellesley in 1905, p. 12.To most Europeans it is the best known of the Japanese silver firs, as it growsabundantly at Chuzenji, a favourite tourist resort. Here at 4000 to 5000 feet it is

768 The Trees of Great Britain and IrelandABIES UMBELLATAAbies umbellata, Mayr, Abiet. Jap. Reiches, 34, t. i, f. 2 (1890).Abies umbilicata, Beissner, ex Mayr, Fremdland. Wald- u. Parkbaume, 2 58 (1906).This species is very imperfectly known; and there are no specimens in the Kewherbarium. The plants distributed in England some years ago under the name byMessrs. Veitch are identical with Abies homolepis, and, when seen by Mayr, werepronounced by him to be a form of A . brachyphylla.According to Mayr this species is distinguishable with difficulty, when in theyoung state, from A . brachyphylla, with which it agrees in the disposition and formof the leaves, and in the characters of the buds and shoots. Mayr states, however,that the leaves are not so white underneath as in A . brachyphylla. I have receivedfrom Herr Spath of Berlin, a specimen of reputed A . umbellata, which agrees generallyin buds, shoots, and foliage with A . brachyphylla ; but has leaves slightly longer than,and not so conspicuously white beneath, as is usual in that species.The cones, according to Mayr, resemble those of A . firma, and are verydifferent in size, colour, scales, and bracts, from those of A . brachyphylla; and aredescribed by him as greenish-yellow when growing, brown when mature, about 4inches long by i^ inch in diameter, cylindrical, the flattened apex having in its centrea raised umbo ; scales about ij inch broad by if inch long; bracts narrowed in themiddle, slightly shorter than the scales, only exserted at the base of the cone.According to Mayr, this species is only found in a few localities in Japan, butgrows in considerable quantity on Mount Mutzumine in the province of Musashi,where it occurs with A . brachyphylla in the beech forests. It grows also onthe lumonji-toge, leading from Musashi to Shinano, and is also probably not uncommon on the neighbouring mountains of Hida and Kai.This species has been united by Sargent, Kent, and others with A . firma; butit is very different from that species in foliage and shoots. It is possibly a hybridbetween A. brachyphylla, of which it has the foliage, and A . firma, which it resemblesin its cones.(A. H.)ABIES VEITCHII, VEITCH'S FIRAbies Veitchii, Lindley, Card. Chron. 1 861, p. 23; Masters, Card. Chron. xiii. 275 (1880), andJourn. Linn. Soc. (£ot.) xviii. 515, t. 20 (1881); Mayr, Monog. Abiet. Jap. Retches, 3 8, t. 2,f. 4 (1890); Kent, Veitch's Man. Coniferce, 541 (1900); Shirasawa, Icon. Ess. Forest. Japan,text 16, t. 5, ff. 23-42 (1900).Abies Eichleri, Lauche, Berlin. Gartenzeit. 1 882, p. 63.Pinus selenolepis, Parlatore, DC. Prod. xvi. 2, p. 427 (1868).Picea Veitchii, Murray, Prof. Roy. Hort. Soc. ii. 347, ff. 52-62 (1862).A tree, attaining 60 to 70 feet in height. Bark of trunk greyish and remainingsmooth even in old trees. Buds small, subglobose, purplish, resinous. YoungAbies 769branchlets smooth, brown, covered with moderately dense, short, erect pubescence,retained on the older branchlets, the bark of which becomes slightly fissured.Leaves on lateral branches arranged almost as in A . Nordmanniana; those on theunder side of the branchlet pectinate; those on the upper side shorter and coveringthe branchlet, the median ones pointing upwards and forwards, and not appressed somuch as in A . Nordmanniana. Leaves, about \ to i inch long, ^ inch wide, linear,flattened, gradually tapering to the base, uniform in width in the anterior half, witha truncate bifid apex; upper surface dark green, shining, with a continuous mediangroove and without stomata; lower surface with two conspicuously white, broadbands of stomata, in nine to ten lines; resin - canals median. On cone - bearingbranches, the leaves are more crowded, and less plainly pectinate below, than is thecase in barren branches.Staminate flowers* \ inch long on a stalk of the same length ; anthers stalked,connective developed into a saddle-shaped flap, from the back of which projects ahorizontal or deflexed spur-like process.Cones sessile or sub-sessile, cylindrical, flattened at the apex, 2 to z\ incheslong, f to i inch in diameter, bluish before ripening, brown when mature. Scalessmall; lamina f inch wide, f inch long, crescentic, with two lateral denticulatewings, which are separated from the narrow obcuneate base by rounded deepsinuses. Bract as long as the scale, obcuneate below, dilated above into a twowingeddenticulate lamina, ending in a short mucro, slightly exserted and reflexed.Seed-wing very broad and short, scarcely the length of the body of the seed ;seed with wing about f inch long.VARIETIESMayr distinguishes two forms of cones :1. Var. typica. Cones large, about z\ inches long; bracts exserted and reflexed.2. Var. Nikkoensis. Cones small, 2 inches long ; bracts scarcely visible, theirfine points projecting only slightly between the scales.Abies nephrolepis, Maximowicz,2 has been united with Abies Veitchii by Masters,and is perhaps a geographical form of the latter species, occurring in Amurland.According to Maximowicz it differs in the leaves of cone-bearing branchlets beingsometimes acute and not bifid, and in the smaller ovoid-cylindrical cones, the scalesof which are longer than the bracts and less in size than those of the Japanese tree.This Manchurian tree has not apparently been introduced into cultivation and is stillimperfectly known.Abies Eichleri, Lauche, was supposed to have been raised from seeds sent fromTiflis to Potsdam; and was considered to be a new species from the Caucasus.Some error, however, had arisen, as the plants turned out to be identical with AbiesVeitchii.Abies Veitchii has been collected according to Beissner 3 by Pere Giraldi at1 Masters, loc. fit.IVMil. Biol. vi. 22 (1866). 3 See Journ. Linn. Soc. (But.) xxvi. 557 (1902).I

77 The Trees of Great Britain and Ireland9800 feet elevation in the Peling Mountains in the province of Kansu in China.The identification of herbarium specimens of Abies is difficult, and the Kansu plantwill probably turn out to be a new and distinct species.IDENTIFICATIONThis species, with leaves covering the branchlet on the upper side, which arevery white beneath, truncate and bifid at the apex, and less appressed than is thecase in A . Nordmanniana, is further characterised by its small resinous buds, medianresin-canals in the leaves, and smooth branchlets with short erect pubescence. Thedistinctions between it and A. Mariesii are given under the latter species.HISTORY AND DISTRIBUTION(A. H.)Abies Veitchii was discovered on Fuji-yama by J. Gould Veitch in 1860.According to Sargent J it was introduced by Mr. T. Hogg into Parson's nurseriesin Flushing, New York, in 1876, and a plant raised there was 16 feet high in1889. It was cultivated in the United States for a time under the name of Abiesjaponica. It was not known in England or on the continent until 1879, whenseeds were sent home 2 to Messrs. Veitch by their collector Maries.The best account of the distribution is given by Mayr, who considers the treeto be the typical silver fir of the cold region of Japan, a zone which doesnot occur in Kiushu, where there are no mountains high enough. In Shikoku,A. Veitchii is very rare, only about 200 trees being known, which grow on thesummit (6600 feet elevation) of Ishitzuchi-yama. It extends in the main island ofJapan over the central mountain chain, from Fuji-yama to lat. 39 , growing atelevations of 6600 feet and upwards. Mayr denies its occurrence beyond lat. 39 ;and states that north of this line it is replaced by Abies Mariesii? which thus intervenes over three degrees of latitude between the southern region, occupied byA. Veitchii, and the northern region, occupied by A . sachalinensis, these two speciesnot meeting at any point, and having no transitional forms. A . Veitchii eitherforms pure woods or is associated with Picea hondoensis and Picea Alcockiana,but never with Piceapolita. Sometimes it is mixed with Tsuga diversifolia or withA. Mariesii.Shirasawa gives its lower limit of altitude in the main island as 5000 feet, andstates that it attains about 70 feet in height by 7 feet in girth.According to Mayr and Matsumura, the Japanese name, which is exclusivelyapplied to this species, is Shirabiso. Shiramomi is also another name for the tree.So far as I could learn the tree is of no special economic value in Japan.1 Garden and Forest, ii. 589 (1889). In this journal, x. 511 (1897) the statement is made that Mr. Hogg introducedit some forty years earlier, evidently a mistake for twenty years. A. Veitchii is very hardy in the United States, where it hasproduced cones. 2 Horttts Veitchii, 3 37 (1906).3 Prof. Miyabe showed me, in his herbarium, a barren specimen with small leaves, from Samani, near Cape Erimo,in eastern Hokkaido, which he believed to be .4. Vcitihii. This was A. sachalinensis, Masters, var. nemorensis, Mayr.Abies 771CULTIVATIONIt seems to grow fairly well though rather slowly on soils which contain nolime; but it will not live on the calcareous soil at Colesborne.We have seen no trees of considerable size. One at Tregrehan near St.Austell, measured 27 feet by 2 feet in 1908; and another at Ochtertyre, Perthshire, measured 30 feet by 2 feet in the same year. Mr. Bean 1 noticed in 1906a specimen 31 feet high at Murthly Castle, and another 20 feet high at DalkeithPalace. Small specimens will be found in most collections of conifers; and theyoung trees at Kew at present appear to thrive better than most species of Abies.(H. J. E.)ABIES MARIESII, MARIES' FIRAbies Mariesii, Masters, 2 Card. Chron. xii. 788, f. 129 (1879), and Journ. Linn. Soc. (Bot.)xviii. 519 (1881) ; Mayr, Abiet. Japan. Reiches, 40, t. 2, f. 5 (1890); Shirasawa, Icon. EssencesForest. Japan, text 15, t. 4. ff. 15-28 (1900); Kent, Veitch's Man. Cotiifera, 5 20 (1900).A tree, attaining in Japan about 80 feet in height and 6 feet in girth.Buds small, globose, resinous; terminal buds on strong shoots are girt at thebase by a ring of ovate, acuminate, rusty-red pubescent scales. Young shootsdensely covered with a rusty-red tomentum, retained more or less in older shoots,the bark slightly fissuring in the third year.Leaves on lateral branches arranged as in A bies Nordmanniana, the medianleaves on the upper side almost appressed to the stem in imbricating ranks, andabout i to f the length of the lower leaves, which spread pectinately outwards andslightly forwards in the horizontal plane. Leaves linear, flattened, tapering at thebase and gradually widening beyond the middle, so that their broadest part is in theupper third ; about f inch in maximum length, ^ to ^ inch wide; apex roundedand bifid ; upper surface yellowish green, shining, with a continuous median grooveand without stomata ; lower surface with two white bands of stomata, each of eight ornine lines; resin-canals marginal. Leaves on cone-bearing branches all appressedmore or less to the shoot, upturned, and shorter than on barren branches.Cones sessile, deep blue with a velvety lustre before ripening, dark brown whenmature, ellipsoid, with an obtuse apex, about 4 inches long by 2 inches in diameter.Scales fan-shaped; lamina i inch wide, £ inch long, upper margin undulate, lateralmargins with two denticulate wings; claw broadly obcuneate. Bract with a broadobcuneate claw, expanding just above the base of the scale, into a broadly ovallamina, which is emarginate at the apex with a short mucro. Seed-wing nearly twicethe length of the body of the seed ; seed with wing about f inch long.The cones show that the tree is nearly related to A bies Webbiana ; but it differsentirely from that species in the characters of the branchlets and foliage.1 Cf. Kew Bulletin, 1 906, pp. 260, 268.2 Allies Mariesii, Masters, Bot. Mag. t. 8098 (1906), described from a tree at Dunphail, Morayshire, is referable toA. Welliana, as mentioned in our account of the latter species.

7 7 2 The Trees of Great Britain and IrelandIDENTIFICATIONThis species is similar in the arrangement and size of the leaves to A . Veitchii;but is distinguishable from that and from all other species of Abies, by the rusty-redor chocolate colour of the densely tomentose branchlets. The leaves are shorter andbroader in proportion than those of A . Veitchii, being widest in their upper third,with their apex rounded and not truncate as in that species. The two species differalso in the position of the resin-canals.(A. H.)HISTORY AND DISTRIBUTIONThis species was discovered 1 in 1878, by Charles Maries, when collecting forMessrs. Veitch, on Mount Hakkoda near Aomori in northern Hondo; and for someyears it was supposed to occur only in the main island of Japan, where Mayr givesits distribution as from lat. 36 to the extreme northerly point of the island. It hassince been found, according to Sargent, 2 by Tokubuchi in 1892 in one place on theshores of southern Yezo ; 3 and Dr. Honda lately discovered it in Formosa on MountMorrison at 10,000 feet elevation.Sargent saw it on Mount Hakkoda, and says that it is common at about 5000feet, scattered amongst deciduous trees, and is the only species of A bies in thislocality, where it forms a compact pyramid, 40 to 50 feet high, with crowded branchesand many large dark purple cones. Maries also found it on Nantai above Nikko,which I had not time to ascend ; Mr. Tome Shirasawa, who was my companion inNorth Japan, says that it grows here in company with Abies Veitchii on the upperzone of the mountain at 7000 to 8000 feet. The tree according to Mayr is thesmallest of all the Japanese silver firs, its maximum height being 80 feet, with agirth of about 6 feet. It is known in Japan as Aomori-todo-matsu, and, so far as Icould learn, has no economic value.CULTIVATIONSeeds were sent home by Maries in 1879, but gave poor results; and we havenot found anywhere in this country a single tree of any size ; but Mr. Bean 4 has seena small but healthy tree at Scone Palace in Perthshire. As seen in the nursery atKew and Coombe Wood, it is very slow and feeble in growth, and apparently is notsuited to the English climate, young plants usually having very small leavesand short shoots. There are, however, three flourishing young trees at Bayfordbury,which were obtained from Hesse's nursery at Weener, in Hanover. It seemsto do very much better in America, where I saw a vigorous tree 5 growing at Mr.Hunnewell's pinetum at Wellesley, Massachusetts. Reputed trees of Abies Mariesiiusually turn out on examination to be Abies Veitchii.(H. J. E.)1 Hortus Veitchii, 336 (1906). * Forest Flora of Japan, 82 ( 1894).3 l!ut Prof. Miyabe told me in 1904 that he had seen no specimens from this place, and doubted its occurrence inHokkaido. lie had specimens in his herbarium from Nambu near Morioka. 4 Card. Chron. xli. 117 (1907).6 Reported to be 9 feet high, by Sargent, in The Pinetum at Wellesley in i9oj, p. 13.Abies 773ABIES GRANDIS, GIANT FIRAbies grandis, Lindley, Penny Cycl. i . 30 (1833); Masters, Card. Chron. xv. 179, ff. 33-36 (1881)xvii. 400 (1882), and xxiv. 563, f. 1 28-131 (1885), and Journ. Linn. Soc. (Bot.\ xxii. 174(1886); Sargent, Silva N. Amer. xii. 117, t. 612 (1898), and Trees N. Amer. 60 (1905);Kent, Veitch's Man. Conifera, 5 10 (1900).Abies Gordoniana, Carriere, Cant/. 2 98 (1867).Abies amabilis, Murray, Proc. Roy. Hort. Soc. iii. 310 (1863) (not Forbes).Pinus grandis, Hooker, Fl. Bor. Amer. ii. 163 (1839).Picea grandis, Loudon, Arb. et Frut. Brit. 2 341 (in part) (1838).A tree attaining in America in the coast regions 300 feet in height and 16 feetin girth ; but on the mountains of the interior rarely more than 100 feet high by6 feet in girth ; often smaller and stunted at high elevations. Bark of young treessmooth, thin, and pale ; of older trees in America, brownish, divided by shallowfissures into low flat ridges roughened by thick appressed scales ; in cultivated treesfissuring into thin irregular plates, exposing the reddish brown cortex. Buds small,conical, obtuse at the apex, resinous, roughened by the raised tips of the scales.Young shoots olive-green, smooth, with a minute, erect, not dense pubescence.Leaves on lateral branchlets pectinate, in two lateral sets in the horizontal plane,each set of apparently two ranks, the upper rank with leaves about half the length ofthose below. Leaves linear, flattened, up to about ij to 2 inches long, ^ to ^ inchin width, narrowed at the base, uniform in breadth elsewhere, with a rounded andbifid apex; upper surface dark green, shining, with a continuous median groove andwithout stomata; lower surface with two white bands of stomata, each of abouteight lines; resin-canals marginal. Leaves on cone-bearing branches crowded, lessspreading or nearly erect, blunt or bifid at the apex, shorter than on sterile branches.Cones 2 to 4 inches long by i to \\ inch in diameter, cylindrical, slightlynarrowed towards the rounded or retuse apex, bright green in colour, with the bractsconcealed. Scales resembling those of Abies Loiviana, but smaller. Bractsituated a little above the base of the scale, quadrangular; upper margin broad,denticulate, deeply emarginate, and with a minute mucro. Seeds f inch long, lightbrown, with pale shining wings about | inch long.IDENTIFICATIONAbies grandis is readily distinguished by the very flat pectinate arrangement ofthe leaves ; those of the upper rank being about half the length of those in thelower rank. Abies Loiviana, when growing feebly, resembles it somewhat in arrangement ; but in this species the upper surface of the leaves has stomatic lines, absentin A . grandis, and the leaves in the upper rank are only slightly shorter than thosein the lower rank.(A. H.)

774 The Trees of Great Britain and IrelandDISTRIBUTIONOn the north-west coast of America this magnificent tree has a wide range,from Vancouver Island, where it grows at low levels and is not, so far as I saw, aconspicuous feature in the forest; through Washington and Oregon as far south asMendocino County in California, where it does not extend far from the coast, andgrows in company with Sequoia sempervirens and Picea sitchensis.Inland it is less abundant on the eastern slopes of the Cascade Mountains, butextends to the Cceur d'Alene and Bitter-root Mountains of Idaho and Montana.In the Flathead Lake Country it is a comparatively small tree, attaining only12 to 15 inches in diameter, and ascending to about 3500 feet.It reaches its maximum development in a damp climate and in sheltered valleys,where I have measured trees much over 200 feet in height, and where, according toSargent and Sheldon, it sometimes reaches as much as 300 feet. So far as I saw,and Sargent confirms this observation, it never grows gregariously, but scatteredamong other species; and rarely forms an important element in the timber.It is easy to recognise when young by the flat arrangement of the leaves,but when its branches are far above one's head I could not distinguish it fromA. amabilis in the Cascade mountains, or from A . Lowiana which seems to take itsplace in southern Oregon and northern California.It grows very fast in its own country, a specimen measured at 2500 feet altitudeon the Cascade Mountains being 140 feet by 16 feet on the stump, at only 106 yearsold. Though the timber is not much valued by lumbermen, it is used for variouspurposes locally, and, according to Sheldon, makes the most durable shakes a nameused for large shingles cleft with the axe used in Oregon.The tree figured (Plate 218) was growing in 1904 on Swallowfield farm, aboutfifty miles north of Victoria, in Vancouver Island, and when I measured it, was215 feet by 19 feet.Abies grandis was discovered on the Columbia river by Douglas in 1825,though he does not seem to have sent seeds to the Horticultural Society until 1831or 1832. Very few of these germinated, and it is doubtful if any of the originalseedlings are still living. 1 The next consignment 2 of seed was sent by William Lobbin 1851 to Messrs. Veitch at Exeter; and about the same time seeds were receivedby the Scottish Oregon Association from their collector Jeffrey.REMARKABLE TREESThere are many fine trees of this species in the warmer and moister parts ofEngland, Scotland, and Ireland; and, with A . nobilis and A . Lowiana, it seemsbest suited of all the American firs to our climate. 81 Murray, Proc. Roy. Hort. Soc. 3 11 (1863), states that there were then living no authentic seedling specimens of A . grandisraised from the seeds sent by Douglas, but a multitude of young plants existed which had been raised from cuttings.2 Horlus Veitchii, 336(1906).3 The most vigorous of all the genus. Thrives admirably on gneiss, free from all trace of disease, is not susceptibleto frosts or Chermes, and as a shade bearer has no equal amongst silver firs. Produces timber, which is white andAbies 775The tallest that I have seen in England grows in Oakly Park nearLudlow, the property of Lord Plymouth, on the rich flat by the river Teme, andmeasured 102 feet by 8£ feet in 1908. Other fine trees are at Fonthill Abbey, whichwas 98 feet by 8 feet in 1906 ; and at Madresfield Court 1 and Eastnor Castle, bothof which are over 95 feet high and 7$ feet in girth. The latter is figured (Plate217). I have seen several others over 90 feet, of which perhaps the one atHeanton Satchville is the largest, though it is too spreading to be a typicalspecimen. In 1903 it was about 94 feet by 9 feet 7 inches, and 56 yards in circumference of the branches. At Castlehill there are some fine trees, one of whichmeasured 92 feet by 7 feet 10 inches in 1904. At Petworth there is a very tall but nota well-grown tree, 94 feet by 6 feet 6 inches. At Eridge Park a very handsometree, planted by Mr. Disraeli in 1868, measures 76 feet by 6£ feet. At Youngsbury,Ware, in Herts, Mr. H. Clinton Baker measured a tree in 1907, which was 91feet in height and 9 feet 8 inches in girth; and at his own place, Bayfordbury,there is another, 73 feet by 5 feet 9 inches in 1905. He also reports two goodtrees at The Heath, Leighton Buzzard, 98 feet by 8 feet 6 inches, and 88 feetby 7 feet 10 inches respectively. There is also a very large tree in a belt by theroad at Flitwick Manor, near Ampthill, Bedfordshire, the seat of Miss Brooks,which is 95 feet by 10 feet. At Welford Park there are two trees which thoughonly planted in 1878, are now about 90 feet high by 7 feet in girth.At Barton a thriving tree, 68 feet by 4 feet 3 inches, is the best we know in theeastern counties, and this is sheltered and drawn up by tall trees around it.At Golden Grove, Carmarthenshire, there is a fine tree which in 1892 was60 feet by 7 feet 8 inches, and when I saw it in 1905 had increased to 80 feet.In Scotland there are many fine specimens, of which one at Riccarton, in Midlothian, was reported at the Conifer Conference in 1891 to have been 83 feet 3 incheshigh and only 3 feet 8£ inches in girth, as carefully measured by the owner, SirJames Gibson Craig; and stated by him to have grown 53 feet in twelve years.Soon after this it was attacked by Chermes and was cut down.At Glenlee, near New Galloway, Mr. T. R. Bruce informs me that there is atree, planted by Mrs. Melville in 1864, which in 1905 measured no less than 95 feetby 10 feet, though, having lost its leader four years previously, it has now threeleads. At Castle Kennedy this species grows much faster than any of theother numerous firs planted there. In 1904, one of two trees, nearly equal in size,was 78 feet high by 6 feet in girth. This tree* was only twelve years old in 1891,when it measured 30 feet by i foot 7 inches.At Benmore, in Argyllshire, one of the wettest places in Scotland, a tree saidto be only thirty-five years planted was, in 1907, 80 feet by 7 feet 4 inches; but thetrunk was infested with scale and did not seem to be healthy when I saw it. AtPoltalloch, there is a fine specimen over 80 feet high, and at Inveraray andsomewhat soft, in great volume, and which is found useful in connection with box-making and other industries in Aberdeen.Specially adapted for cultivation for profit where a large volume of timber is a desideratum. (J. D. CROZIER.)1 A note signed J. N. in the Trans. Scot. Art. Soc. xx. 126 (1907) states that this tree, in Sept. 1906, was 114 feetby 8 feet 4 inches. When I measured it in 1904, I made it 96 feet by J % feet; and though owing to the ground I could notget a level base line, I can hardly believe that it is now so tall as stated.2 Jonrn. Roy. Hort. Soc. xiv. 547 (1892).

776 The Trees of Great Britain and IrelandArdkinglas, in the same county, are trees over 70 feet, which in that wet climateflourish exceedingly.In the Keillour Pinetum, near Balgowan, in Perthshire, a tree growing in boggysoil was, in 1904, 90 feet high by 7 feet 3 inches. It is not well furnished above,and is perhaps beginning to suffer from the nature of the soil. At Keir, Dunblane,there is a tree which in 1904 measured 82 feet high by 9 feet 3 inches in girth.This tree 1 was twenty-eight years old in 1891, and then measured 55 feet by 4 feet2 inches. At Abercairney, Perthshire, there is a fine tree, which in 1904 was 91feet by 8 feet 4 inches. This tree 2 was about thirty years old in 1891, and thenmeasured 58 feet by 4 feet 6 inches.At Durris, Aberdeenshire, there is a good tree, which Mr. Crozier measuredin 1904 as 82 feet high by 9 feet 6 inches in girth. When 1 saw it in 1907 ithad increased to nearly 90 feet.The largest tree in Ireland was formerly at Carton, which was reported in1891 to be 80 feet high by 6 feet in girth. The top was blown off by the gale ofFebruary 1903, and when seen by Henry in the autumn of that year, the treemeasured 67 feet by 9 feet 6 inches. At Kilmacurragh, Co. Wicklow, a finespecimen was, in 1906, 86 feet by 7 feet 2 inches ; and at Coollattin, in the samecounty, another measured 63 feet by 6 feet 4 inches. At Powerscourt I measuredone in 1903 which was about 87 feet by 7^ feet.Abies grandis thrives very well in north-western Germany, and according toCount Von Wilamitz-Mollendorf s grows at Gadow faster than any other silver fir,a specimen figured being 25 metres by 1.40 metre when only twenty-five yearsold. It also succeeds in some parts of Denmark, where Hansen 4 states that aspecimen planted in 1864 had attained, in 1891, 53 feet by 6 feet. (H. J. E.)1 Journ. Roy. Hort. Soc. xiv. 531 (1892).3 Mitt. D. Dcndr. Ges., 1907, p. 138.1 Ibid. 527.4 Journ. Roy. Hort. Soc. xiv. 469 (1892).Abies 777ABIES CONCOLOR, COLORADO FIRAbies concolor, 1 Lindley and Gordon, Journ. Hort. Soc. v. 210 (1850); Masters, Journ. Linn. Soc.(^/.)xxii. 177, ff. 8-11 (1886), and Card. Chron. viii. 748, ff. 147, 148 (1890); Sargent,Silva N. Amer. xii. 121, t. 613 (1898) (in part), and Trees N. Amer. 6 2 (1905) (in part);Kent, Veitch's Man. Conifera, 501 (1900).Picea concolor, Gordon, Pinetum, 1 55 (1858).Picea concolor, var. violacea, Roezl, ex Murray, Card. Chron. iii. 464 (1875).Pinus concolor, Engelmann, ex Parlatore, in DC. Prod. xvi. 2, p. 427 (1868).A tree attaining in America 100 to 125 feet in height, with a girth of 9 feet.Bark of old trees fissuring into small irregular plates. Buds, much larger thanthose of A . Lowiana, broadly conical, rounded at the apex, brownish, resinous,and slightly roughened by the raised tips of the scales. Young shoots smooth,yellowish-green, with a minute scattered pubescence, variable in quantity and oftenabsent from the greater part of the branchlet. Second year's shoot greyish andirregularly fissuring.Leaves on lateral branchlets irregularly arranged and not truly pectinate; mostof the leaves extending laterally outwards and curving upwards, a few on the lowerside directed downwards and forwards, some on the upper side directed upwards andforwards ; those above shorter than those below. Leaves up to 2 to 3 inches long,fa inch broad, glaucous on both surfaces, linear, flattened, slightly tapering at thebase, uniform in width elsewhere ; apex acute or rounded and not bifid, thoughoccasionally a slight emargination is discernible with a lens; upper surface slightlyconvex, not grooved, with fifteen to sixteen regular lines of stomata; lower surfaceconvex with two bands of stomata, each of about eight irregular lines, not conspicuously white ; resin-canals marginal. Leaves on cone-bearing branches shorter,thicker, falcate, all curving upwards.Cones, 3 to 5 inches long, i^ inch in diameter, cylindrical, narrowed at bothends, rounded or obtuse at the apex; greenish or purple before ripening, brownwhen mature. Scales of native Colorado specimens much broader than long ; laminaabout I inch wide by £ inch long, upper margin entire, lateral margins rounded anddenticulate, gradually passing into the obcuneate claw or with a slightly auricledtruncate base. Bract, at the base of the scale, rectangular, denticulate, with truncateupper margin and a minute mucro; in some specimens deeply bifid above. Seeds^ inch long, with broad shining pinkish wings, about £ inch long. In cultivatedspecimens, both brown and purple cones occur.The following varieties have arisen in continental nurseries :r. Var. falcata, Beissner, 2 leaves sickle-shaped, curving upwards.2. Var. glabosa, Beissner, 2 globose in habit, with symmetrical short branches.1 According to the view taken here, Abies concolor includes only the tree found in Colorado, Utah, and SouthernCalifornia. Sargent and other American botanists combine with this species the tree found in the Californian Sierras, whichis considered by us to be a distinct species, A. Lowiana. The two forms differ remarkably in buds and foliage ; and it ismost convenient to regard them as distinct species. t Mitt. Deut. Dendr. Ces. 1905, p. 112.IVK

778 The Trees of Great Britain and Ireland3. Var. aurea, Beissner, 1 young foliage golden yellow, gradually changing to asilvery grey colour.4. Var. brevifolia, Beissner, 1 leaves short, thick, obtuse, twice as broad as in thetypical form.DISTRIBUTIONAbies concolor occurs in the Rocky Mountains of southern Colorado and extendssouthwards over the mountains of New Mexico and Arizona into northern Mexico,being the only silver fir in the arid regions of the Great Basin and of southernNew Mexico and Arizona. It occurs also in Utah in the Wasatch Mountains, andin southern California, in the San Bernardino and San Jacinto Mountains. Itis accordingly confined to dry regions, while Abies Lowiana, which is in allprobability only a geographical form of it, occurs in the more rainy regions of theSierra Nevada of California and the southern mountains of Oregon. According tosome opinions, the three species, Abies grandis, Abies Lowiana, and Abies concolorare only geographical forms of one large species.Sargent says, of Abies concolor, that it endures heat and dryness best of all thesilver firs of North America, and its distribution is accordingly more southerly thanthat of the other species, which occur in the United States.HISTORY AND CULTIVATIONThis species was discovered by Fendler, near Sante F6, in 1847, and was firstclearly described by Parlatore, who adopted for it Engelmann's MS. name, Pinusconcolor. I t does not appear to have been introduced 2 into cultivation until about1872. Syme mentions 8 two-year-old seedlings of it as a new species in 1875.Roezl, apparently in 1874, sent specimens and seeds, which were labelled Piceaconcolor violacea,* from New Mexico to Messrs. Sanders and Co., St. Albans. Thisspecies has been much confused with A . Lowiana, which was introduced considerablyearlier. It is probable that there are no trees of true A . concolor in cultivation,older than 1873 or 1874.Abies concolor, according to Sargent, is the only American silver fir, which isreally successful in cultivation in the eastern part of the United States, where itgrows better than A. Lowiana.We have seen few trees of large size, though one at Highnam Court, Gloucestershire, of no great age, was 44 feet by 2 feet 9 inches in 1908.It is less common in cultivation than A . Lowiana, which it much excels inbeauty of foliage. Mr. Crozier says that young trees growing at Durris are quitehealthy. (A. H.)1 Mitt. Deut. Dendr. Ges. 1906, p. 144.2 Roezl sent a few seeds in 1872. Cf. Lavallee, Nouveaux Coniferes du Colorado et de la Califomit, vnjourn. Sac.Cent. Hort. France, viii. (1875). 3 Card. Chron. iii. 563 (1875). * IKd- 464-Abies 779ABIES LOWIANA, CALIFORNIAN FIRAbies Lowiana, A. Murray, Proc. Key. Hort. Soc. iii. 317 (1863).Abies lasiocarpa, Masters (not Nuttall or Murray), Card. Chron. xiii. 8, f. i (1880).Abies grandis, Lindley, van Lowiana, Masters, Journ. Linn. Soc. (-Bot.), xxii. 175, ff. 6, 7 (1886).Allies concolor, Sargent, Silva N. Amer. xii. 121 (1898), and Trees N. Atner. 62 (1905) (in part).Abies concolor, Lindley and Gordon, var. lasiocarpa, Beissner, Handb. Conif. 7 1 (1887).Abies concolor, Lindley and Gordon, var. Lowiana, I .emmon, W. Amer. Cone-Bearers, 64 (1895);Kent, Veitch's Man. Conif. 5 02 (1900).Picea Lmviana, Gordon, Pinet. Suppl. 5 3 (1862).Picea Parsonsiana, Barren, Catalogue, 1 859, and Card. Chron. \. 7 7 (1876).Pinus Lowiana, M'Nab, Proc. R. Irish Acad. ii. 680 (1877).A tree, attaining on the Californian Sierras 200 to 250 feet in height, with atrunk often 18 feet in girth. Bark in cultivated specimens as in A . concolor; in wildtrees becoming, near the ground, on old trunks, very thick and deeply divided intobroad, rounded, scaly ridges. Buds ovoid, blunt at the apex, brownish, resinous,roughened by the raised tips of the scales. Young shoots yellowish green, smooth,covered with a minute scattered pubescence.Leaves on lateral branchlets pectinately arranged, each lateral set of about tworanks, directed almost horizontally outwards, or curving upwards and outwards, so asto assume above a V-shaped arrangement. None of the leaves are directedirregularly in the middle line; and those of the upper rank are only slightly shorterthan those of the lower rank. Leaves, up to 2^ inches long, about TX2 inch broad,linear, flattened, slightly tapering at the base, uniform in width elsewhere, roundedand bifid at the apex; upper surface with a wide median furrow, usually notcontinued to the apex, and with eight lines of stomata in the furrow; lower surfacewith two white bands of stomata, each of eight to nine lines ; resin-canals, marginal.Leaves on cone-bearing branches, upturned.Cones, according to Sargent, not distinguishable from those of Abies concolor.Wild specimens, however, from California slightly differ, in having larger scales andbroader bracts. Cultivated specimens in England bear cones which are chestnutbrown,and apparently never purple, as is often the case in Abies concolor.IDENTIFICATIONAbies Lowiana is regarded by Sargent and other American botanists as a formof A . concolor. As seen in cultivation it is very distinct from that species : moreover, it has a different distribution in the wild state. We have kept it separate, asbeing more convenient to cultivators.In practice it can only be confused with A . grandis, and true A . concolor. Thecharacters distinguishing it from A . grandis are given under this species on p. 773.In A . concolor the arrangement of the leaves is irregular, not being trulypectinate. Many of the leaves in the middle line, both above and below, are not

780 The Trees of Great Britain and Irelanddirected outwards, but point forwards parallel to the axis of the branchlet. In A.concolor the leaves are entire at the apex, and their convex upper surface shows sixteen lines of stomata, and is without a groove; whereas, in A . Loiviana, the apex ofthe leaves is bifid, and their upper surface is grooved, showing eight lines ofstomata. The buds are smaller in the latter species. (A. H.)DISTRIBUTIONAbies Lowiana is found on the Siskiyou Mountains in southern Oregon, and onMt. Shasta and the Sierra Nevada ranges in California. Its northern limit is the dryinterior of southern Oregon, near the divide between the headwaters of the Umpquaand Rogue rivers, which, according to Sargent, is the real northern boundary of theCalifornian flora. 1 With Abies magnified it forms in great part one of theprincipal forest belts on the west slope of the Sierra Nevada Mountains for 450 miles,and extends from 4000 to 9000 feet above sea-level. Here I saw it on my way intothe Yosemite Valley in 1888, but did not then measure any trees. I found it inSeptember 1904 in company with A . magnified, Pinus ponderosa, and PinusLambertiana abundant on Mount Shasta, from about 3000 to 6000 feet; and here itwas of moderate size, the largest that I measured being 140 feet by u feet 8 inches.It attains, however, 200 to 250 feet on the Sierra Nevada, and as much as 200 feetin Oregon.HISTORY AND CULTIVATIONAbies Lowiana was introduced from the Sierra Nevada of California by WilliamLobb in 1851 ; and about the same time seeds were sent from southern Oregon byJohn Jeffrey, who collected for the Scottish Oregon Association. The plants raisedfrom Lobb's seeds were distributed by Messrs. Veitch of Exeter as Picea lasiocarpa,while those raised in Scotland from Jeffrey's seeds were distributed as Piceagrandis?Messrs. Parsons of Flushing, United States, received seeds from California in1853; and plants raised from these were imported to England in 1855 by Messrs.Low of Clapton. These passed into commerce as Picea Parsonsiana, a name whichfirst appeared in Barren's Catalogue in 1859, and as Picea Lo-wiana, the name givenby Gordon in 1862.Of all the western silver firs this seems to be the most accommodating to thevaried conditions of England, growing well on soils where A . nobilis will not thrive,and in a drier climate than A . grandis prefers. It is usually grown under the nameof A . lasiocarpa, in the pineta which I have visited, and generally seen in goodhealth and with a symmetrical top; as it is not so liable to become stunted by theproduction of cones as A. nobilis.According to Sargent, the Californian form of A . concolor grows in the easternAbiesStates with less vigour and rapidity than the Colorado form ; but is equally hardy,and has attained 40 to 50 feet in height in New England.REMARKABLE TREESAmong the numerous trees that I have measured I find it difficult to saywhich is the finest specimen. The one at Linton Park was the largest recorded atthe time of the Conifer Conference, when it was 64 feet by 8 feet 7 inches. In 1902I found it to be 85 feet by ro feet 6 inches, a great increase in ten years (Plate 219).There is, however, a tree at Fonthill Abbey which I believe to be A . Loiuiana,though I could not reach the branches in order to identify it, which, in 1906,measured 90 feet by 6£ feet, and resembled, by its short branches, the typical habitof A . magnified.At Highnam Court, Gloucester, there is a fine specimen which was figured byKent; according to Major Gambier Parry, it measured 77 feet by 9 feet 2 inches in1906. I made it 80 feet by 9^ feet in 1908. Another at Eastnor Castle is about88 feet by 7 feet 4 inches.In Herts there are several good trees, one at Essendon Place being 82 feethigh by 5 feet 9 inches in 1907 ; another at Youngsbury, Ware, which was plantedin 1866, being 68 feet by 5 feet 6 inches in the same year; and a third at Bayfordbury,which was 69 feet by 6 feet 9 inches in 1905.A very remarkable specimen, narrow and almost columnar in habit, which wasplanted twenty-six years ago, was seen by Henry at Crowsley Park, Oxfordshire,the seat of Colonel Baskerville, and measured, in 1907, 71 feet by 6 feet.In Wales there is a very fine tree at Hafodunos, which Henry measured, in1904, as 87 feet by 7 feet 9 inches; and I saw one at Glanusk Park in Breconshire,which was over 80 feet high in 1906.In Scotland this species is not usually so large as in the south, though itgrows well even in the west, where I have seen good trees at Inveraray andPoltalloch; and in the reports of the Conifer Conference it is generally described asthriving, and several trees of 40 to 50 feet high are mentioned. The largest wehave heard of is at Abercairney, mentioned by Mr. Bean 1 as 65 feet by 5 feet.In Ireland the tree does not appear to have been often planted, and the largestreported at the Conifer Conference in 1891 was growing at Abbeyleix in Queen'sCounty, and measured 45 feet by 6 feet ro inches. At Coollattin, Co. Wicklow,another was, in 1906, 52 feet by 4 feet 9 inches; and, in the same year, a finespecimen at Castlewellan, Co. Down, measured 67 feet in height and 9 feet ingirth. (H. J. E.)1 fCcw Bulletin, 1 906, p. 258, and Card. Chron. xli. 168 (1907), where it is named through inadvertence A. concolor.1 The fir named A. concolor by Plummer in his Report on the Mt. Rainier Forest Reserve, p. 101 (Washington, 1900),is evidently A. grandis, which he does not mention, and all his references to white fir no doubt relate to that species.' Cf. Hoitus Veitchii, 39, 335 (1906).

782, The Trees of Great Britain and IrelandABIES AMABILIS, LOVELY FIRAbies amabilis, Forbes, Pitut. Woburn. 1 25, t. 44 (1840); Masters, Journ. Linn. Soc. (£ot.) xxii.171, t. 2 (1886), and Card. Chron. iii. 754, f. 102 (1888); Sargent, Silva N. Amer. xii. 125,t. 614 (1898), and Trees N. Amer. 5 9 (1905); Kent, Veitch's Man. Conif. 489 (1900).Abies grandis, Murray, Proc. Roy. Hort. Soc. iii. 308 (1863) (not Lindley).Pima amabilis, Douglas, Comp. Bot. Mag. ii. 93 (name only) (1836); Antoine, Conif. 6 3(1846).Pinus grandis, Don, in Lambert, Pimts, iii. t. (1837).Picea amabilis, Loudon, Arb. et Frut. Brit. iv. 2342 (in part) (1838).A tree sometimes attaining in America 250 feet in height and 18 feet in girth,but at high altitudes and in the north usually not more than 80 feet. Bark thin,smooth, pale or silvery white ; becoming, on very old trunks, thick near the groundand irregularly divided into small scaly plates. Buds small, globose, resinous,smooth, with purple scales all immersed in the resin, except occasionally two orthree, small and keeled, at the base of the bud. Young shoots grey, smooth,densely covered with short, loose, wavy pubescence.Leaves on lateral branches arranged as in A . Nordmanniana, up to i^ to i J inchlong by jJj inch broad, fragrant, linear, flattened, gradually tapering from themiddle to the base, slightly broader in the anterior half, with a truncate and bifidapex ; upper surface very dark green and lustrous, with a continuous median grooveand without stomata; lower surface with two broad white bands of stomata, each ofeight to ten lines; resin-canals marginal. Leaves on vigorous leading shoots acutewith long rigid points, closely appressed or recurved near the middle. Leaves oncone-bearing branches upturned, acute or acuminate.Cones ovoid-cylindric, slightly narrowing to the rounded apex, dark purplewhen growing, brown when mature; 3^ to 6 inches long by 2 to 2^ inches indiameter. Scales, i to i^ inch wide, nearly as long as broad, inflexed at the upperrounded margin, gradually narrowing towards the base. Bracts rhombic or obovateoblong;lamina situated just above the base of the scale and ending in a longacuminate tip, which reaches half the height of the scale. Seeds light yellowishbrown, J inch long, with oblique pale brown shining wings about J inch long.Abies amabilis resembles A . Nordmanniana in the arrangement and size of theleaves; but is readily distinguished from it by the small globose resinous buds.The leaves are also much darker, shining above, more truncate at the apex; andemit, especially when bruised, a strong fragrant odour which resembles that ofmandarin orange peel.(A. H.)DISTRIBUTION AND HISTORYAbies amabilis occurs on mountain slopes and terraces from British Columbiasouthward along the Cascade Mountains to northern Oregon, and on thecoast ranges of Oregon and Washington. According to Sargent, it attains itslargest size on the Olympic Mountains, where it is the most common silver fir,Abies 783extending from 1200 feet up to timber line at about 4500 feet, and forming, with theWestern Hemlock, a large part of the forest between 3000 and 4000 feet. In theCascade Mountains it extends south to about 20 miles north of Crater Lake whereMr. Coville found it on the east side of Diamond Mountain. It occurs 1 in theextreme south-eastern end of Alaska, at the Boca de Quatre inlet, ranging fromsea-level to 1000 feet altitude ; but has not yet been found between this point and thenorthern end of Vancouver Island. It is the common fir - i n south-western VancouverIsland, where it grows abundantly from sea-level up to the summits of the highestmountains. Near the sea it often forms groves of almost pure growth, the treesstanding close together and having very tall slender trunks, about 3 feet in diameterat the base, and often unbranched to a height of 100 feet or more. At an altitudeof 3000 feet it is a comparatively small tree, often clothed with branches to thebase. Plate 220, taken from a photograph, for which I am indebted to Mr. J. M.Macoun of the Geological Survey of Canada, shows the tree as growing nearKamloops, in British Columbia.Sargent says, " unsurpassed among fir trees in the beauty of its snowy bark,dark green lustrous foliage, and great purple cones, Abies amabilis can never beforgotten by those who have seen it in the alpine meadows covered with lilies,dog's-tooth violets, heaths, and other flowers which make the valleys of the northernCascade Mountains the most charming natural gardens of the continent. "Engelmann in a letter, dated " Portland, Or., August 6," 1880, and quoted inGardeners' Chronicle of December 4, 1880, says of it: " A. amabilis, on the samemountain where Douglas discovered it, just south of the Cascades of the Columbia, isa magnificent tree, at about 4000 feet, attaining 150 to 200 feet high with a trunk4 feet in diameter, branching to the ground and forming a perfect cone. The barkof old trees is i^ to 2 inches thick, furrowed and reddish grey, that of younger trees,less than 100 years, is quite thin and smooth, light grey or almost white. It iscertainly very closely allied to A . grandis, but readily distinguished by its verycrowded dark green foliage and its large dark purple cones. It has the purplecones and sharp-pointed leaves (on fertile branches) of A . subalpina, but this latterhas much smaller cones, and not such crowded leaves."Though I saw this tree in abundance on Mount Rainier I cannot say that Iknow how to distinguish it in the forest from A . nobilis without the leaves andcones. It has, according to Plummer, a wider range of elevation than that species,and grows from 800 up to 5500 feet. The cone is as large as that of A . nobilis,but without the projecting bracts. From A . lasiocarpa? with which it was mixedin the upper part of its range, it is distinguished by its habit, which is muchless slender and spiry, by its greater size, and by its cones, which are nearlytwice as large. Plummer says that it attains 200 feet in height by 15 feet in girth,but I saw none so large as this that I could identify. It is a slow-growing tree, one20 inches in diameter having 288 rings.1 U.S. Forest Service, Sylvical Leaflet, 22, p. i (1908). Its most southerly point in the coast range is Saddle Mountain,25 miles south of the mouth of the Columbia River.z Cf. Butters, Conifers of Vancouver Island, in Postelsia, 1 87 (St. Taul, Minn., 1906).3 Sargent, Silva, xii. 126, adnot., mentions the occurrence in a wild state of a hybrid between these two species.

784 The Trees of Great Britain and IrelandAs there were no cones on any of the firs on this side of Mount Rainier in1904 I was unable to procure seed of either of these species, though Prof. Aliensent me both of them in 1905.The wood is yellowish and can, according to Plummer, be distinguished fromthat of A . lasiocarpa, by its darker colour. It is soft and perishable, and of nocommercial importance at the present time.Abies amabilis was discovered in 1825 by Douglas on a high mountain south ofthe Grand Rapids of the Columbia River ; but it was not until 1830 that he succeededin sending to England seed, from which a few plants were raised in the garden ofthe Horticultural Society at Chiswick; and of these original trees hardly anynow survive. For many years afterwards the tree was not seen by any traveller orcollector ; and seeds of reputed A . amabilis sent to Europe invariably turned outto be some other species; and much confusion resulted in the nomenclatureof the western American silver firs. In 1880 the tree was re-discovered bySargent in company with Engelmann and Parry, who found it on Silver Mountainnear Fort Hope on the Fraser River; and a few days later Sargent himself observedit on the mountain where it had first been seen by Douglas. Large supplies of seedwere sent from Oregon in 1882, and young trees are not now uncommon.REMARKABLE TREESOf the original trees, those raised from seed sent home by Douglas, Kent knewonly of two surviving in 1900, one at Dropmore and another at Orton Longueville.The latter, as I was told by Mr. Harding, was cut down in 1905, when it measured5 feet 9 inches in girth.The tree at Dropmore, which was received from the Royal Horticultural Society,and planted in 1835, was cut down four years ago. Mr. Page informs us that thetrunk in the timber yard measured 36 feet long by 8 feet 4 inches in girth at 5 feetfrom the lower end. A cutting from the tree was raised in 1847 by the late Mr.Frost, and is now growing at Dropmore, and measures 50 feet high by 7 feet3 inches. For a time, up to 1873, it promised to be a better tree than its parent;but it is now a miserable object, being badly affected by " knotty " disease. 1 Thisdisease has attacked also all the young trees of this species at Dropmore, somefourteen or fifteen in number, which were planted a few years ago.1 Dr. Masters, in Card. Chron. xvii. 812, xviii. 109, figs. 19, 20 (1882) states that Mr. Barren had proved the goutyswellings on branchlets of A . amabilis and A . nolilis to be due to a woolly aphis, and had succeeded in killing the pest, inhis nursery at Borrowash, by applications of fir-tree oil. A petroleum emulsion is recommended in Card. Chron. xxvii. 190(1900). I am indebted to Prof. Borthwick of Edinburgh for a paper on the subject (in Nat. ZeitscJir. Forst. a. Landwirthschaft,1908, p. 151, figs. 1-4) by Dr. E. VVolz, who states that these swellings are caused by a Chermes which Cholodkovskyhas named C. picea, var. Bouvieri. The life-history of this insect does not seem to have been fully worked out; and itmay not be identical with the Chermes picece, which attacks the bark of silver firs, and is said by Gillanders (Forest Entomology,333) t be common in the nursery on young plants of A. pectinata and of A. Nordmanniana. The figures given byWolz, however, of Abies nobilis, attacked by the disease, represent exactly the swellings which I have seen on that speciesat Carlisle, and which is present on most of the trees of A . amabilis in England. E. R. Bnrdon, in Journal of EconomicBiology, 1 908, ii. 132, states that Cholodkovsky's drawing looks more like the effect attributed to Êciditim elalinum. Cf.Hartig, Diseases of Trees, 1 80, fig. 109 (1894), who states that no formation of spores ever takes place on these swellings.Abies 785We have, however, found several other old trees, none of which are finespecimens, and may have been planted later.A tree at Bayfordbury, with a broken top, is about 20 feet high. AtBrickendon Grange, Herts, there is a remarkable specimen, only a foot in height,with long branches spreading over the ground for about 12 feet. This curiosity isprobably very old; and its peculiar form is possibly due to the leader having beenrepeatedly bitten by animals.At Pencarrow, Cornwall, a tree is growing, which I made in 1905 47 feet highby 7 feet 10 inches in girth. Mr. Bartlett, in a letter dated February, 1906, givesthe following interesting particulars concerning this tree: " According to SirW. Molesworth's catalogue of the trees at Pencarrow, the A bies amabilis was plantedin 1843. The soil is well-drained loam, and the tree stands in a sheltered position.For many years it was a strikingly beautiful specimen, quite symmetrical andfeathered to the ground. A few years ago it was attacked by Chermes, and isnow in a poor state and likely to be completely ruined by the disease in a few years.The tree bore a few cones near the top, four years ago ; but these contained no goodseeds. The cones were resinous, dark blue in colour when growing, fading to a dullbrown towards autumn. The bark of the trunk and branches is covered with resinblisters,which exude a liquid resembling golden syrup in colour and consistency.The buds are late in unfolding." Mr. Bartlett states that there is, at Lamellan, innorth Cornwall, a perfectly healthy but stunted example of Abies amabilis, growingon very poor soil on the edge of a quarry. This tree was probably raised from acutting of the Pencarrow tree. At Menabilly, in the same county, there is anothertree, the flowers of which have been figured. 1 In 1908 it measured 37 feet by3 feet 7 inches.At Brocklesby Park, Lincolnshire, Henry measured in 1908 a tree, 50 feet by4 feet, the date of planting of which is unknown. Though very healthy in generalappearance, some of the lower branches are beginning to suffer from knotty disease.The bark is very smooth and covered with numerous resin blisters, differingmarkedly from the rough bark of an A . Nordmanniana, of the same size, growingbeside it.At Smeaton-Hepburn, in East Lothian, there is a tree, 2 which was planted in1843 ; but its top was blown off in 1859, and it is now only 31 feet high, but hasa girth of 8 feet (o inches. It produced staminate flowers in 1886.At Castle Kennedy, Abies amabilis takes on a low creeping bushy habit, possiblydue to the plants being raised from cuttings, and I saw a similar dwarf stunted plantat Moncreiffe, which I believe to be A . amabilis.On the whole this species appears to be a failure in cultivation, in Europe; anddoes not succeed any better in New England, where, according to Sargent, 3 it hasproved rather tender and grows very slowly. (H. J. E.)1 Card. Chron. iii. 755, f. 102 (1888).2 Cf. Sir Archibald Buchan-Hepburn's account in Prof. Berwickshire Naturalists1 Club, xviii. 207, 210 (1904). It was8 inches high at the time of planting, when it was supposed to be A. grandis.3 Sargent, in The Pinetum at IVellesley in SQOJ, p. 12, mentions a small healthy specimen, which was raised in theVeitcliian nurseries near London, from seeds collected in Oregon by C. S. Pringle in 1882.IV r

786 The Trees of Great Britain and IrelandABIES NOBILIS, N OBLE FIRAbies nobilis, Lindley, Penny Cycl. i. 30 (1833); Masters, Card. Chron. xxiv. 652, f. 146 (1885),and/0«w. Linn. Soc. (Bot.) xxii. 188 (excl. habitat Mt. Shasta, and var. magnified) (1886);Sargent, Silva N. Amer. xii. 133, t. 617 (1898), and Trees N. Amer. 6 5 (1905); Kent,Veitch's Man. Coniferce, 5 21 (1900).Pinus nobilis, Douglas, in Cotnp. Bot. Mag. ii. 147 (1836).Picea nobilis, Loudon, Arb. etFrut. Brit. iv. 2342 (1838).A tree, attaining in America occasionally 250 feet in height with a girth up to 24feet, but more usually 150 to 200 feet high. Bark smooth on young trees, becoming onold trunks reddish-brown and deeply divided by broad flat ridges, irregularly brokenby cross fissures and covered with thick closely appressed scales.Buds concealed by the leaves at the tips of the branchlets, ovoid-globose;terminal bud resinous above and surrounded at the base by a ring of lanceolateacuminate or subulately pointed pubescent brown scales; lateral buds with ovatebasal scales. Young shoots smooth, densely covered with minute rusty browntomentum, which is retained in the second year.Leaves on lateral branches pectinate below, extending outwards in the horizontalplane in two lateral sets; above, the leaves in the middle line, much shorter, completely cover the shoot, from which they arise curving upwards, after being appressedto the branchlet for a short distance near their bases, their tips usually having aslight inclination forwards. Leaves up to about i£ inch long, ^ inch wide, linear,flattened, narrowed at the base, uniform in width elsewhere, rounded and entire atthe apex; upper surface with a continuous median groove and variable as regards thestomata, which are sometimes in two definite bands each of six to eight lines or sometimes present as a few irregular lines, or rarely absent; lower surface with twonarrow bands of stomata, each of five to six lines ; resin-canals marginal.Leaves on cone-bearing branches all upturned, thickened, and with sharpcartilaginous points.Staminate flowers reddish. Pistillate flowers with broad rounded scales, muchshorter than the nearly orbicular bracts, which are erose in margin and contractedabove into slender elongated reflexed tips.Cones cylindrical, but narrowing towards the full and rounded apex; 4 to 5inches long by 2 inches in diameter on wild trees, 6 to 10 inches long by 3 to 4 inchesin diameter on cultivated trees ; pubescent, purplish-brown with green bracts whengrowing, the bracts becoming bright chestnut brown in the mature fruit. Scales:lamina, i£ to i^ inch broad, i inch long, variable in shape; gradually narrowing tothe base with straight lateral margins, or rounded and denticulate on the sides abovethe middle and contracted below ; claw short, clavate. Bracts exserted and stronglyreflexed, covering the greater part of the scale next below ; lamina, broad, full androunded above, fimbriate in margin, and with a conspicuous midrib prolonged into amucro about J inch long ; claw long and cuneate. Seeds pale brown, about J inchAbies 787long, with similarly coloured obovate-cuneate wings, which in cultivated specimensare considerably longer than the body of the seed.This tree can only be confused with A . magnified, which has a differenthabit. The difference between these two trees in the shape and disposition of theleaves is given in the Key, p. 718. (A. H.)DISTRIBUTIONAccording to Sargent, this species forms extensive forests on the CascadeMountains in Washington, extending southwards to the valley of the MackenzieRiver, Oregon. It also occurs on the coast ranges of Washington, and the SiskiyouMountains of California. It is most abundant on the western slopes of the CascadeMountains, and ranges from 2500 to 5000 feet above sea level, attaining its largestsize at 3000 to 4000 feet. It is less abundant and of smaller size on the northernand eastern slopes of these mountains. It commonly attains 200 feet in height; andoften grows to 250 feet; Sheldon says, even to 300 feet.In the Cascade Range Forest Reserve 1 the noble fir forms about 6 per cent ofthe total, and is an important element in the mixed forests of the middle zone on thewestern slope, where it often comprises 15 or 20 per cent of the forest. It crossesthe summit in lat. 45 where a moist climate prevails, but cannot compete withpine and larch in the drier areas. It is closely associated with the lovely fir, andamong lumbermen both species are called larch. Some individuals attain as muchas 8 feet in diameter, but the average size is about 150 feet high by 12 feet in girth atthe base. Langille states 2 that this tree cannot hold its own against the lovely fir(A. amabilis} and hemlock, which are superseding it, and that a sapling is seldom seen.A tree growing at 6000 feet elevation was 163 years old and 125 feet in height, witha diameter of 4 feet 5 inches at the base.In the forests of Mt. Rainier in Washington, Plummer says that the noble firis the finest timber tree and is found from 1800 to 5200 feet. The largest that hemeasured was 225 feet by 18 feet. But when I ascended this mountain fromLongmire's Springs I did not see it, or perhaps I did not distinguish it in theabsence of cones from Abies amabilis. I n the watershed of the Washougal andRock Creek rivers, however, which are very heavily timbered, it forms, according toPlummer, 25 per cent of the timber. The cones here measure about 4^ inches longby 2\ inches wide, not so large as some I have seen in England.I saw this tree at its best in the Cascade Mountains above Bridal Veil innorthern Oregon in June 1904. In this district the tree is known to the lumbermen as larch, and grows in thick forest, more or less mixed with Douglas fir andhemlock ; with Acer circinatum and other shrubs as underwood, where there is lightenough for any to exist. The largest trees I saw here were above 200 feet inheight, and were clear of branches for at least two-thirds of their height, as in theillustration, which was taken from a tree at this place which measured 210 feet by 131 Forest Conditions of the Cascade Range Forest Reserve, U.S. Geological Survey, Washington, 1903.8 Ibid. p. 35.

788 The Trees of Great Britain and Irelandfeet. (Plate 221.) A stump close by showed 360 rings on a diameter of 4 feet, thefirst fifty being twice as wide as any of the later ones. I could find no seedlingsof the noble fir in this part of the forest, and my guide said that he had seen noneexcept at higher elevations.The wood of this tree, though not of equal value to that of Douglas fir, isbeginning to be more appreciated, and I saw it being cut up at the mill at BridalVeil where the owner, Mr. Bradley, told me it was worth twenty to twenty-fivedollars per 1000 feet, and was sent east to be used for the same purposes as whitepine.HISTORYThis tree was discovered by David Douglas on the south side of the Columbiariver in September 1825, and introduced by him five years later on his secondjourney. Ravenscroft, 1 after quoting Douglas's account of the collection of the seeds,which was published in the Companion to the Botanical Magazine, vol. ii. p. 130,says that the seeds arrived in good condition, and were successfully grown anddistributed among the Fellows of the Royal Horticultural Society for whom at thattime Douglas was working. " Extravagant prices were paid for the plants, fifteen andtwenty guineas being then no unusual price." As it usually does, the demand calledforth a supply, but for a long time this supply was in a great measure obtained bymaking grafts and cuttings from the older plants. Plants grown from this source,however, seldom have the same beauty as seedling trees.The next importation was a small package of seed sent by Mr. Peter Banks,who was drowned soon after. After him Jeffrey sent a quantity to the OregonAssociation, but not a plant came up, as the seeds had been destroyed inthe cone by the larva of a hymenopterous insect, Megastigmus pint, and the samething happened to the greater part of the seeds sent by William Murray andBeardsley. Afterwards Lobb and Bridges sent more consignments.Ravenscroft says that plants raised from home-grown seeds are not so strongand healthy as those from imported seed, and have often died from a fungoidattack.CULTIVATIONAmong the silver firs of North America none has had a greater success as anornamental tree than this, but it is only after many years of cultivation that we areable to say with confidence, what are the conditions of soil under which it will preserveits beauty.When first introduced it became so popular that seedlings could not be procured in sufficient quantity to supply the demand, and grafting was resorted to bynurserymen ; the silver fir being usually the stock selected. These trees grew well fora good many years, and some grafted trees are still thriving; but the majority of1 In Lawson, Pinet. Brit. ii. 184.Abies 789them have shown a tendency to produce cones in such quantity and so prematurely,that the trees have ceased to produce a straight leader, and have often becomeunsightly and ragged. This applies specially to those which were planted on lawnsor on pleasure-grounds, without much shelter.An avenue of this tree was planted in 1868 at Madresfield Court, Worcestershire, with grafted trees of the glaucous variety from the Worcester nurseries. Itwas figured in Veitch's Manual of Coniferce, ed. 2, p. 524. Though every care hasbeen taken by top-dressing, and removing the cones to keep these healthy, they donot seem likely to remain so, as the lateral branches are, in many cases, covered withthe knotty swellings described under Abies amabilis, p . 784, note i.Mr. W. E. Gumbleton of Belgrove, near Queenstown, tells me that many yearsago when Abies nobilis was still scarce, the Duke of Leinster, whose treewas one of the first to produce cones, sold the seed of it for ,£40. Thecones were artificially fertilised by shaking out the pollen from the malecatkins at the foot of the tree, and dusting it from a ladder on the female flowersat the top.It is often stated that this is one of the few silver firs which grows well onlimestone, but my own experience disproves this, and I have never seen a really finetree where there was much lime in the soil. A deep sand resting on rock or a hillside, where good drainage is combined with plenty of humus, seem to be the bestconditions for the noble fir ; and if the glaucous variety, of which seedlings are difficultto obtain, is desired, I would graft it on A. Nordmanniana, which is usually a mostvigorous grower, and endures spring frosts better than the common silver fir.In woods the noble fir is often healthier than in the open, and in some caseshas reproduced itself, though not abundantly. I have raised numbers of seedlingsfrom grafted trees, but they were always sickly and died young on my soil, and inany case their growth is slow at first, six to ten years being required to produce treesfit to plant out. But in Scotland seedlings raised from home-grown seed are healthyand vigorous.The tree is quite hardy in all parts of the country, even in the severe climate ofupper Deeside, where at Balmoral it thrives well, and has endured several degreesbelow zero without injury. 1 It enjoys a fairly wet climate, but will also grow wellin the drier parts of England if the soil is deep and cool.REMARKABLE TREESThe largest noble fir that I know of in England is at Tortworth, where, on adeep bed of sand sloping down to the lake, it had attained in 1901 a height of about100 feet and 9 feet 6 inches in girth in forty-seven years from the date of planting.1 A . nobilis, one of the hardiest and best wind-resisting conifers in cultivation, thrives well on gneiss or granite, and maybe planted on the most exposed sites. It is the most prolific of all silvers in seed bearing, and readily reproduces itself.Commercially it may be placed next to A. gi-andis amongst exotic firs. The timber, like all the west North American treesof the genus, is white, soft, and light, but closer in texture than A . grandis. Root formation ruined by frequent transplanting.(J. D. CROZIEE.)

79 The Trees of Great Britain and IrelandThis tree has suffered to some extent from an attack of Chermes, with which thetrunk was covered in 1903, but when I last saw it this had mostly disappeared.Lord Ducie had the tree accurately measured by a man climbing it in May 1908,and informs me that it was then 103 feet 9 inches high, by 9 feet 11 inches in girth.It was planted in 1854, and was 7 feet high in 1855 and 23 feet in 1864.At Highnam Court, Gloucestershire, there is also a fine specimen in thepinetum, measuring 75 feet by 8 feet, but the trees here seem, as they do in manyother places, to have nearly exhausted the soil they grow in, and are beginning togo off. At Miserden Park, the seat of A. Leatham, Esq., in the same county,there is an avenue of grafted trees on dry oolite soil, which were so laden withcones in the year 1900 that they have suffered much in consequence, thoughhitherto they have borne the exposed situation well.At Chatsworth Mr. Robertson has measured a tree 85 feet by 8 feet 5 incheswith a fine clean stem containing 195 cubic feet. At Walcot, the seat of the Earl ofPowis, in Shropshire, I measured in 1906 a very fine glaucous specimen which, thoughgrafted, was 86 feet by 10 feet 9 inches. At Beauport, Sussex, there is a tree, alsografted, 86 feet by 8 feet i inch in 1905. At Linton, Kent, there is a tree 90feet by 8 feet 6 inches in 1902. At Barton there is a tree 80 feet by 7 feet,sheltered in a high wood, and growing fast.In Fulmodestone Wood, on the Earl of Leicester's property, there is a tree74 feet by 9 feet 6 inches, from which a self-sown seedling has sprung up, whichat eleven to twelve years old was, in 1903, 3 feet 6 inches high ; another self-sownseedling in the same place was 20 feet high at about 23 to 25 years old.At Sandringham there are two fine trees in a shrubbery near York House, thelargest of which, in October 1907, measured 85 to 90 feet by 8 feet 10 inches.At Twizell, Northumberland, once the property of Selby, the author of BritishForest Trees, I saw in 1906 a tree 80 feet by 8 feet, the top of which, however, wasdamaged by wind.In Wales it seems to thrive well both at Penrhyn and Hafodunos, in the north ;and at Uinas Mawddwy in Merionethshire, where in 1906 1 measured a veryflourishing tree 75 feet by 5 feet 8 inches.In Scotland 1 it generally succeeds better than in England, and where it hassufficient shelter seems likely to attain a great size and age. By far the finest thatI have seen, are some trees growing at the foot of a sheltered bank on deepsandy soil, in the Dolphin walk at Murthly, four of which certainly exceed 100 feetin height, and the tallest was, as nearly as I could measure it, from 105 feet to 110feet by 7 feet 11 inches in September 1906.A tree growing at Ballindalloch Castle, Banffshire, the seat of Sir J. Macpherson-Grant, is said to be the finest in the north of Scotland, and is stated to have measuredin August 1907, 94 feet by 9 feet i \\ inches, and to be only forty-seven yearsplanted. 1The next largest we have seen is at Keir,2 Perthshire, which was, in 1905, 991 Trans. Roy. Scot. Arb. Sac. xxi. 98 (1908).* This tree was reported to be forty years old in 1891, and then measured 82 feet by 5 feet 8 incbes (Jotim. Key. Hort.Soc. xiv. 531 (1892)).Abies 79feet high by 7 feet 5 inches in girth, remarkable for its clean stem and short branchesoccurring only on the upper half of the tree. Another tree at Keillour, in the samecounty, was 91 feet by 7 feet i inch in 1904; and at Castle Kennedy, Wigtonshire,another measured in the same year 80 feet by 7 feet 10 inches. Sir ArchibaldBuchan-Hepburn reports one at Smeaton-Hepburn, East Lothian, which measured,in 1908, 84 feet by 8 feet 10 inches. It was planted in 1843.At Blair Castle, a tree planted by the Duke of Atholl about forty-two years agowas, in 1904, 70 feet by 5 feet; and at Balmoral, though of no great size, it seems tobe the best of the silver firs, and has endured a temperature of 15 without injury.In Ireland, Abies nobilis thrives well. At Churchill, Armagh, there is amagnificent specimen, which in 1904 was covered with cones, and measured 73 feetby 8 feet 4 inches. At Curraghmore, Co. Waterford, a tree measured, in 1907, 75feet in height by 10 feet in girth. At Powerscourt, a tree in 1903 was 59 feet by6^ feet. At Carton, in the same year, a tree measured 61 feet by 6 feet. At BirrCastle, King's County, there is a very tall tree, which was reported 1 in 1891 to be83 feet high and 6 feet in girth. There is an avenue of this species at Woodstock,Kilkenny; and good specimens are growing at Castlewellan in Down. In aplantation behind the old deer park at Castle Martyr, Co. Cork, there is a very largetree of the glaucous variety, which, though I could not measure the height accurately,seems to be about 75 feet high and is 10 feet 4 inches in girth. (H. J. E.)1 Journ. Roy. ffoit. Soc. xiv. 557 (1892).

The Trees of Great Britain and Ireland Abies 793Var. shastensis, Lemmon, West. Amer. Cone-bearers, 6 2 (1895); Sargent, SilvaN. Amer. xii. 138, t. 620 (1898), and Trees N. Amer. 6 7 (1905).ABIES MAGNIFICA, RED FIR, SHASTA FIRAbies magnified, A. Murray, Proc. R. Hort. Soc. iii. 318, ff. 25-33 ( l86 3) > Masters, Card. Chron.xxiv. 652, f. 148 (1885); Sargent, Silva N. Amer. xii. 137, tt. 618, 619(1898), and TreesN. Amer. 66 (1905); Kent, Veitch's Man. Conifera, 5 16 (1900).Abies nobilis, Lindley, var. magnified, Kellogg, Trees of California, 2 8 (1882); Masters, Journ. Linn.Soc. (£ot.) xxii. 189, t. 5, ff. 19-21 (1886).Abies shastensis, Lemrnon, Garden and Forest, x. 184 (1897).Picea magnifica, Gordon, Pinetum, 2 19 (1875).Pinus magnified, M'Nab, Proc. JR. I. Acad. ii. 700 (1876).Pinus amabilis, Parlatore, in DC. Prod. xvi. 2, p. 426 (in part) (1868).A tree, attaining in America 200 feet in height and 30 feet in girth. Bark,buds, and branchlets similar in all respects to those of A foes nobilis.Leaves on lateral branchlets arranged as in A . nobilis; but with the medianleaves above not so densely crowded as in that species, portions of the branchletbeing visible from above, whereas in A . nobilis the branchlet is completely concealed ;moreover, these median leaves are appressed to the branchlet at their bases for ashorter distance than in the other species. Leaves longer than in A . nobilis, up toabout if inch long, ^ inch wide, tapering gradually to the base, uniform in widthelsewhere ; apex rounded, entire; obscurely quadrangular in section ; upper surfacewith a central ridge and several (often eight) rows of stomata; lower surface withtwo bands of stomata, each of four to six lines ; resin-canals marginal. Leaves onleading shoots erect and acuminate, with long rigid points pressed against the stem.Leaves on fertile branches much thickened, crowded, upturned, acute with shortcallous tips.Staminate flowers dark reddish. Pistillate flowers with rounded scales muchshorter than their oblong pale green bracts, which end in elongated slender tips.Cones very large, 6 to 9 inches long, 3 to 5 inches in diameter, cylindrical, butslightly narrowing to the rounded, truncate or retuse apex; purplish-violet whengrowing, brown when mature, pubescent. Scales fan-shaped; lamina, i^ to iînch broad, i inch long, upper margin rounded and incurved, the sides graduallynarrowing to a cordate base ; claw nearly ^ inch long, narrowly obcuneate. Bracts,in the usual form of the species, about two-thirds as long as the scale and notexserted; variable in shape; upper expanded part oval, acute or acuminate,terminated by a mucro ; claw sharply contracted below the lamina. Seeds brownish,more than |- inch long, slightly shorter than their pink obovate-cuneate wings.Var. xanthocarpa, Lemrnon, Third Report, ex Masters, Journ. Roy. Hort. Soc. xiv. 193 (1892),and Card. Chron. xii. 114, figs. 51, 52, 53 (1907).Abies shastensis, Lemmon, Garden and Forest, x. 184 (1897); Coville, Garden and Forest, x.516 (1897).Abies nobilis robusta, Masters, Card. Chron. xxiv. 652, f. 147 (1885) (not Carriere).Abies nobilis, Lindley, var. magnifica, Masters, Journ. Linn. Soc. (Zto/.) xxii. 193, PI. 5 (1886).This differs from the type only in the cones, which have much longer bracts,yellow in colour, rounded or obtusely pointed (not acute), exserted, usually reflexed,and covering about half the outer surface of the scales.This variety, which is known as the Shasta Fir, occurs on the mountains ofsouthern Oregon, in the cross and coast ranges of northern California and on thesouthern Sierra Nevada. In Oregon it is met with in the lower parts of themountains; but in the other localities it only occurs at very high elevations.It is rare in cultivation in England, or at any rate has been rarelynoticed. A tree at the Cranston Nursery, near Hereford, produced cones 1 of thiskind in 1878, which were figured 2 by Dr. Masters. Another is growing at DurrisCastle, Aberdeenshire, where Mr. Crozier states that intermediate forms betweenthis and A. nobilis exist.IDENTIFICATIONThis species is only liable to be confused with A . nobilis; but in large trees, asseen in cultivation, the difference in habit between the two species is remarkable.The formal arrangement of the branches in A . magnijica, though difficult to describe,when once seen can seldom be mistaken. The differences in the foliage are given inthe Key, p. 718. (A. H.)DISTRIBUTIONThe most northerly point at which this tree has been found is on the mountainseast of Odell Lake in about lat. 44 N. in southern Oregon, where Dr. Covillecollected it in 1897, many miles south of where A . nobilis occurs ; and it is not mentioned among the trees of the Cascade Forest Reserve, so that it really belongs tothe Californian rather than to the North Pacific flora. It becomes common on theTrinity Mountains, and on Mt. Shasta is the only fir besides A . Lo-wiana. Thetree extends along the entire length of the western slope of the Sierra Nevada,from 6000 to 9000 feet above the sea, and extends to the eastern slope at highelevations.The northern form has been separated by Lemmon under the name ofA. shastensis, on account of the bracts which protrude from the scales; being in thisrespect, as in its geographical distribution, midway between A . nobilis and A . magnifica;1 Card. Chron. 1 878, p. 343.2 Journ. Linn. Sec. (Bol). xxii. 193, plate v. (1886).IV MI

794 The Trees of Great Britain and Irelandbut this character is variable, and I judged from the specimens shown me by MissA. Eastwood, that the two forms cannot always be defined.On the west slopes of Mount Shasta the tree occurs higher up than A . Lo-wiana,mixing with it at about 6000 feet, and at 8000 feet it is the only species of fir. It isnot on Mount Shasta a very large tree, the biggest that I measured being not muchover 100 feet in height, and 15^ feet in girth, the average being 80 to 100 feet high,by 6 to 8 feet in girth. I could not see very much difference in the bark; thoughA. magnified, is known as the red, and A . Lowiana as the white fir, but the verymuch larger cones of the former distinguish it at once. These were borneonly near the summit of the trees, and could only be procured by shooting themoff with a rifle, or by felling trees on purpose. They were fully formed but unripe in the first week of September. The soil here was very rocky, and drierthan that of any mountain which I have ascended in this latitude; and there waslittle herbaceous vegetation, though the snow is said to lie deep from Novemberuntil May or June.INTRODUCTIONThis species was introduced in 1851 by John Jeffrey, who believed it to beA. amabilis; and the seedlings were distributed under this name amongst themembers of the Scottish Oregon Association. The tree in Scotland is frequentlylabelled A . amabilis, in consequence of this error.W. Lobb 1 sent seed in 1852, also under the name of A . amabilis; but later onthe plants were found to differ from that species, and were distributed by Messrs.Veitch as A . nobilis robusta.REMARKABLE TREESThough it is quite possible that larger trees exist, which have been mistakenfor A. nobilis, yet we have identified none in England which at all approach thatspecies in size, and all the best we have seen are in the eastern and southern counties.The largest perhaps is one at Fulmodestone, Norfolk, a handsome and well-shapedtree growing in a damp soil and well sheltered situation, which in 1905 was 61 feetby 5 feet 9 inches, and bore cones near the summit.At Bayfordbury, on a much drier soil, it has flourished better than A . nobilis, andin 1905 was 56 feet by 5 feet 9 inches (Plate 222). Mr. H. Clinton Baker recentlymeasured a good specimen, 61 feet by 4 feet 4 inches, at Flitwick Manor; andanother, 60 feet by 4 feet 9 inches at High Leigh, near Hoddesdon. At Grayswood,Haslemere, a tree planted as recently as 1881, measured in 1906 56 feet by 4 feetii inches; and at Petworth, in 1905, there was a slender and less vigorous tree 47feet by 3 feet i inch.At Eridge Park, Kent, a tree planted in 1880 by Count Gleichen was, in 1905,34 feet by 3^ feet.In a pinetum close to Presteign, Radnorshire, planted about fifty years ago,Abies 795now the property of Mr. J. H. Wall, I saw a good specimen of Abies magnifica in1906, which measured 53 feet by 5 feet 7 inches.The largest reported 1 at the Conifer Conference in 1891 was at Revesby Abbey,Lincolnshire, and then measured only 40 feet by 5 feet.In Scotland it is more numerous and larger. The late Malcolm Dunn, who hadan exceptionally wide experience in the cultivation of conifers in Great Britain,wrote of it as follows in a paper 2 which he sent to the Conifer Conference:" It is in truth a stately tree and one of the handsomest of all the taller-growingconifers for ornamental purposes. It is one of the very hardiest of the firs, and isseldom affected by spring frost, and the timber being straight, clean-grained, and ofgood quality, it will no doubt be a useful forest tree." But this latter opinion hasnot so far received any proof so far as we know, for the tree is, and seems likely toremain, difficult to obtain, and like most of its congeners is slow and costly to raisefrom seed.Probably the finest trees in Scotland are one at Durris,8 Aberdeenshire, whichwas, in 1904, 80 feet high by 6 feet 6 inches in girth, and when I measured it in 1907had increased to about 85 feet; and another (Plate 223) at Bonskeid, near Pitlochry,of which Mr. J. Forgan has been good enough to send me a photograph, and whichmeasured, in 1908, 87 feet by 8 feet. When he first knew it thirty-five years ago itwas about 12 feet high ; it has not produced cones. Mr. Bean 4 noticed in 1906 atree at Abercairney 70 feet high, and another at Blair Castle 60 feet high.At Farthingbank, Drumlanrig, there is, growing on clay loam at 650 feet abovesea-level, a tree 50 feet by 5 feet 3 inches in 1905, which was planted, accordingto Mr. Menzies, the forester, thirty-one years previously.The tree is rare in Ireland, but there is a specimen 8 at Castlewellan, which was47 feet by 6 feet in 1906; and at Powerscourt, a tree, planted thirty-five years ago,was 57 feet by 6 feet 8 inches in 1906, and is said to bear cones nearly every year.(H. J. E.)1 Journ. Roy. Hort. See. xiv. 568 (1892). * Journ. Roy. Hort. Soc. xiv-. 83 (1892).3 A . magnified closely resembles A. nobilis, but in strong contrast as regards seed-bearing. It does not seem as if thetree is likely to become acclimatised in this respect as, although planted in considerable numbers throughout the policygrounds and plantations, and most of those trees now between fifty and sixty years of age, cones have been produced only onone occasion, and that on only a few trees. The timber when closely grown is closer in texture, richer in colour, and betterin quality than A. nobilis. Like that species it is impatient of side shade and sheds its branches freely. Constitutionally itis less robust than its near relative, and also less accommodating in its demands on site and soil. (J. D. CROZ1ER.)4 Kew Bulletin, 1 906, pp. 264, 267.6 Figured in Garden, June 28, 1890, p. 591.1 Hortus Veitchii, 336 (1906).

796 The Trees of Great Britain and Ireland Abies 797ABIES BRACTEATA, BRISTLE-CONE FIRAbies bracteata, Nuttall, Sytoa N. Amer. iii. 137, t. 118 (1849); Hooker, Bot. Mag. t. 4740(1853)5 Masters, Card. Citron. \. 2 42, f. 44 (1889),and vii. 672, f. 112 (1890); Kent, Veitch'sMan. Coniferce, 493 (1900).Abies venusta, Koch, Dendrol. ii. 210 (1873); Sargent, Silva JV. Amer. xii. 129, tt. 615, 616(1898), and Trees N. Amer. 63 (1905)-Pinus venusta, Douglas, Comp. Bot. Mag. ii. 152 (1836).Pinus bracteata, Don, Trans. Linn. See. xvii. 442 (1837).Picea bracteata, Loudon, Arb. et Fritt. Brit. iv. 2348 (1838); Coleman, Garden, xxxv. 1 2, with fig.(1889).A tree attaining in America 150 feet in height and 9 feet in girth. Barkbrown, smooth; becoming, near the base in old trees, slightly fissured and brokeninto thick appressed scales. Buds unique in the genus, elongated, fusiform,broadest near the base, and gradually tapering to a sharp point, about \ to f inchlong, brown in colour, non-resinous; scales thin, membranous, glabrous, looselyimbricated, obtuse at the apex, shorter at the base of the bud, gradually lengtheningabove. Young shoots glabrous, greenish, with slightly raised pulvini and inconspicuous furrows. Base of the shoots usually ringed with the scars of theprevious season's bud-scales, which in most cases all fall off and do not persist inart, as is usual in other species.Leaves on lateral branches pectinately arranged, those below spreadingoutwards in two sets in the horizontal plane; those above slightly shorter, falcate,directed outwards and slightly upwards and forwards, forming a shallow V-shapeddepression on the upper side of the branchlet. Leaves, up to 2 inches long, £$ inchwide, rigid, thin, flat, linear, ending in long spine-like cartilaginous points, neverbifid; widest in the lower third, gradually tapering to the apex, and abruptlynarrowed close to the base; upper surface dark-green, shining, slightly concave inthe lower half and flat near the apex, no definite median groove being formed ; lowersurface with two wide white bands of stomata, each of 10 to 12 lines; resin-canalsmarginal. Leaves on cone-bearing branches upturned, falcate.Male flowers, i£ to i^ inch long, cylindric, shortly-stalked, surrounded at thebase by numerous lanceolate, fawn-coloured parchment-like scales, similar tothose of the leaf-buds. Pistillate flowers, with oblong scales rounded above andnearly as long as the cuneate obcordate yellow-green bracts, which end in slenderelongated awns.Cones, remarkable for the long spiny rigid tips to the bracts, ovoid, roundedand full at the apex, 3 to 4 inches long, about 2 inches in diameter, glabrous, 1resinous, purplish brown. Scales, about i inch broad by ^ inch long, almostreniform; upper margin incurved, with a short obtuse denticulate cusp; clawobcuneate. Bracts oblong-obovate, adnate to the scale to beyond the middle andRemarkable, as all the other species of Abies have the scales of the cones pubescent.deciduous with it, terminating in linear, rigid spines, i to 2 inches long, which in theupper half of the cone point towards its apex, and in the lower half are spreadingand often recurved. Seeds dark reddish-brown, about g inch long and nearly aslong as their pale reddish-brown shining wings. (A. H.)DISTRIBUTION AND HISTORYAbies bracteata has perhaps the most restricted distribution of all the silverfirs, as, according to Sargent, it only occurs in a few isolated groves along themoist bottoms of canons at about 3000 feet elevation on both slopes of the westernridge of the Santa Lucia Mountains in Monterey County, California. The mostnortherly point where it is now known to grow is in Bear Canon, twenty-five milessouth of the Los Burros mines; the other localities mentioned by Sargent are in theSan Miguel Canon and in a gorge at the head of the Nacimiento river.The discovery of this tree is assigned by Don, Sir VV. J. Hooker, 1 andSargent to Dr. T. Coulter, who, according to a letter z of Douglas to Hooker datedNovember 23, 1831, arrived at Monterey after he began the letter in question.Douglas also, in a letter 8 dated October 1832, states 4 that he found the tree, whichhe called Pinus venusta, in the preceding March " on the high mountains of California," and that it is never seen at a lower elevation than 6000 feet above sea-level,in lat. 36 , where it is not uncommon.But Kent says, 5 in a note, that a comparison of the dates shows that Douglaswas the first discoverer, which, however, is not proved ; as, according to Douglas'sown showing, Coulter was at Monterey, near to the place where the tree grows, threemonths before Douglas found the tree himself. Prof. Hansen 6 also has incorrectlystated the date of Douglas's discovery of this tree as March 1831 instead of March1832.William Lobb, when collecting for Messrs. Veitch in 1853, introduced it tocultivation, and in a letter in Gardeners' Chronicle, 1 853, p. 435, describes it as "themost conspicuous ornament of the arborescent vegetation. On the westernslopes, towards the sea, it occupies the deep ravines, and attains the height offrom 120 to 150 feet, and from i to 2 feet in diameter, the trunk as straightas an arrow, the lower branches decumbent. The branches above are numerous,short, and thickly set, forming a long tapering pyramid or spire, which gives tothe tree that peculiar appearance not seen in any other kind of the Pinus tribe.Along the summit of the central ridges, and about the highest peaks, in the mostexposed and coldest places imaginable, where no other pine makes its appearance,it stands the severity of the climate without the slightest perceptible injury, growingin slaty rubbish, which to all appearance is incapable of supporting vegetation. Insuch situations it becomes stunted and bushy. The cones are quite as singular asthe growth of the tree is beautiful; when fully developed the scales, as well asi Bot. Mag. t. 4740(1853)-3 Ibid. 1 51.6 Veitch's Man. Coniferce, 497, note (1900).* Comp. Bot. Mag. ii. p. 149.4 Ibid. 1 52." Journ. Roy. Hort. Soc. xiv. 459 (1892).

798 The Trees of Great Britain and Irelandthe long leaf-like bracts, are covered with globules of thin transparent resin.Douglas was mistaken in saying that this tree does not occur below 6000 feetelevation ; on the contrary, it is found as low as 3000 feet, where it meets Taxodiumsempervirens."In 1856 another expedition to collect seeds was made by W. Beardsley, whogives a good account of his journey, which is quoted from by Murray. 1 In themiddle of October the seeds were already shed, and Murray says that Mr. W.Peebles, who went for the same purpose on September 17, 1858, found the cones soripe that when the tree was felled they fell to pieces.According to Beardsley, the soil on which it grows is " exclusively thecalcareous districts, abounding with ledges of white, veined, and grey marble."CULTIVATIONA. bracteata has never been a common tree in English gardens and, owing tothe difficulty of procuring seeds in California, it is rarely to be had from nurseries.It seems to be quite hardy as regards winter cold, but susceptible to spring frosts;and all the good specimens I have seen are in sheltered and rather elevated situationson well drained soil.The seedlings which I have raised from English-grown seeds have not thrivenon my soil, though the tree does not appear to dislike a moderate amount of lime.All the best specimens we know of are in the south and west of England, and inIreland. A list of them is given by Kent, 2 and they all are probably of about thesame age, being raised from William Lobb's seeds by Messrs. Veitch in 1854.REMARKABLE TREESThe finest trees in England are in the valley of the Severn, the largest beingat Eastnor Castle (Plate 224), where two are growing. They were stated 8 by thelate Mr. Coleman to have been planted in 1865, and the best of them was 40 feethigh in 1889. It first bore cones in 1888; when I measured it last in 1908, I foundit to be 78 feet by 9 feet, and though very healthy and handsome in appearance, thetop had become forked. It bore cones freely in 1900, from which I raised numerousseedlings, but these have grown very slowly, and do not seem able to make rootsfreely on my soil.At Highnam Court, Gloucester, the seat of Sir Hubert Parry, there is anotherfine tree, difficult to measure on account of its situation, but, in 1908, I made it64 feet by 6 feet 5 inches. It has several times produced cones, four being bornein 1907, from which seedlings were raised.At Tortworth Court there is a tree which Lord Ducie believes to have beenplanted between 1858 and 1862, and in 1908 was 63 feet by 6 feet. It is growing onold red sandstone, about 250 feet above sea level in a situation much exposed tothe south-west wind.Abies 799At Nevill Court, near Tunbridge Wells, I measured a tree which, though only48 by 4^ feet in 1906, is one of the best shaped I have seen, with a very slenderspire, as described by Lobb in California.At Fonthill Abbey there is a tree about 72 feet by 5^ feet, in a sheltered thoughelevated situation on greensand. At Osborne, in the Isle of Wight, a tree was 60feet by 5 feet 9 inches in 1908, but this does not appear to be thriving, on accountperhaps, of the dry soil.At Ponfield, Hertford, the seat of P. Bosanquet, Esq., Henry saw in 1906 atree, very thriving and about 25 feet in height; and in the same district, at HighCanons, near Shenley, Mr. Clinton Baker showed me a tree 53 feet by 4 feet whichbore about twenty cones in 1907. At Pampisford, Cambridgeshire, there is atree, about 25 feet high, growing in a sheltered position, and very thriving.At Monk Coniston, in Westmoreland, the seat of Victor Marshall, Esq., I haveseen a tree which has borne cones, and which measured, in 1906, 60 feet by 5 feet.Several others are mentioned by Kent: at Kenfield Hall, 1 near Canterbury ; atNew Court, and at Streatham, near Exeter; at Upcott, near Barnstaple; and atWarnham Court, near Horsham. A large tree at Orton Hall, Peterborough, was,before it was cut down in 1905, 59 by 6 feet, but became unhealthy owing to the soilbeing too heavy.In Wales, where the species should grow well, I have seen no trees of any size.In Scotland the only specimens I have seen are at Castle Kennedy and atCawdor Castle, neither of which are large, and the climate of Scotland generallyseems to be too cold for it.2In Ireland Henry has seen specimens at Fota, in the south-west, a fine youngtree which, in 1903, was 48 feet by 4 feet; at Castlewellan, in the north-east, another,in 1906, was 35 feet by 3 feet 2 inches; and a smaller one also exists at Glasnevin.On the continent of Europe this tree is very rare, the only fine one I haveseen being a tree at Pallanza in the nursery grounds of Messrs. Rovelli, which, in1906, was about 70 feet by 7 feet, but not very healthy and bearing no cones.M. Parde states that there is a fine specimen in the domain of the NationalSociety of Agriculture at Harcourt (Eure); and I saw a small one in M. Allard'scollection at Angers. (H. J. E.)1 This tree produced cones in 1886. Cf. Card. Chron. xxvi. 85 (1886).2 The one specimen now remaining at Durris between forty and fifty years of age if it can possibly be taken as afair example of the growth of the tree in this locality, proves it of little use for planting. It is quite healthy, but its growthis slow in proportion to that of A. pectinata. (J. D. CROZIER.)1 Edin. New Phil. Jotirn. x. I, pis. I and 2 (1859).3 Garden, 1 889, xxxv. 12.2 Veitch's Man. Conifertf, lof. cit.

8oo The Trees of Great Britain and IrelandABIES LASIOCARPA, ROCKY MOUNTAIN FIRAbies lasioearpa, Nuttall, Sylva, iii. 138 (i»49); Masters, Card. Chron. v. 172, ff. 23-27, 32,(1889), and/0« . Bot. xxvii. 129 (1889); Sargent, Silva N. Amer. xii. 113, t. 611 (1898),and Trees IV. Amer. 6 1 (1905); Kent, Veitch's Man. Conifera, 5 15 (1900).Abies bifolia, Murray, Proc. Roy. Hort. Soc. iii. 320 (1863).Abies subalpina, Engelmann, Am. Nat. x. 555 (1876).Abies arizonica, Merriam, Proc. Biol. Soc. Wash. x. 115, ff. 24, 25 (1896).Pinus lasioearpa, W. J. Hooker, fl. Bar. Am. ii. 163 (1839).Picea bifolia, Murray, Card. Chron. iii. 106 (1875).Picea lasioearpa, Murray, Card. Chron. iv. 135 (1875).A tree, attaining occasionally 175 feet in height, with a trunk 15 feet in girth,but usually not over 80 to 100 feet high. Bark of young trees smooth and silverygrey; of old trees shallowly fissured and roughened by reddish brown or whitishscales; in some trees becoming corky and white in colour. Buds small, about £ inchlong, ovoid-conical, obtuse at the apex, brownish, resinous; scales embedded in theresin but roughening the surface of the bud by their raised tips. Branchlets swollenat the nodes, those of the first year ashy grey, smooth, and covered with amoderately dense short wavy pubescence. Branchlets of the second year retainingsome pubescence, darker grey, smooth, with the bark slightly fissuring.Leaves on lateral branchlets irregularly arranged; sometimes irregularlypectinate with some of the leaves above and below not directed outwards, butforwards at an angle with the axis of the shoot; usually with most of the leavesdirected upwards, those in the middle line above covering the shoot and standingedgeways with their apices almost vertical, a few leaves in the middle line belowpointing forwards and downwards. Leaves linear, up to i£ inch long by T^ inchbroad, uniform in width except at the gradually tapering base; apex rounded andeither entire or with a slight emargination ; upper surface with a shallow continuousmedian groove, and with four to five lines of stomata on each side of the groove inits anterior half, the lines fewer in number and broken in the basal half; undersurface with two bands of stomata, each of six to eight lines; resin-canals median.The stomatic lines above give the foliage a glaucous appearance; the bands belowvary very much in whiteness. Leaves on leading shoots closely appressed to thestem with their tips directed forwards, flattened in section, and ending in longslender rigid points. Leaves on cone-bearing branchlets upturned, directedforwards, usually acute and not more than ^ inch long.Cones sub-sessile, cylindrical; rounded, truncate or depressed at the slightlynarrowed apex; 2 to 4 inches long by i^ inch in diameter, dark purple andtomentose, with the bracts concealed. 1 Scales very variable in size and shape,from f inch long by f inch wide to ^ inch long by i inch wide: lateral marginsrounded or with sinuses, usually auricled on each side of the short obcuneateclaw. Bract situated at the base of the scale or slightly above it, quadrangular or1 According to Piper, Contrib. U.S. Nat. Herb. xi. 93 (1906), cones on trees growing in the Olympic Mountainshave exserted bracts.Abies 801orbicular, denticulate, emarginate with a long slender mucro. Seed ^ inch long, withdark purplish shining wings, which vary in length according to the height of thescale which they cover almost completely.Var. arizonica, Lemmon, Bull. Sierra Club, i i. 167 (1897) ; Masters, Card.Chron. xxix. 86, 134, ff. 52, 53 (1901).Abies arizonica, Merriam, Proc. Biol. Soc. Wash. x. 115, ff. 24, 25 (1896); Purpus, Gartenwelt,v. 4, 26 (1896).This form occurs in the San Francisco mountains in Arizona, where it iscommon between 8500 and 9500 feet elevation, and occasionally ascends to 12,000feet. It is remarkable for the creamy- white thick corky bark of the trunk. As seenin cultivation, young plants differ from the type, in the leaves being emarginate atthe apex, whiter beneath, and more regularly pectinate in arrangement. Sargent 1states that bark equally corky occurs in trees of Abies lasioearpa in other regions,as in Colorado, Oregon, South Alberta, and British Columbia ; and, as there is nodifference in the cones, he does not assign even varietal rank to the Arizona tree.The best account of this variety is by Prof. Purpus in Mitt. D. D. Ges., No. 13,p. 47 (1904), who visited the San Francisco mountains in 1901, and introducedthe tree to Europe. It seems to be a strictly alpine tree, growing on basaltic andtrachytic rocks, where the soil is never quite dry, either scattered or mixed withPopulus tremuloides, Pinus flexilis, Pseudotsuga Douglasii, and Picea Engelmanni.It attains a height of 60 to 70 feet with a girth of 6 to 9 feet. The bark isvery corky and corrugated, in old trees milk-white or silver-grey in colour. It isreplaced in these mountains at 7000 to 8000 feet by Abies concolor.This form has only recently been introduced into cultivation. Plants were forsale in the Pinehurst Nurseries, North Carolina, in 1901 ; and Dr. Masters saw astock of young plants in Moser's nursery at Versailles in 1903. It is too soon yetto form any opinion as to the suitability of this variety for ornamental gardening.IDENTIFICATIONAbies lasioearpa is perhaps most readily distinguished by the conspicuous bandsof stomata on the upper surface of the leaf, which separate it clearly from the otherspecies 2 with median resin-canals and long narrow leaves. The following points arealso noteworthy : the irregular arrangement of the leaves, which are usually quiteentire at the apex ; the ashy-grey pubescent shoots ; and the resinous obtuse buds.(A. H.)DISTRIBUTIONThis is essentially the alpine fir of the Rocky Mountains and higher ranges onthe west coast of North America, and is the most widely distributed fir of the NewWorld, occurring from about lat. 61 N. in Alaska to Arizona and New Mexico. Itdoes not occur in California. 8 In the west it extends to the summits of the Olympicl Silva, xii. 113.IV2 As A. sibirica and A. sachalinensis, which it somewhat resembles in general appearance.3 U.S. Forest Service, Syhiical Leaflet I , Alpine Fir.N

802 The Trees of Great Britain and IrelandMountains in Washington, and in the east to the mountains of Idaho, Montana(Plate 225), Wyoming, Colorado, and Utah. Everywhere it grows up to or verynear the timber line, and on the shores of Lake Bennett in northern BritishColumbia descends to 2500 feet. In Colorado it reaches 10,000 feet.Macoun states that it crosses the Rocky Mountains into the Peace River region,and the country between the Little Slave Lake and the Athabasca River; and thatin the Rocky Mountains of Alberta it occurs with Picea Engelmanni, but is lesscommon; and in a letter states that it is an enormous tree at Glacier, but becomesdwarfed at higher elevations, ascending to 7000 feet in that region.Wilcox J writes of it as follows :" The balsam fir has about the same range as the white spruce (PiceaEngelmanni) but is less common. At a distance it is hardly to be distinguishedfrom the spruce, but the bark on branches and young trees is raised in blisters whichcontain a drop or two of balsam. This balsam exudes from the bark wherever it isbruised. At first it is a very clear liquid, regarded by old trappers and woodsmenas a certain cure, when brewed with hot water, for colds and throat troubles. Onexposure to the air it hardens into a brittle resin, which the woodsman melts intopitch to seal boxes or mend leaky canvas. The camper-out makes his bed frombalsam boughs, as they are more springy and less rigid than those of the spruce."I saw this tree in perfection in the Paradise valley on the south-west slopes ofMount Rainier in August 1904. An excellent illustration of this locality is given byC. O. Piper in Garden and Forest, vol. iv. p. 382, which shows the tall slender spiryhabit of the fir. Here it lives in company with A . amabilis in the lower part ofits range, and with Tsuga Pattoniana, and Cupressus nootkatensis higher up;growing in small clumps and groves, as shown in the illustration referred to. Itseems to be a very slow grower, a tree felled by Plummer being only 15 inchesin diameter at 125 years old. The tallest that I measured here was 77 feet by5 feet 8 inches, but Sargent says that it occasionally attains 175 feet in height(probably in the Olympic Mountains). The seedlings, which I usually found growingon rotten logs, were very slow in growth, and must be often eight to ten years oldbefore their roots reach the soil.HISTORY AND CULTIVATIONAbies lasiocarpa was discovered by Douglas in 1832, and his specimen, whichis the type of Pinus lasiocarpa of W. J. Hooker, the first name applied to thespecies, is preserved in the herbarium at Kew.Seeds were first collected about 1863, by Dr. Parry in Colorado; but it is notknown if any plants raised from these still survive. The first plants raised in theArnold Arboretum date from 1873, tne largest of them being now only 10 to 12 feetin height. Roezl collected seeds in 1874 in Colorado. 2 According to Syme, 8 a small1 The Rockies of Canada, 62 (1900).2 Masters, Journ. Bot. xxvii. 135 (1889), refers these seeds doubtfully to New Mexico; but there is no doubt that theywere collected in Colorado. Cf. La valise's article on JVouvcanx Coniftrcs du Colorado et de la Californie, in Journ. Soc. Cent.Hart. France, 1 875. 3 Card. Chron. iii. 586 (1888).Abies 803plant of this origin was alive in Perthshire in 1888 ; but it was only 2\ feet in height,forming a wide spreading bush, though it was growing in rich black loam.No trees of this species are recorded by Kent; nor were any specimens sent tothe Conifer Conference in 1891. It appears to be unsuitable for cultivation in thiscountry. Young trees at Kew, a few feet in height, are stunted and dying.Waterer had a large stock of plants in 1889 in the nursery at Bagshot; but theyall did badly and were thrown away, only one or two surviving and showingthe same wretched appearance as the young trees at Kew. Plants cultivatedsome years ago at Glasnevin have since died. Henry, however, lately saw in thePinetum at Hatfield, Herts, a tree, planted in 1893 when it was about 3 feet high,which is now 20 feet in height and 15 inches in girth. It has thriven well hitherto,but is slightly attacked by knotty disease. The best specimen we have seen is oneat Bayfordbury, about 14 feet high, and fairly thriving. A small tree atOchtertyre bore cones in 1906. The tree appears to succeed better in Germany.I have raised seedlings from cones sent by Prof. Alien in 1905 from MountRainier. (H. J. E.)ABIES BALSAMEA, BALSAM FIRAlnes bahamea, Miller, Diet. No. 3 (1768); Sargent, Silva N. Amer. xii. 107, t. 610 (1898), andTrees N. Amer. 5 8 (1905)5 Masters, Card. Chron. xvii. 422, figs. 57-60 (1895); Kent,Veitch's Man. Coniferee, 492 (1900).Abies balsamifera, Michaux, Fl. Bor. Am. ii. 207 (1803) (in part).Pinus bahamea, Linnaeus, Sp. PI. 1 002 (1753).Picea bahamea, Loudon, Arb. et Frut. Brit. iv. 2339 (1838).A tree, commonly 50 to 60 feet in height and 3 to 5 feet in girth, but sometimeslarger, with spreading branches, usually forming an open broad-based pyramid. Bark,greyish brown and with numerous blisters; on old trees broken on the surface intosmall scaly plates. Buds small, globose or occasionally dome-shaped, reddish, shiningand resinous. Young shoots smooth, ashy grey, with very short scattered pubescence ; on the shoots of the second year some of the pubescence is retained, andthe bark fissures slightly between the pulvini. The branchlets when cut havea very resinous odour.Leaves on lateral branches pectinately arranged, in two sets directed outwardsin the horizontal plane ; upper leaves of each set shorter than the others, and directedalso slightly upwards, thus forming a shallow V-shaped arrangement. Leaves linear,flattened, uniform in width except at the tapering base; rounded and slightly bifid atthe apex, up to about i inch long and ^ to TV inch wide ; upper surface dark green,shining, with a median continuous groove, and with two or three broken rows ofstomata in the middle line towards the apex; lower surface with two narrow, greyishbands of stomata, composed of six to eight lines ; resin-canals median. Leaves oncone-bearing branches more or less upturned, stouter and broader than those onbarren shoots, acute and not bifid at the apex.

8 04 The Trees of Great Britain and IrelandStaminate flowers yellow, tinged with purple. Pistillate flowers with nearlyorbicular purple scales, shorter than the serrulate greenish-yellow bracts, which areemarginate above and end in long, recurved tips.Cones sub-sessile, ovoid-cylindrical, tapering both at the base and towards theround or flattened apex; purple l in colour, 2 to 4 inches long, about an inch indiameter. Scales, about f inch wide and long; lamina fan-shaped, rounded andundulate above, lateral margins denticulate and curving to the truncate or auricledbase; claw wedge-shaped. Bracts variable in length, exserted or concealed betweenthe scales; claw oblong; lamina trapezoidal and denticulate, ending in a mucro.Seeds purplish, about ^ inch long ; wing about as long as the body of the seed.In the wild state considerable variation occurs in the habit of the tree, whichbecomes a mere shrub at high altitudes. The cones vary both in size and in thelength of the bracts, which are either slightly exserted, or quite concealed betweenthe scales. Prof. Balfour found on the same tree at Keillour cones both with longand with short bracts.Var. Hudsonia, Engelmann, Trans. St. Louis Acad. iii. 597 (1878).Abies Hudsonia, Bosc. ex Carrifcre, Conif. i. 200 (1855).According to Engelmann this is a sterile dwarf form which occurs above thetimber line on the White Mountains in New Hampshire. Whether this isidentical with the A . Hudsonia, which occurs in cultivation, is uncertain.The latter, according to Sargent, 2 is of unknown origin, but is probably, thoughit has never produced cones, a depauperate form of A . balsamea. I t has denselycrowded branches, short numerous branchlets, and small broad leaves, about £ inchin length; and is a dwarf spreading shrub, only a foot or two in height. It differsfrom A . balsamea in having marginal resin-canals.Var. macrocarpa? This was discovered near the Wolf River, Wisconsin, andraised by Robert Douglas at Waukegan nursery; it is said to be a distinct andbeautiful form with longer leaves and larger cones than the type.DISTRIBUTIONThe balsam fir extends far to the northward in the Dominion of Canada, itsnortherly limit being a line drawn from the interior of Labrador north-westward tothe shores of the Lesser Slave Lake. It occurs in Newfoundland and in theprovinces of Quebec and Ontario, and descends in the United States in the westthrough northern Michigan and Minnesota to northern and central Iowa, and inthe east extends through New England and New York, along the Catskill andAlleghany mountains to south-western Virginia. It is common and often forms aconsiderable part of the forest on low swampy ground, while on well-drained hill-sidesit is met with as single trees or small groves chiefly in the spruce forests. It ascendsto 5000 feet on the Adirondacks. (A. H.)1 In cultivated specimens the cones are occasionally olive-green in colour, and rarely exceed 2 inches in length.2 Garden and Forest, x. Jio (1897). 3 Ibid. v. 274 (1892) and x. 510 (1897).Abies 805REMARKABLE TREESThe most noted trees of this species in cultivation were those growing in theKeillour Pinetum, Perthshire, now the property of Captain Black of Balgowan.This pinetum was visited by Prof. Balfour 1 in 1895, who found about 30 trees stillliving out of 200, which were planted in 1831. The largest tree was about 60 feethigh with a girth of 5 feet i inch at three feet from the ground. There were severalothers over 4 feet in girth. In 1904, when Henry made a hurried visit to theKeillour Pinetum, where there was much of interest to be seen, he only saw one treeof A . balsamea, with the top broken and in a dying state. Mr. W. Causand informedhim that in 1903 there was a tree 68 feet by 5 feet.The finest specimen of which we have any account in Great Britain is recordedin the Conifer Conference Report, Journ. Roy. Hort. Soc. xiv. 511, as havinggrown at Saltoun Hall, East Lothian, the seat of A. Fletcher, Esq., until 1891,when it was swept away by a flood on the river Tyne. This tree was supposed tohave been given by Bishop Compton, who introduced the species in 1697, to BishopBurnet, formerly incumbent of the parish of Pencaitland, and was thus somethinglike 190 years old. It was 68 feet high though it had lost its top, and at ten feetfrom the ground no less than 7 feet 10 inches in girth, and was said to have beenhealthy and growing vigorously up till the time of its destruction.In England we have never seen a tree of any great size or age, the largest beingat Bicton, 52 feet by 4 feet 4 inches in 1908 ; and this species seems to have beenneglected and forgotten by modern planters, as it is only twice mentioned in thenumerous reports sent to the Conifer Conference.Loudon states that it arrives at maturity in twenty to twenty-five years, afterwhich it soon dies, though he mentions trees of 30 to 40 feet high as then existing atSyon, Whitton, and Chiswick.It appears therefore to be of no horticultural value in this country, though if theSaltoun report was correct it may be grown successfully in some parts of Scotland.2In Norway, according to Schiibeler, the Balsam fir succeeds better than here.He mentions three at Bogstad near Christiania, planted about 1772, of which thelargest was 55 feet by 6 feet 4 inches, and another 8 feet 2 inches in girth ; butwhen I visited this place in 1904 I could not find these trees, and do not knowwhether they are still living. Hansen 3 states that specimens, about 50 years oldand 40 feet high, are to be met with in Danish Gardens.TIMBER, RESINSargent describes the wood 4 as being light, soft, coarse-grained, and perishable,and only used for cheap lumber. From the blisters on the bark, a straw-coloured1 See Card. Chron. xvii. 422 (1895), which gives an interesting account of this remarkable pinetum.2 A . balsamea was planted at Durris freely about fifty years ago, the largest trees now being from 40 to 45 feet in height.Timber of good quality, and contains an exceptionally small percentage of water in a green state. I have seen no accounttaken of the latter fact, but it has been a continual surprise to me in handling timber in a green state. (J. D. CROZIER.)3 Journ. Roy. Hort. Soc. xiv. 458 (1892).4 H. von Schrenk, in Missouri Bot. Garden Report, 1 905, p. 117, describes and figures logs of this timber, felled inMaine for pulpwood, which show on cross-section irregular areas, perfectly smooth and shining as if they had been planed.

806 The Trees of Great Britain and Irelandtransparent resin, known as Canada balsam, is collected by Indians and poor whitesin the province of Quebec. This resin, which was formerly largely used in medicineon account of its stimulating action on the mucous membrane, is now chiefly used formounting objects to be examined under the microscope, for which, and kindredpurposes, it is specially suitable by reason of its transparency. (H. J. E.)Abies 807have broader bands of stomata than in that species eight to twelve lines inA. Fraseri, usually only six lines in A . balsamea. The cones differ mainly in thelarger bracts, which are much exserted and refiexed over the edges of the scalesnext below ; whereas in A . balsamea the bracts are either concealed, or, if slightlyexserted, are never refiexed. (A. H.)ABIES FRASERI, SOUTHERN BALSAM FIRAbies Fraseri, Poiret, in Lamarck, Diet. Suppl. v. 35 (1817); Forbes, Pine turn Woburnense, i n,t. 38 (1840); Sargent, Silva N. Amer. xii. 105, t. 609 (1898), and Trees N. Amer. 5 7 (1905);Masters, Card. Chron. viii. 684, fig. 132 (1890); Kent, Veitch's Man. Coniferce, 5 09 (1900).Finns Fraseri, Lambert, Genus Pinus, ii. t. 42 (1837).Picea Fraseri, Loudon, Arb. et Frut. Brit. iv. 2340 (1838).A tree attaining in America 70 feet in height and 7 feet in girth, with ratherrigid branches, forming an open symmetrical pyramid. Bark smooth and withnumerous blisters in young trees, becoming on older trunks covered with thinappressed reddish scales. Buds small, broadly ovoid or globose, reddish, resinous.Young shoots smooth, yellowish grey, densely covered with reddish, short, twistedor curved hairs, the pubescence being retained on the older branchlets.Leaves on lateral branches pectinately arranged, as in A. balsamea; linear, flattened, shorter than in that species, rarely exceeding f inch long and ^ inch broad,uniform in width except at the shortly tapering base, rounded and bifid at the apex;upper surface dark green, shining, with a continuous median groove and withoutstomata; lower surface with two broad conspicuously white bands of stomata, eachof eight to twelve lines; resin-canals median. Leaves on cone-bearing shootsupturned, crowded, broader than on barren shoots, rounded and entire at the apex.Staminate flowers yellow tinged with red. Pistillate flowers, with roundedscales, shorter than the oblong bracts, which are broad and rounded above, endingin long slender tips.Cones sub-sessile, ovoid, cylindrical, slightly tapering at the base and towardsthe rounded or flattened apex, purple, about 2 inches long by ij inch in diameter,with the bracts conspicuously exserted and reflected. Scales as in A. balsamea, butwider in proportion to their length. Bracts ; claw oblong; lamina broad, trapezoidal,denticulate in margin and bifid above with a mucro in the emargination. Seed withwing about \ inch long ; wing purplish, broadly trapezoidal, denticulate in the uppermargin, about twice as long as the body of the seed.DISTRIBUTIONAbies Fraseri is very restricted in its range of distribution, being only found inthe Alleghany Mountains of south-western Virginia, North Carolina, and easternTennessee, where it often forms forests of considerable extent at elevations of 4000to 6000 feet above sea-level. These forests are usually pure ; but occasionally thisspecies grows mixed with black spruce, birch, and beech. The tree averages about40 feet in height; it only rarely attains 70 feet.Sargent in an article 1 on this species gives a good illustration of a forest, atabout 5000 feet altitude on the Black Mountain range, a spur of the Blue Ridge inNorth Carolina; which is very like some forests that I saw when I visited thismost interesting region in 1895.HISTORY AND CULTIVATIONAbies Fraseri was discovered by the Scotch traveller and botanist whose nameit bears, John Fraser, in the first decade of the nineteenth century; and plants of itwere first distributed from Messrs. Lee's nursery, at Hammersmith, in 1811. Theexcellent figure in Pinetum Woburnense, was taken from the original tree in thisnursery, where it had then attained 16 feet in height, at about twenty-eight years ofage.The tree is short-lived, and the plants of the first introduction are probably alllong since dead. According to Sargent, 1 seeds of A. balsamea, collected in Pennsylvania and Canada, where specimens are occasionally found, in which the tips of thebracts of the cone are slightly exserted, have been very generally sold as A. Fraseri.Seedlings of the Carolina tree were, however, distributed by the Arnold Arboretuma few years prior to 1889. We know of no trees of any size now living in thiscountry. Some seedlings which I brought from N. Carolina in 1895 soon died.1 Garden and Forest, ii. 472, fig. 132 (1889).(H. J. E.)IDENTIFICATIONThis species can readily be distinguished from Abies balsamea by the differentpubescence on the young hranchlets and the shorter, more coriaceous leaves, whichLogs with this so-called " glassy" appearance are occasionally rejected; but examination showed that this peculiarity wassimply due to the presence of ice, which follows the radial lines on the hcaled-over branches of the logs.

808 The Trees of Great Britain and IrelandABIES RELIGIOSA, M EXICAN FIRAbies religiosa, Schlechtendal, Lintuea, v. 7 7 (1830); Lindley, Penny. Cyd. i. 31 (1833); Seemann,Bot. Voy. 'Herald^ 3 35 (1852-1857); Hooker, Bet. Mag. t. 6753 (1884); Masters, Card.Chron. xxiii. 56, f. 13 (1885), and ix. 304, ff. 69, 70 (1891), and Journ. Linn. Soc. (Bot.)xxii. 194, t. 6 (1886); Kent, Veitch's Man. Coniferee, 5 36 (1900).Abies hirtella, IJndley, loc. fit. ( 1833).Finns religiosa, Humboldt, Bonpland et Kunth, Nov. Gen. et Spec. i i. 5 (1817); Parlatore, in DC.Prod. xvi. 2, p. 420 (1868).Pinus hirtella, Humboldt, Blonpland et Kunth, loc. at. ( 1817).Picea religiosa, Loudon, Arb. et Frut. Brit. iv. 2349 (1838).Picea hirtella, Loudon, loc. cit. ( 1838).A tree, attaining in Mexico 150 feet in height and 18 feet in girth. Bark 1greyish-white, rough, divided into small roundish plates.Buds shortly cylindrical, rounded at the apex, covered with white resin. Youngshoots brown on the upper surface, olive green beneath, covered with minute erectpubescence; pulvini prominent. Second year's shoots reddish-brown, smooth, andstriate between the pulvini, which are no longer raised.Leaves on lateral branches, arranged as in A . Nordmanniana; but with themedian upper leaves much fewer than in that species, covering the upper side ofthe branchlet, and pointing forwards and slightly upwards ; lower leaves in twolateral sets, spreading outwards and slightly forwards in the horizontal plane.Leaves twisted above the base, linear, flattened, gradually narrowing in the anteriorhalf to an obtuse apex, which is usually entire or rarely minutely bifid; uppersurface dark green, shining, with a median groove (usually not continued to theapex) and without stomata; 2 lower surface with two greyish bands of stomata,each of eight to ten lines; resin-canals marginal. The upper leaves are abouthalf the length of those below, the latter about an inch in length and about ^6 inchbroad. Leaves on cone-bearing branches similar to those on barren branches.Cones on short stout stalks, 4 inches long, 2 inches in diameter, conical, broadestnear the base and gradually tapering to an obtuse and narrowed apex, bluish beforeripening, dark brown when mature, the large reflexed bracts being then of a chestnutbrown colour. Scales broadly fan-shaped, nearly i£ inch wide by f inch long ; uppermargin almost entire ; lateral margins laciniate and denticulate ; base broad with asinus on each side of the short obcuneate claw. Bract: claw wide, obcuneate ;lamina quadrangular, denticulate, emarginate with a short triangular cusp. Seed withwing about f inch long; wing broad and ij times the length of the body of the seed.DISTRIBUTIONThis species extends throughout the mountains of Mexico, from near Durangoin the Sierra Madre range (lat. 24 ), where it was collected by Seemann, 3 to the1 In this tree, as in the other species with prominent pulvini on the branchlets, the bark of the trunk speedily becomesscaly and like that of a spruce, not remaining smooth for a considerable period, as in the common species of silver fir.2 On leaves towards the tip of the shoot, short irregular lines of stomata are present on their upper surface near the apex.Some of these leaves turn their ventral surfaces upwards towards the light. 3 Bot. Voy. Herald, 335 (1852).Abies 809mountains of northern Guatemala (lat. 15 ), where it was observed by Hartweg 1 andcollected by Skinner. It is known to the natives as Oyamel, and occurs mainly inforests at 8000 to 10,000 feet, though it occasionally descends to 4000 feet. Itapparently reaches its best development on the Campanario, the highest point ofthe mountains of Angangueo, a range about 100 miles west of the city of Mexico.Here Hartweg found trees 150 feet in height and 5 to 6 feet in diameter. Parry andPalmer collected it in the province of San Luis Potosi in Central Mexico, and gaveits range as from 6000 to 8000 feet. Linden found it on the peak of Orizaba, inlandfrom Vera Cruz, growing between 9000 and 10,000 feet elevation.Stahl, in Karsten and Schenk's Vegetationsbilder, 2 Reihe, Heft 3, gives a goodaccount of this tree, which he found growing near Orizaba between 2600 and 3500metres above sea-level, and in the higher mountains round the valley of Mexico,in pure forests or mixed with pines, oaks, and alders. He gives no dimensions, andthe two excellent figures 17 and 18 taken in the Sierra de Ajusco, near Salazar, atabout 9500 feet, show the trees to be smaller there than those which Elwes saw onPopocatapetl. 2Dr. Gadow s found it growing in the mountains of Omiltelme, at 8000 feet; anddescribes the trees as " veritable giants, from 5 to 6 feet in diameter, as straight as amast, and may be 100 feet high."Humboldt supposed that there were two species, one with glabrous and theother with pubescent branchlets; but Seemann and Hartweg were convinced thatthis distinction is unfounded; and the type specimen of Pinus religiosa, the supposedglabrous form, according to Bolle, has pubescent branchlets.The branches of the tree, which are very elegant, are used in Mexico fordecorating churches at the times of religious festivals.This species was discovered in 1799 by Humboldt, who saw it near the city ofMexico in two localities, at 4000 feet elevation between Masantla and Chilpantzingo,and near El Guardia at 8400 feet. It was introduced into cultivation in 1838 byHartweg, who collected for the Horticultural Society of London.REMARKABLE TREESAbies religiosa is tender and will not live, except in the warmer parts of theseislands, close to the sea coast, where the temperature never falls much below freezingpoint. Trees planted long ago at Kew and Bayfordbury, do not now survive.Murray mentions 4 in 1876 specimens growing at Woodstock in Kilkenny, Highnam1 Trans. Hort. Soc. iii. 123, 138 (1848).* I believe that this was the silver fir which clothes the lower slopes of the volcano of Popocatepetl, in Mexico, which Iascended to the limit of vegetation, about 13,000 feet, in March 1888, with my wife and Mr. F. D. Godman. The treesformed in some places dense forests at an elevation of 9000 to 10,000 feet, bnt though my recollection is that they grew to agreat size, we took no measurements, being at the time engaged in collecting birds and insects. In the dry volcanic soil inwhich they grow we found abundantly Pinguicula rosea, one of the most charming ornaments of our greenhouses; and higherup lupins and pentstemons were the most conspicuous plants. (H. J. E.)3 Through Southern Mexico, 378 (1908).* Card. Chron. v. 560 (1876).IVO

8io The Trees of Great Britain and IrelandCourt in Gloucestershire, Munches in Kirkcudbright, and Hafodunos in Denbighshire,places which we have visited; but none of these trees can now be found.There are several trees in Cornwall. Specimens with cones were sent to Kew in1899 from Trevince, near Redruth, the residence of Mr. E. B. Beauchamp. There isalso a tree 1 in Mr. Boscawen's garden at Lamorran, which produced cones in iSgo. 2There is a small tree at Mr. Rashleigh's garden, Menabilly, which was figured in theGardeners Chronicle ; 1 and at Tregothnan, the seat of Viscount Falmouth, there is alarge tree which Elwes measured in 1905 as 56 feet by 6^ feet. Though bearingcones, this tree did not seem healthy, and its top was broken by the wind.A tree at Castle Kennedy is fairly large in size; but it was blown down someyears ago, and then replaced in position. It is in consequence very irregular in shape.It produces cones freely, but the seeds are never fertile.There were formerly two trees at Fota, which differed somewhat in colour ofthe foliage and hardihood ; one 8 has since been blown down. The surviving tree(Plate 226) is a handsome one, though its trunk was broken at about thirty-six feetup, and it has now four leaders: when measured by Elwes, in 1908, it was 66 feethigh by 7 feet 3 inches in girth. It is branched to the ground, one very largebranch coming off near the base. The foliage is variable in colour, being bluishgreentowards the ends of the branchlets, and elsewhere of a light or dark greencolour, so that there are three tints visible on the tree. It was bearing in August1904 numerous cones and male flowers, scattered all over the tree. The cones exudea white resin ; and are peculiar, as the scales do not all fall at the same time, someremaining at the base and apex of the axis for two or three years.The tree does remarkably well on the shores of the Italian Lakes. 4 Carrieresays it is killed by frost at Paris; but at Cherbourg 5 there was a tree 30 feet high in1867. It seems, however, to be very rare if at all existing in France.(A. H.)1 Card. Chron. ix. 304, figs. 69, 70 (1891).2 This tree was the first in Britain to produce cones, which were exhibited at the Royal Horticultural Society in 1876.3 This tree was much more tender than the other, and had the top and some lateral branches killed by frost in thewinter of 1880-1881. See Osborne, in Card. Chron. xxiii. 56 (1885).4 Sargent, Silva N. Amer. xii. 97, adnot. ( 1898). But Elwes saw none that he could identify in the neighbourhood ofFallanza.6 Hickel and Parde, in Bull. Sac. Dendr. France, 1 908, pp. 206, 224, state that the trees of this species in theneighbourhood of Cherbourg died in the severe winter of 1879-1880; and believe that there are now no living specimensin France.PSEUDOTSUGAPseudotsuga, Carriere, Conif. 2 56 (1867); Bentham et Hooker, Gen. PL iii. 441 (1880); Masters,Jonrn. Linn. Soc. (Sot.) xxx. 35 (1893).Abies, section Peucoides, Spach, Hist. V£g. xi. 423 (1842).Pinus, section Tsuga, Endlicher, Gen. PI. Suppl. iv. Ft. ii. 6 (1847).Tsuga, section Peucoides, Engelmann, Trans. St. Louis Acad. ii. 211 (1863).Abietia, Kent, Veitch's Man. Conifercz, 4 74 (1900).EVERGREEN trees belonging to the tribe Abietineae of the order Coniferae. Branchesirregularly whorled. Branchlets of one kind; pulvini slightly projecting, persistent,and showing, when the leaves have fallen, an oval scar at their apex. Buds spindleshaped,acute at the apex, brownish, shining, glabrous ; one terminal larger, and oneto four lateral and smaller in the axils of the uppermost leaves; scales numerous,imbricated, rounded and entire at the upper margin, increasing in size from belowupwards; some of the scales persistent for three or four years at the base of thebranchlets, ultimately falling and leaving ring-like scars. Leaves arising in spiralorder; but on lateral branches, thrown by a twist of their bases into two spreadingranks; persistent for four to eight years; linear, flat, narrowed at the base; uppersurface green and longitudinally furrowed; lower surface with a prominent midriband two stomatic bands ; fibro-vascular bundle single, resin-canals two on the undersurface next the epidermis.Flowers, arising from buds formed in the previous summer, erect, solitary,surrounded at the base by involucral scales. Male flowers axillary, scattered alongthe branchlets, cylindrical; pedicel short at first, ultimately elongated; composed ofnumerous spirally arranged short-stalked globose anthers, opening obliquely: connective surmounted by a short spur; pollen-grains globose, without air-sacs. Pistillate flowers, terminal or in the axils of the uppermost leaves, composed of numerousspirally imbricated rounded scales, much shorter than the acutely three-lobed bracts;ovules two on each scale, inverted. Fruit, a woody pendulous cone, ripening in thefirst season, ovoid-oblong, acute at the apex, rounded and narrowed at the base;peduncle short and stout; scales persistent on the axis after the fall of the seeds,small and sterile towards the base and apex of the cone, rounded, concave, rigid;bracts conspicuous, exserted, oblong, three-lobed at the apex, the middle lobe awnlikeand longer than the two lateral lobes. Seeds, two in shallow depressions, whichoccupy about half the surface of the scale, triangular, without resin-vesicles, winged.Cotyledons, six to twelve, linear; with a prominent midrib and stomatiferous on theupper surface.811

812, The Trees of Great Britain and IrelandThe genus comprises three species, two inhabiting western North America,and the third restricted to small areas in Japan and Formosa.In the absence of cones, they are distinguishable as follows :1. Pseudotsuga Douglasii, Carriere. Western North America.Branchlets usually pubescent, occasionally glabrous. Leaves straight,undivided at the apex.2. Pseiidotsuga macrocarpa, Mayr. Southern California.Branchlets covered with short, stiff pubescence. Leaves curved,undivided at the apex.3. Pseudotsuga japonica, Sargent. Japan, Formosa.Branchlets glabrous. Leaves straight or curved, bifid at the apex.The latter two species, not being yet introduced into England, will now be brieflydealt with.PSEUDOTSUGA MACROCARPA, Mayr, Wald. Nordamer. 2 78 (1890); Sargent, SilvaN. Amer. xii. 93, t. 608 (1898), and Trees N. America, 54 (1905).Abies Douglasii, var. macrocarpa, Torrey, Ives' Rep. pt. iv. 28 (1861).Abies macrocarpa, Vasey, Gardener? Monthly, xviii. 21 (1876).Tsuga macrocarpa, Lemmon, Pacific Rural Press, xvii. No. 5, p. 75 (1879).Pseudotsuga Douglasii, var. macrocarpa, Engelmann, in Brewer and Watson, Bot. Calif, ii. 120(1880).Abietia Douglasii, var. macrocarpa, Kent, Veitch's Man. Conifera, 4 78 (1900).A tree usually 50, rarely 80 feet high, with a trunk 3 to 4 feet in diameter. Itdiffers from the common species in the following characters: Branches comparatively larger and more remotely placed. Branchlets covered with a short, stiff,white pubescence. Leaves, f to i inch long, resembling those of P. Douglasii,except that they are distinctly curved. Buds short and broad, usually not morethan \ inch long. Cones very large, 4^ to 6 inches long ; scales \\ to 2 inches wide,thick, very concave, puberulous on the outer surface; bracts, only slightly exserted,short, narrow, with broad midribs produced into short flattened flexible tips. Seeds,£ inch long, dark brown or nearly black and shining above, pale brown below ; wing\ inch long.This species 1 occupies an isolated area in the arid mountains of southernCalifornia, at 3000 to 5000 feet elevation, forming open groves or growing inmixture with oak and pines on western and southern slopes. Its distribution extendsfrom the Santa Inez Mountains near Santa Barbara on the coast to the CuyamacaMountains on the southern border of California.PSEUDOTSUGA JAPONICA, Sargent, Silva N. Amer. x ii. 84, adnot. 2 (1898);Shirasawa, Icon. Ess. Forest. Japon, text 21, t. 7 (1900).Tsuga japonica, Shirasawa, Tokyo Bot. Mag. ix. 86, t. 3 (1895).This species is not represented by dried material in European herbaria; but Ihave seen a specimen 2 recently sent from Japan by Capt. L. Clinton Baker, R.N.1 A view of a forest of this species is given in Garden and Forest, x. 24, f. 5 (1897).The buds on this specimen were not developed ; but the scales of the previous season's buds remained persistent at thebase of the branchlets, and resembled those of P. Douglasii.Pseudotsuga 813It is distinguished from the other species by its glabrous branchlets and by its leavesbifid at the apex. The leaves are pectinately arranged, f to i inch long, ~fato -^ inch wide, straight or curved, yellowish green above, conspicuously whitebeneath, broadest near the contracted base, and gradually tapering to an acute apex,which is minutely bifid. The cones are small, ijto if inch long, i inch in diameter;scales few, about twenty in number, more woody in consistence than those of P.Douglasii, glabrous externally ; bracts strongly reflexed, the central awn-like lobeonly slightly larger than the lateral lobes. According to Shirasawa, its discoverer, 1the tree attains a height of 100 feet and a diameter of 3 feet, and occurs at 1000 to3000 feet elevation in the mountains of the provinces of Ise, Yamato, and Kii inJapan. It grows in mixed forests, composed mainly of Tsuga, Oak, Beech, Magnolia, and other broad-leaved species. Elwes, when at Koyasan, endeavoured toreach the habitat of this species, but owing to the distance, the heavy rain, andinability to find a guide, was unsuccessful. According to Hayata, 2 this speciesoccurs also on Mount Morrison in Formosa. Its Japanese name is Togasawara.Young plants are reported by Beissner 3 to be in cultivation in Ansorge'snursery, at Flottbeck near Altona, and in the Botanic Garden at Hamburg. Twosmall branches, recently sent to Kew from Flottbeck and from Herr Langen'snursery at Grevenbroich, are only distinguishable from those of the American speciesby some of the leaves being bifid at the apex. Apparently in the young stage, theleaves are acute or mucronate and entire, the bifid character only being assumedafter two or three years.Except for its botanical interest this species does not seem likely to have anyvalue in this country. 4 (A. H.)1 Shirasawa discovered this species in July 1893, on the road between Owashi (in Kii province) and Yoshino (Yamatoprovince), about 10 miles from the coast. He states that the forests in which it occurs are small in area and very inaccessible.2 Tokyo Bot. Mag. xix. 45 (1905).3 Mitt. Dent. Dendr. Gesell. 1 902, p. 53, and 1906, pp. 84 and 144. Mayr, in Fremdldnd. Wald- u. Parkbdume, 406(1906), states that seeds of the Japanese species have never germinated in Europe. The young plants, however, referred toabove, are unquestionably this species.4 While the above was passing through the press, Mr. H. Clinton Baker writes that he had just received from Fallanzafour plants of /'. japonica, about 2 feet high, which are being planted at Bayfordbury. The buds on these plants are about\ inch long, shining brown, and without resin ; and the leaves are nearly all bifid at the apex.

814 The Trees of Great Britain and IrelandPSEUDOTSUGA DOUGLASII, DOUGLAS FIRPseudotsuga Douglasii, Carriere, Conif. 2 56 (1867); Mayr, Fremdland. Wald- u. Parkbdume, 396(1906).Pseudotsuga Lindleyana, Carriere, Rev. Hort. 1 868, p. 152, fig.Pseudotsuga taxifolia?- Britton, Trans. N. York Acad. Sc. viii. 74 (1889); Sargent, £ot. Gazette,xliv. 226 (1907).Pseudotsuga mucronata, Sudworth, Contrib. U.S. Nat. Herb. iii. 266 (1895); Sargent, Silva N.Atner. xii. 87, t. 607 (1898), and Trees N. Amer. 5 3 (1905).Pseudotstiga glaucescens, Bailly, Rev. Hort. 1 895, p. 88, fig.; Andre, Rev. Hort. 1 895, p. 159;Bellair, Rev. Hort. 1 903, p. 208, f. 85.Pseudotsuga glauca, Mayr. Mitt. Dent. Dendr. Ges. 1 902, p. 86, and Fremdland. Wald-u. Parkbamne,404 (1906).Pinus taxifolia, Lambert, Pinus, i. 51, t. 33 (1803) (not Salisbury).Pinus Douglasii, D. Don, in Lambert, Pinus, iii. t. (1837).Abies taxifolia, Poiret, in Lamarck, Diet. vi. 523 (1804).Abies mucronata, Rafinesque, Atlant. Journ. 1 20 (1832).Abies Douglasii, Lindley, Penny Cycl. i. 32 (1833); Loudon, Art. et Frut. Brit. iv. 2319 (1838).Picea Douglasii, Link, Linncea, xv. 524 (1841).Tsuga Douglasii, Carriere, Conif. 1 92 (1855).Tsuga Lindleyana, Roezl, Cat. Conif. Mex. 8 (1857).Tsuga taxifolia, Kuntze, Rev. Gen. PI. ii. 802 (1891).Abietia Douglasii, Kent, Veitch's Man. Conif. 4 76 (1900).A tree, attaining in the moist climate of the Pacific coast 250 to 300 feet inheight and 40 feet in girth ; but in the dry regions of the interior and at highaltitudes rarely more than 100 feet high and 10 feet in girth. Bark of young stemsthin, smooth, shining, grey; on older trunks, 2 to 12 inches in thickness, corky,divided by deep longitudinal furrows into broad oblong scaly ridges. Youngbranchlets usually pubescent, occasionally glabrous. Buds \ to | inch long. Leavesf to \\ inch long, straight, rounded or obtuse, rarely acute at the apex; variable incolour, the stomatic bands beneath either dull grey or conspicuously white.Cones, 2 to 4^ inches long; scales thin, slightly concave, rounded or slightlyprolonged at the apex, about f inch wide; pubescent on both surfaces; beforeripening bluish below, purple towards the apex and bright red on the closelyappressed margins, the bracts being pale green ; scales and bracts brown when ripe.Bracts variable in length; the three-pointed apex always, however, extending beyondthe scale, usually appressed, but occasionally reflexed. Seeds, about \ inch long,reddish brown and shining above, paler and with whitish spots below; wings longerthan the body of the seed, dark brown, rounded at the apex.VARIETIESi. Var. glauca, Beissner, Nadelholzkunde 4 19 (1891), Colorado Douglas fir.In the interior of the continent, the Douglas fir, growing in a dry climate at1 According to the rules of botanical nomenclature adopted by the Vienna Congress of 1905, P. taxifolia is the correctname for the species, as pointed out by Sargent, in Bot. Caz. xliv. 226 (1907) ; but we prefer to use P. Douglasii, thename which is universally in use amongst foresters and arboriculturists.Pseudotsugahigh elevations in the Rocky Mountains, through Montana, Colorado, Utah, Arizona,New Mexico, and Mexico, is a smaller tree than the form which occurs in the moistclimate of the Pacific coast region. It bears small cones, 2 to 3 inches in length,which in rare cases have the bracts reflexed, but resemble in all essential characters,except size, the cones of the coast form. The leaves are usually thicker in textureand are very glaucous beneath; but the bluish tint visible on the upper surfaceof the leaves, which is supposed to be characteristic, while common in certainlocalities, and in others occurring on scattered individual trees, is no more constantthan the similar coloured variation which is met with in trees like Picea pungens andCedrus atlantica. Mayr has separated the Rocky Mountain form as a distinctspecies, P. glauca ; but the differences, being rather physiological than morphological, do not entitle it to rank as more than a variety. The main difference liesin the rate of growth and the hardiness of the tree, when seeds of it are raised incountries remote from its native habitat.Dr. C. C. Parry discovered this variety of the Douglas fir in the outer rangesof the Rocky Mountains in 1862 ; and in the following year seeds were sent to theBotanic Garden of Harvard College, from which plants were raised, that haveproved perfectly hardy and vigorous in growth in New England. In the northeastern States the Pacific Coast form, whether introduced by seeds collected inOregon or produced by trees growing in England, has not proved hardy.The exact date of the introduction of the Colorado Douglas into Europe isuncertain; but it appears to have been unknown in 1884, when the first edition ofVeitch's Manual was published, and was described as a distinct variety by Beissnerin 1891. Seeds were apparently sent from Mexico by Roezl in 1856, and plants 1raised from these on the continent do not seem to differ from the Colorado Douglas.According to the experiments of Johannes Rafn, of Copenhagen, the germinationof the seed of Douglas fir from Colorado is quicker and much better than that fromthe Pacific coast. 2In England young plants of the Colorado Douglas 3 have ascending branches,and are more narrowly pyramidal in habit than the Oregon Douglas, which haswide-spreading horizontal branches. Owing to its slowness of growth, the Coloradovariety has short internodes between the branches, which give it a bushy appearance.The blue tint of the foliage can scarcely be relied on as a distinctive character, as itis variable in intensity and often disappears with age. The leaves are usuallythicker, but do not differ in length or shape from those of the Oregon Douglas, thesharp-pointed apex being characteristic of both forms in the young stage. Theyoung branchlets of the Colorado variety are often either quite glabrous or show onlya few minute hairs under the lens, whereas those of the other form are distinctlypubescent. In wild trees, judging from herbarium specimens, this distinction doesnot occur.1 Pseudotsuga Lindleyana, Carriere, raised from Mexican seed sent by Roezl, and P. glaucesccns, Bailly, also probablyfrom Mexican seed, belong to var. glauca, and bear cones with strongly reflexed bracts.' Trans. Roy. Scot. Arbor. Soc., xvi. 408 (1901).3 The Colorado Douglas in cultivation in England has been supposed by Schwappach (cf. Richardson in Trans. Key.Scot. Arbor. Soc. xviii. 195, with figure) to be Pseudotsuga macrocarpa; but there is no evidence to support this opinion.

816 The Trees of Great Britain and IrelandThe Colorado Douglas displays in cultivation well-marked peculiarities, whichare mentioned in detail on pages 825, 826.2. Lemmon 1 has described several other wild varieties, as var. suberosa fromArizona and New Mexico, var. elongata from the base of Mount Hood in Oregon,and var. palustris from swamps in the Lower Columbia Valley.3. A considerable number of cultivated varieties have been distinguished byCarriere and Beissner, most of which are not worth mentioning, as their distinctivecharacters are trifling and inconstant. Fastigiate and pendulous forms are known,but are rarely met with.Var. Stairii, with yellowish foliage, originated z at Castle Kennedy.Var. Fretsii, Beissner, 8 is very peculiar in the foliage, as the leaves are shortand broad, only \ inch in length, very obtuse at the apex, and resembling those ofTsuga Sieboldii. This originated in the seed-bed, and was sent out by Messrs.Frets & Sons of Boskoop, Holland. (A. H.)Other varieties occur in cultivation which, though very distinct in habit, are not,in my opinion, worth naming. Among the best of the pendulous forms is one atBury Hill, Dorking, the seat of R. Barclay, Esq., which in 1908 was 88 feet high.I noticed in October 1907, near Boldrewood in the New Forest, on the northside of the drive, two trees, one of which was a typical Oregon Douglas fir withdrooping branches, and leaves very silvery on the under side. Another close by hadmuch narrower, stiffer, and darker foliage, and denser branches, the leaves much lesssilvery, and the cones closely packed near the summit of the tree in a mannerunusual in this species. The former tree measured 70 feet by 6 feet 10 inches, thelatter 66 feet by 5 feet 2 inches. Others of the latter type, standing near the gateleading into Mark Ash, bore no cones at all.Near Eggesford House, in the higher walk, I saw a Douglas fir tree, sodistinct in habit that it might be easily mistaken for another species. It had thingreyish foliage, pendulous branchlets, and very few cones, and measured 80 feet by 5feet 7 inches, whilst ordinary Douglas firs planted close by were much thicker in proportion. I believe that by selecting such trees as seed-bearers we may ultimatelysucceed in obtaining distinct races which, for economic planting, will be much morevaluable than trees of unknown origin. (H. J. E.)DISTRIBUTIONThe Douglas fir has an extremely wide distribution in western North America,extending from north to south over 33 of latitude, between the parallels of 55 and22 , and ranging from the Pacific coast to east of the Rocky Mountains. It occupiespractically all of this vast territory except the higher elevations of the mountains andthe desert and prairie regions of lower altitudes, where the rainfall is slight. It isthe dominant tree of the great western forest, always growing in mixture with other1 West American Conebearers, 57 (1895).2 This variety is fully described in Card. Chron. 1 871, p. 1481. I have seen the original, which is now a smallunhealthy looking tree ; as are all those we have seen elsewhere. The best, perhaps, is a large dense bush rather than a tree,growing in Wood's nursery, near Buxsted, Sussex. (H. J. E.)3 Milt. Dent. Vend. Cesell. 1 905, p. 74, f. 8.Pseud otsuga 817conifers, which have a much more restricted distribution. In Montana it is associatedwith the western larch : in California it encroaches on the redwood belt; in southwestern Oregon it is mixed with the Lawson cypress; while in the rest of the greatforest of this state, and of Washington and northern Idaho, Thuya plicata is usuallyits constant companion. The various silver firs, hemlocks, and the Sitka spruce alsotake part, in different localities, in the mixture of coniferous trees in the Douglasforest. Towards the edges of the prairie regions and in the drier parts of themountains, the Douglas fir gradually gives place to Pinus ponderosa, which is thecharacteristic tree of dry soils, where a very moderate rainfall prevails.The northern limit of the Douglas fir extends from near the head of the SkeenaRiver, latitude 54 , in the coast range of British Columbia to Lake Tacla in the RockyMountains, latitude 55 , reaching its most easterly point near Calgary in Alberta. Inthe coast range, the tree grows at some distance inland north of latitude 51 ; whilesouth of this line it is common on the coast of the mainland and in the island ofVancouver ; and in this region, and in Washington and Oregon, between the westernfoothills of the Cascades and the sea, it is most abundant and of its largest size.It attains its maximum development, 300 feet in height, in Vancouver Island andon the northern slopes of the Olympic Mountains in Washington, where the rainfallis excessive ; whereas, on the Cascades and in the interior of the continent, it rarelyexceeds 150 feet in height. It is common, but only of moderate size, in the forestsof northern Idaho and of western Montana, 1 ascending to 6000 feet.The Douglas fir extends southwards along the Rocky Mountains, in theYellowstone Park in Colorado, where it grows between 6000 and 11,000 feetaltitude; in Utah, to the east of the Wasatch range; in northern and central NewMexico and northern Arizona, where it is common between 8200 and 9000 feet,being rare and of small size in the southern parts of these two states, where itascends to 6000 or 7000 feet; in the Guadalupe Mountains of western Texas, whereit is abundant; and it spreads into Mexico, along the Sierra Madre range ofChihuahua and the mountains of Nuevo Leon, reaching its most southerly pointnear the city of San Luis Potosi.In California it extends southward in the coast mountains 2 as far as PuntaGordain Monterey county, but is not abundant, and is rarely over 150 feet in height;inland it extends to the Sierra Nevada, 8 where it grows to a large size and ascendsto 7000 feet. It does not occur in the arid tracts of Nevada and Utah, which liebetween the Sierra Nevada and the Wasatch ranges. (A. H.)So little seems to be known by British foresters as to the conditions underwhich the tree grows in America, that though I quite agree with the precedingaccount, it may be as well to add some of my personal experience of the tree as I sawit on my last journey in 1904. In the Blackfoot valley of Montana it is associated1 At \Vhitefish, Montana, an average tree, growing with the western larch, was 140 feet in height and 8 feet in girth,and showed 245 annual rings ; the sapwood, £ inch wide, containing 45 rings; the bark was 2^ inches in thickness.2 Jepson, in Flora Western Mid. California, 1 9 (1901) says that it is frequent in the Santa Cruz mountains; but isnot known in the Mt. Diabolo and Mt. Hamilton ranges, or in the Oakland hills.3 Sargent in Garden and Forest, x. 25 (1897), says that it does not extend in the Sierras, south of the head of King'sriver, or within loo miles of the territory occupied by P. macrocarpa. Jepson (op. cit. 20), makes its southern limit on theSierras, about the head-waters of Stevenson Creek, which is not far from the head of King's river.XV p

818 The Trees of Great Britain and Irelandwith Larix occidentalis, on the damper and shadier slopes of the mountains, at 4000to 6000 feet, giving place to Pinus ponderosa in drier and sunnier situations, and neverattains, so far as I could see or learn, more than 140 to 150 feet in height.In Washington and British Columbia it is not seen in the dry country east ofthe Cascade range, but appears 'as soon as the forest begins to thicken near thewatershed ; and on the western slopes of the mountains, from about 6000 feet downwards, is almost everywhere, except in swampy land, the dominant tree of theforest, attaining 200 to 300 feet in height from sea level to about 2000 feet.It grows usually in mixture with Thuya plicata, Tsuga Albertiana, Piceasitchensis, and Abiesgrandis ; sometimes with a smaller proportion of Pinus monticola,and in drier situations with Pinus ponderosa ? but in all the coast forests which I sawin Oregon, Washington, and British Columbia, including Vancouver Island, it outnumbers all the other conifers, except where forest fires have destroyed it, and itsplace is being to some extent taken by the hemlock, whose seeds seem able togerminate and grow in denser shade and in deeper humus than the young plants ofthe Douglas fir can endure. Wherever the soil becomes too dry and rocky forhemlock and Thuya, the Douglas fir is able to grow, climbing up to the dry ridgesand sunny slopes until it meets the more alpine species of conifers. Its habit andsize vary according to the soil and situation; but I never observed any trees even inthe most open situations, whose branches extended so far from the trunk as they doin English parks and gardens, and it does not attain anything like its full size unlessit has a deep soil, a sheltered situation, and has been drawn up in youth by thestruggle for existence, which prevails everywhere in the forest.I saw a section of bark in the Washington State exhibit at the St. LouisExhibition, taken from a tree cut at M'Cormick in Lewis Co., Washington, in thespring of 1904 ; which was said by Mr. Baker, who was in charge of it, to have been390 feet high. The same tree was recorded, however, in a Washington newspaperas having been 340 feet high and 42 feet in circumference (probably at three to fourfeet from the ground), and above 300 years old. The tree is said to have contained79,218 feet board measure, equal to above 8000 cubic feet, quarter-girth measure.The discrepancy in the account of the height and that given me by Mr. Baker mayarise, in part, from the tree in falling, having jumped some distance from its stump.Another tree even more remarkable, though not so large, was cut by Mr.Angus M'Dougall of Tacoma for the Chicago Exhibition in 1893. This grew inSnohomish Co., Washington, and measured on the stump only 4 feet in diameter. Infalling it broke off at a height of 238 feet, where it measured 17^ inches in diameter,and was nearly free from branches to a height of 216 feet, which length was sent toChicago.The largest tree I have ever seen myself, which is said to be perhaps the largestknown in Vancouver Island, grows by the roadside at Mr. P. Barkley's farm atWestholme, about 40 miles north of Victoria and 4 miles south of Chemainus Station.1 In the Bow river of Alberta it grows mixed with aspen (Populus tremuloides), and cottonwood (P. talsamifera).Wilcox, T!ie Rockies of Canada, p. 65.Pseudotsuga 819What its height may originally have been is impossible to say, as it is broken offat about 175 feet. This tree has a very swelling base, which does not show sowell as I could wish in the photograph (Plate 227). At the ground it measures 21paces in circumference. Above the swelling, at about 6^ feet, I made it 41 feet 5inches in girth. Assuming this tree to be 24 feet in girth at 100 feet high, and tohave had a top at all in proportion to its girth, it must have contained 7000 to 8000cubic feet of timber, or even more. The soil in which it grows is a deep fertile loam,and the timber standing in the valley is some of the finest in the island. Plate 228,from a photograph also taken in Vancouver Island, gives an idea of the forest, andshows on the right the trunk of a typical Douglas Fir; on the left, a trunk ofThuya plicata.In the eastern part of the Washington Forest Reserve, Mr. Martin W. Germanfound this species up to 6000 feet, and measured a tree growing at 5500 feet, 132years old, which was i8f inches diameter on the stump, with the bark 3 inches thick.Another tree at 1200 feet elevation, 244 years old, was 43 inches in diameter, withbark 6 inches thick. In the dry region the tree ranges from 70 to 120 feet high andfrom 20 to 50 inches diameter. He remarks that the species resists fire better thanany other conifer of this region, and bears fertile cones at an earlier age than anyother, a tree of only twelve years old having well developed cones.Observations on the rapid growth of Douglas fir at various ages in itsown country are given by Mayr. 1 In southern Oregon, on the best sandy loam, witha rich humus, he measured Douglas firs 130 feet high at eighty years old; a fallenstem, 100 feet high, contained 135 cubic feet.The wood of the Douglas fir is known in the European, South African, andAustralian markets as Oregon pine or Oregon fir, on the Pacific coast of NorthAmerica as red fir or yellow fir, in Utah, Idaho, and Colorado as red pine,and in California is sometimes incorrectly called spruce or hemlock. It producesprobably a larger quantity of commercial lumber than any other conifer in thenew world, or at least on the Pacific coast; and is likely to continue the principalsource of supply for most purposes, as the white pine (P. Strobus) of the NewEngland states and Canada, and the long-leaved or pitch pine of the southernstates become scarcer; and as its timber is likely in the future to become animportant article of trade in Europe, both as an imported and home-grown product,I think it may be useful to give some particulars of the way in which the immensesawmills of Oregon, Washington, and British Columbia are managed.First, as regards the growth of the timber, Prof. Sheldon, the Oregon Stateforestry expert, has published in the Oregon Timberman, May 1904, a valuablepaper, which entirely confirms my own much more limited observations, and goesto show that the two forms locally known as red and yellow fir are not inany way distinct, but are simply the result of different conditions of growth.When the trees grow in an open space, and have the annual rings, as isusually the case in youth, pretty far apart, they may attain at the butt 16 to 18 inchesdiameter at forty years. In such trees the thickness of the sap-wood is from 2 to 31 Fremdldnd. Wald- u. Parkbaume, 398 (1906).

820 The Trees of Great Britain and Irelandinches, and the thickness of the bark, which under such conditions is comparativelysmooth and greyish in colour, is about 5- inch. The timber of such trees would beknown as red fir. When the tree, however, becomes crowded by its neighbours, andits girth increment is much slower, all the energy of the tree being devoted toupward growth, the rings become much closer, and trees of fifty to sixty years of agemay be only i foot in diameter. The bark in such cases is much thinner, and thequality of the timber from the point at which the slower growth began muchbetter, so that it would be classed as yellow fir when sawn up. Prof.Sheldon gives figures showing sections of such trees, his Plate 6 showing theinfluence of light, room, and nourishment on the growing tree. The tree fromwhich this section was made was 143 years old with a diameter of only 16înches. For 116 years it had stood in a crowded forest with large trees 4 and 5 feetin diameter all round it. Twenty-seven years ago the large trees were felled, andthe growth immediately became much more rapid. The sap-wood in this case is 3^to 4 inches and the bark i^ inch thick. He says, " The result of this study is toconclude that the rapid growth of Oregon fir in the open produces red fir, and thesubsequent growth when the trees begin to crowd each other produces yellow fir.Trees grown in dense clumps crowded all their life produce solid yellow fir. Thegrowth of the upper portion of the tree may show larger annual growths in the centrethan are found near the butt of the same tree. This is of interest in accounting forthe immense height of the Oregon fir in many places, as trees 300 to 350 feethigh are found in the forests of Oregon and Washington."I asked experienced loggers whether they could distinguish red from yellow firas they grew, and my impression was that they could not, though they said a veryfew blows of the axe would soon show the difference in the hardness of the wood.With the object of finding out the age at which the tree comes to maturity, I measuredthe rings of several trees recently felled at the logging camps which I visited. I ammuch indebted to the managers of these mills, for the facilities which they gave me tosee the whole operations of a modern west coast lumberman. Among them Mr.Bradley of the Bridal Veil Company, Oregon ; Mr. Browne, president of the St. Pauland Tacoma Sawmills, Tacoma, Washington, and his logging contractor, Mr.M'Dougall; Mr. Palmer of the Chemainus Mills, Vancouver Island; and Mr. KennethRoss, manager of the Big Blackfoot Lumber Company, Montana, were all mostobliging and hospitable.I found that the average age of mature trees 4 to 6 feet in diameter on the stumpis 300 to 500 years. At an age of from 400 to 500 years, and possibly muchsooner in some cases, the trees begin to decline in health, and some of thosefelled are more or less hollow. In all cases the annual rings for the first fifty toseventy years are very much thicker than for the next 300 years, the best treeshaving from four to five rings to the inch at first, and afterwards as many asfifteen to twenty. The better class trees are clear of branches up to about 120to 150 feet, and in such cases produce wood free from knots, or " clear lumber " as itis called in the trade. Such clear lumber, however, even when a large number oftrees are rejected by the fellers, does not exceed 15 to 30 per cent of the totalPseudotsuga 821product, and is worth a much higher price than the more or less knotty lumber knownas " merchantable."The business of lumbering which has been carried on for many years on a verylarge scale is, on the Pacific coast, as in most parts of North America, conducted in away which, though perhaps necessary in order to meet the severe competition forprice which everywhere prevails, would shock the feelings of any European forester,on account of its wastefulness and the absolute disregard which is paid to the futureof the forest; which is in most cases abandoned to fire, as soon as the soundest,cleanest, and most accessible trees have been extracted.A tract of land having been first surveyed, and its probable contents roughlyestimated by the " cruiser," on whose judgment in selecting the best field of operationsmuch of the success of the business depends, is purchased or leased from theowner on the basis of so much per thousand feet board measure. This estimateruns in most cases from 20,000 to 70,000 feet per acre, and as far as I could judgeis rarely more than half, and often much less than half, of the actual contents, whichin favourable situations amounts to as much as 300,000 to 500,000 feet per acre.Unless the timber to be felled is near the sea, in which case it is on Puget Soundoften slid direct into the salt water, made up into rafts, and towed by steamers to thesawmill, the next operation is to build a railroad up the valley to bring the logsfrom the forest to the sawmill. Sometimes the mill is in the forest itself, and awooden flume of many miles in length is built, by which the sawn boards can befloated down to the nearest railway station. Sometimes the logs themselves arefloated to the mill, where a large enough river exists ; or a combination of railway,river, and flume may have to be adopted as the distance from the mill or stationincreases. The cost of extracting the logs from the forest and bringing them totheir shipping point, governs the value of the growing timber, which is rapidlybecoming less and less accessible as the best areas are cut over.When the means of transport are completed, a " skid road" or a temporarytramway is built right up to where the trees grow, and powerful movable donkeyengines are used, which are able, with a steel-wire rope, to drag logs of 40 to 50feet long to a distance of 1000 yards or more from where they fall. Felling thencommences and is managed as follows : The most experienced man in the gang,having marked the trees to be felled, cuts a deep notch into one side at 4 to 6 feetfrom the ground, after carefully considering which way the tree should fall, so asto run least risk of lodging, or of breaking in falling. Both the undercutting andthe sawing which follows, are done on spring boards fixed into a notch cut intothe butt at 3 to 4 feet from the ground. When the two fellers, who sometimesmake the notch themselves, have got within 5 to 6 inches of it, they insert largeiron wedges in the sawcut, carefully watching the top of the tree to see wherethe wedges should be driven, so as to fell the tree with least danger to themselvesand the log. After a few blows on the wedges the tree begins to lean and themen jump clear, calling out to warn others who may be near. There is somerisk of large branches being torn off the falling tree or adjacent trees, and manyaccidents occur.

822 The Trees of Great Britain and IrelandWhen the tree comes down, it is cross-cut by other men paid at a lower ratethan the fellers, into such lengths as seem best. The smaller end of the log isthen bevelled off and two deep notches cut, into which a pair of iron claws are fixed,and attached to the wire rope of the donkey engine. A signal is then given bywire from the men in charge of the log to the engineer, who commences windingup the rope, and with frequent stoppages caused by the log being jammed amongstumps and other obstructions, it is at last dragged either to a prepared skid road,where another donkey engine hauls it to the loading point, or direct to wherethe trucks are able to load it. The loading is managed by building a rollingstage of heavy timbers down which the logs can be slid, or up which they arerolled by a donkey engine on to the trucks. Sometimes a dam is built and a pondformed, into which all the logs are dragged and rolled out on to the trucks. Infact there is no end to the ingenuity of the logging contractor in devising mechanicalmeans for handling these great logs, often 4, 5, and 6 feet in diameter, with the leastexpense and trouble. Many logs which to an inexperienced eye would be thoughtvaluable, are left either because they would cost more than they are worth to getout, or because they are more or less faulty; and in all cases that I saw, the work isdone without the least regard to the younger trees, or to the future. Sometimeshalf the trees are left standing and as much is left after felling as is taken. Theprice per 1000 feet at the sawmill is the one governing idea.When the logs reach the sawmill they usually go into a pond, from which theyare hauled as required up an inclined plane to the saw bench. In the largest andmodern mills the band-saw has replaced the gang-saws formerly used, and worksat an incredible speed, saving a great deal of wood which was formerly eaten upby the saw. Some of the band-saws are double-edged; and after taking off theslabs and squaring the log, it is then converted into whatever sized lumber iswanted; the best quality being cut into vertical grained decking or flooring, 4 to6 inches wide.The ingenious arrangements by which everything in these great mills isarranged so as to save manual labour, must be seen to be appreciated. I foundmany of the men employed were Japanese, who are said to be excellent workmenand to possess both nerve and pluck.When the boards are cut, the best are sorted out and sent to the drying kilnwhere they are dried for four or six days in order to prepare them for planing,tonguing, and grooving; which is usually done in another part of the sameestablishment by machinery, before the finished wood is put in cars for transport tothe interior.Much of it now goes to the middle states, and a great deal to South Africa, China,and Australia ; but whenever very large-sized balks, masts, or piles are wanted, thePuget Sound mills are called on to fill the order, because no others in the world cansupply timbers of such great size at so cheap a rate. Logs of 24 inches square andup to 100 feet long are regularly quoted.The Douglas Flagstaff, 1 in Kew Gardens, came from Vancouver Island, and was1 Cf.Jburn. K. Hort. Soc. xiv. 452 (1892).Pseudotsuga 823presented in 1861 by Mr. Edward Stamp. It is 159 feet high, about 12 feet beingunderground, and is about 4^ feet in girth at ground-level. It weighed 4 tons 8 cwt.,and was about 250 years old. In the British Museum of Natural History there is asection cut in 1885, 7 feet 7 inches in diameter, including bark, on which 533 annualrings may be counted. There is also in the Timbers Museum at Kew a finesection, 8 feet in diameter, cut from a tree on Puget Sound.A technical report on the strength, weight, and structural value of Douglastimber is given by Halt in U.S. Burecui of Forestry, Circular No. 32 (1904), fromwhich it appears that the possibility of obtaining long and large pieces, combinedwith the exceptional strength and stiffness of the material, compared with its verymoderate weight, renders it an ideal timber for structural purposes, and durable onexposure to weather.In a report on the Forest Products of the United States for 1906 (issued MarchI9O8) 1 I find that this species now comes second in the quantity of timber produced,being only surpassed by "yellow pine," under which heading are included all thevarious pines of the south and east except white and Norway pine (P. Strobus andP. resinoso). The quantity cut in 1906 was estimated at 5 billion feet, valuedat 70 million dollars, of which the state of Washington yielded 68.5 per cent, Oregon27.2 per cent, and all the other states together less than 5 per cent. The increasein production was very rapid in the last few years, and the average value hadincreased from 8.67 dollars per 1000 in 1899 to 14.20 dollars in 1906.I am informed by Mr. R. S. Kellogg of the United States Forestry Bureau,Washington, that on the Pacific coast all masts except the smallest, and on the eastcoast the largest masts, are made of Douglas fir, which is transported overland fromthe Pacific coast.It is the opinion of Lieut.-Commander Williams of the Bureau of Constructionand Repair, U.S. Navy Dept., that there is practically no difference in the strengthof Douglas fir and long-leaf pine (P. pahistris) ; the latter, however, is considerablyheavier. This appears to be now generally recognised by yacht - builders inEurope who use Douglas fir in preference to any other timber for the masts ofracing yachts.A letter on the timber of this tree in Gardeners Chronicle, 1 862, p. 452, givesthe results of experiments made at Cherbourg by M. Serres on twelve specimensof squared mast timber sent from Vancouver, which showed that in strength itwas almost equal to Florida pitch pine, and stronger than Baltic or Canadianpine. The weight of a compound mast made up of pitch pine in the centre andBaltic or Canada pine on the outside was about 12,200 kilos., whilst a solidmast of the same dimensions, made of Douglas fir, weighed only 8900 kilos. Thecost of material and workmanship of the latter was very much less.Mayr's comparison 2 of the wood as grown in various parts of Europe, withthat grown in America, and also with that of silver fir, spruce, and larch, is wellworth studying; but the age of the trees was insufficient to make the comparisonconclusive.1 U.S. Dept. Agr. Forest Service Bull. 7 7. 2 Fremdland. Wald- u. Parkbaume, 399, 400 (1906).

824 The Trees of Great Britain and IrelandWith regard to the future of the Douglas fir forests it is very hard to sayto what extent or for what period the present supply will last. Axe and fire arecertainly destroying them at a great rate, but the reproduction all over the coastregion is so good, and the danger of fire in dense young growths of trees sosmall, that many places cleared twenty to forty years ago are already coveredwith healthy young trees; and though the size and quality of these will probablynever equal those of the virgin forests, yet there is no reason why, withreasonable care, the forests should be devastated as they are now. On the driermountains of the interior, the danger of destruction is greater; and it seems tome that whilst Douglas pine is the dominant tree of the coast region, Pinusponderosa, owing to its thicker bark and greater adaptability to dry soils andclimates, will replace it in the interior.INTRODUCTIONThe Douglas fir was discovered by Menzies at Nootka Sound in 1797. Seedswere, however, first sent home by Douglas in 1827, from which plants were raisedby the Horticultural Society of London and distributed throughout the country.According to a note by Mr. Frost l the tree at Dropmore, which is usuallyconsidered to be the oldest in England, was raised from seed sown by himselfin the winter of 1827-1828.CULTIVATIONThe best account 2 of the cultivation of Douglas fir yet written is by Mr. Crozier,forester on the Durris estate in Kincardineshire, who has paid special attention tothis tree, and is one of the most experienced foresters in Scotland. Heprefers to collect home-grown seed, and considers that much may be done toimprove the type of the tree commercially, by selection of the best varieties as seedbearers; and states that the production of seed in good years is enormous, no less than15,000 cones having been counted on an outlying specimen tree about 40 years old.The seed ripens about the beginning of October, when the cones should begathered without delay before the seed escapes. After storage in a dry loft throughthe winter, the cones are exposed to sun heat, which causes them to shed the seed.In the beginning of May the seed is sown in beds 3 to 3^ feet wide, one pound beingallowed to every 8 or 10 yards. The seedlings are transplanted at two years old, andMr. Crozier prefers to plant them out in the month of April one year later bynotching, or if the ground is liable to be covered with bracken or herbage,by pitting in plants a year older.So far as my own observations go this tree will not grow well on clay or onthe oolite formation, but it thrives on greensand, and on sandstone of the Llandoverygroup at Tortworth. If desired to grow to a large size, it should be planted in awell-sheltered situation, where the soil is of sufficient depth and fertility to keep thetrees growing for a long period, but in exposed situations the tops are ruined by the1 Card. Chron. 1 871, p. 1360. 2 Trans. Roy. Scot. Arb. Soc. xxi. 3 1 (1908).Pseudotsuga 825wind. All attempts to grow this tree into timber on bare, exposed, or barrendowns and hillsides will, I believe, prove futile. 1The Colorado or glaucous variety has been so much spoken of, and is recognisedso universally in cultivation as a distinct form, that we must speak of its peculiaritiesin full. It is usually supposed to be known by its colour, which is variable in allraces of the tree; and I know of a case in which colour alone was considered bya forestry expert, to be sufficient to condemn as seed-bearing parents, a largenumber of vigorous healthy trees of great size, which were certainly of Pacific coastorigin.The Rocky Mountain forms, of which the Colorado one may be taken as typical,are constitutionally able to endure a continental climate ; namely, one characterised byextremes of summer heat and winter cold; whilst the coast form is less hardy,though it will endure the extremes of climate in most parts of Great Britain, andis a very much larger, faster-growing, and, from a forester's point of view, morevaluable tree.They are at Colesborne equally liable to suffer from late spring frost aftergrowth has commenced; but Mayr, whose experience of both is considerable, saysthat the Colorado form in Germany, does not suffer like the other, from thefreezing of the immature shoots in autumn and early winter; and wherever this is acommon cause of injury to the coast form, the mountain form should be tried instead.Such places, however, are rare in England; and on this subject I cannot do betterthan quote the opinion of Mr. Crozier. In a letter to me he says, "That there aretwo well-defined forms no one with practical experience of the tree will deny, butwhether that known as ' Colorado' is confined to the state of that name seemsdoubtful. As a timber tree, however, my experience convinces me that in thenorth of Scotland at least it is a failure, and whatever advantages it may possessover the Oregon variety in its nursery stages, is really of no moment, as after atrial of between thirty and forty years, under the most favourable conditions of soil,shelter, etc., it has failed to make timber on this estate; while the Oregon variety,under much less favourable conditions, has never failed to make good headway.The cone 2 also differs from that of the Oregon variety in some important respects,being much smaller, with the bracts a great deal longer and reflexed."" I made a further experiment with this tree some years ago, and may give youthe dimensions of average specimens at the present time of Oregon and ColoradoDouglas and Norway spruce, grown under exactly similar conditions side by side.The age of the Colorado Douglas and Norway spruce is twelve years, while theOregon Douglas is ten years from sowing.Height.Three last years' growth. Girth at 6 inches high.Oregon Douglas 15 feet 6 inches 8 feet 10 inches 9^ inchesColorado Douglas 10 ,, n ,, 5 5 6^Spruce 8 10 ,, 4-5 » 5i '.1 According to Mr. Bean, in Kew Bulletin, 1 906, p. 268, this species is used as a hedge plant at Monzie Castle, andanswers the purpose very well, heing dense and well-furnished.2 The cones on cultivated trees are very variahle. Cf. Card. Chron. xxviii. 12 (1900).IVQ

826 The Trees of Great Britain and IrelandThese have been planted seven years, and though for a time the Colorado heldtheir own with the spruce they are now being left behind."" We have raised some millions of the Oregon variety and find it sufficientlyhardy for all practical purposes. It does frequently make a second growth in thenursery stages, and these may be killed back, but the damage done in this respect isnot serious. After being planted out and established in the plantation, they arecapable of bearing a greater degree of exposure than the Norway spruce, and maybe seen on the lower spurs of the Grampians easily beating the latter. In thetreeless district of Buchan it does not do well, but neither does any other tree;but for general planting in Scotland, and with ordinary precautions, it is quitevaluable."" A member of an old firm of nurserymen informed me that it is about fifteenyears since the Colorado variety first began to be sent to this country in quantity,and they only found out the mistake after the seedlings came up. To speak ofthe Colorado as ' glaucous' and the Oregon as the green variety would beincorrect, as both vary in colour. The Colorado may be found of all shades fromgreen to a rich glaucous, while the Oregon runs from a dark bluish tint to a lightgreen."A most striking instance of the different rate of growth of the two trees may beseen in Dr. Watney's avenue at Buckhold in Berkshire, where Oregon Douglasabout 3^ feet high were planted in the winter of 1882-83, in trenched ground ona gravelly soil with some clay, underlaid at a depth of 10 to 12 feet by chalk. Fiveof the best of these average in 1908 59 feet 8 inches in height by 4 feet in girth. Thelargest was 65 feet by 5 feet 3 inches, showing 2^ feet of annual height increase fortwenty-four years. Colorado Douglas (so-called) planted on the same land at thesame time, were, when I saw them, not above half this size.In the Great Bear plantation, on the same estate, planted October 1895, andsteam cultivated 15 inches deep, Dr. Watney has measured six average ColoradoDouglas, planted about 3 feet high, now 13 feet by 6§ inches ; six average Scotspine, planted about i^ feet high, now 18 feet by 12 inches ; six average larch, plantedabout 2 feet high, now 19 feet 7 inches by 9^ inches. According to his experiencethe Colorado have many small branches which extend but a short distance from thestem, whilst the Oregon are distinguished by wide-spreading branches set muchfarther apart on the stem. He says that the latter is the fastest-growing tree heknows, whilst the former is probably the most useless of all the common conifers hehas grown; and yet he is told by a leading nurseryman that about one-third of theseed he buys produces plants which are apparently of the Colorado variety. Thesetrees are sold and planted somewhere, to the great ultimate loss and disappointmentof the unwary planter.The Douglas fir is usually healthy and little liable to insect or fungus attacks.However, of late years, a fungus, Botrytis Douglasii, Tubeuf, which is known asthe Douglas fir blight,1 has caused considerable danger to young trees growing innurseries. The leaves, especially those on the upper shoots, wilt and fall off; and1 Fisher, Schlich's Man. Forestry, iv. 461 (1907).Pseudotsuga 827the plants frequently die. There is an illustrated article on this fungus in the Journalof the Board of Agriculture for June 1903.I am informed by Capt. the Hon. R. Coke that in January 1907 there wasa bad attack of this fungus, on two-year-old plants in the nurseries at Weasenham,Norfolk, and on some trees of the same age which were planted out in theprevious autumn. He was advised at Kew to burn all the affected plants, andspray the remainder with "Violet Mixture." 1 About 25 per cent of the infectedplants died or were removed as worthless; the remainder outgrew the disease, andare now (June 1908) looking well, though the fungus has not entirely disappeared.Capt. Coke adds that, after trying the so-called Colorado variety, he will plantno more of them; and that as seedlings of the Oregon variety vary a gooddeal, he prefers those which show a tendency to stop growing in time to ripen theirleader.The seeds are liable to be destroyed by the larva of an insect,2 Megastigmusspermotropkus, which has been introduced into Europe from Oregon. The eggsare laid by the insect in the young cones, and one larva develops in each seed anddestroys it. This pest has been observed at Mariabrunn, and has done greatdamage in Denmark ; and during 1905 and 1906 was so serious at Durris in Aberdeenshire,that no seed was worth collecting there.REMARKABLE TREES AND PLANTATIONSThe largest tree that we have heard of in Europe, is at Eggesford, in Devonshire (Plate 229). This tree must be as old as any existing, for it was reported 8 in1865 to be then about forty years old and too feet high. This, however, was anexaggeration, as three years later it was recorded 4 by Mr. A. Spreadbury, as being93 feet by 12 feet at three feet from the ground. I measured it carefully in company with Mr. Asprey, agent to the Earl of Portsmouth, in April 1908, and foundit, by the mean of two measurements from opposite sides, to be 128 feet by 18^ feet.About four feet from the ground two very large spreading branches come off, which attwo feet from the trunk are 6 feet 9 inches and 5 feet in girth. At 30 feet fromthe ground, the stem is still 13 feet 5 inches round, and at 100 feet it girths 3 feet3 inches; so that it must contain about 700 feet of timber. It grows on a lawnfacing east, a little above the river Exe, on a soil which is evidently deep andfertile; and if the top is not broken may become a much larger tree, though it hasonly increased 35 feet in height in forty years.The largest tree in the grounds at Endsleigh was reported by Mr. R. G. Forbesto be, in 1906, 100 feet high, with a quarter-girth of 26 inches in the middle; butin remeasuring it by climbing in 1908, he informs me that it is only measurable to aheight of 87 feet. The quarter-girth over bark at 43^- feet is 26^ inches. Allowing1 This is composed of sulphate of copper, 2 Ibs. ; carbonate of copper, 3 Ibs.; permanganate of potash, 3 oz.; soft soap,\ Ib.; rain water, 18 gallons.2 Cf. Card. Chron. xxxix. 57 (1906), Traits. R. Scot. Arb. Soc. xix. 52 (1906), and Jonrn. Board Agriculture, xii. 615(1906), where an article on the insect with figures is given by Dr. R. Stewart MacDougall.3 Trams. Scot. Arb. Soc. iii. 80. * Card. Chron. 1 868, p. 1189.

828 The Trees of Great Britain and Ireland2^ inches for bark, its contents are therefore 87 feet by 24^ inches, making 355 cubicfeet, instead of 469, as stated in Quarterly Journal of Forestry, i. 107 (1907).At the time of the Conifer Conference in 1891, a tree 1 at Dropmore was statedto have been then 120 feet by 11 feet; but I measured this tree in 1905 and couldnot make it more than 107 feet by i r^ feet, a considerable part of the top havingbeen, as I was told, broken off by the wind. I measured it again carefully in June1908, when it was no feet by 12 feet in girth.At VValcot there is a very large tree planted in 1842, of which the Earl of Powisgave me a series of measurements. The first taken in 1860 was 74 feet by 7 feet ;the second in 1872, 85 feet by 8 feet 10 inches; the third in 1892, 107 feet by 12 feet9 inches; the fourth in April 1906, taken with a theodolite, was given as 122 feet by15 feet 6 inches; all the girths taken at 4 feet. The cubic contents were 393 feet.[ measured what I believed to be the same tree carefully from both sides, in March1906, and made it 114 feet by 14 feet 2 inches at 5 feet, and noticed that the tophad been somewhat broken. Thus it is evident in both these trees that after they hadattained about sixty years old, the height increased much more slowly.There are two trees at Powis Castle, one of which on the rabbit bank,near the park gate, I made from 112 to 115 feet by n feet 10 inches (this is themean of two measurements from opposite sides as the tree ~ leans a good deal), andthe other in a thick plantation, close to a pond, which, though I cannot, owing toits position, be confident that it is over 130 feet, may be 5 feet or more higher, andis more likely to increase in height than any Douglas fir that I have seen. It isonly 9 feet 5 inches in girth and quite the finest timber tree of the sort I know inEngland. 8There is a tree at Highclere which, in 1903, was about 100 feet by 13 feet 8inches. At Barton, Suffolk, a tree planted in 1831 measured in 1904, 107 feet highby 10 feet i inch in girth, and, while beautifully clothed to the base, was rather thinat the top with a divided leader. At Bury Hill, near Dorking, are perhaps the oldestand largest trees in Surrey, which, as Mr. R. Barclay told me, were planted by hisfather about 1832. The largest in 1908 measured 104 feet by 12 feet 2 inches, andappeared to have lost its leader recently.At Albury, Sussex, there are two trees, which in 1904 measured 95 feet by 6 feet2 inches and 82 feet by 8 feet 3 inches. At Cassiobury, Herts, there is another,which, according to the label, was planted in 1830, and had attained in 1904,99 feet in height and 11 feet 3 inches in girth. This has now lost its leader,and has remarkably pendent branches, with leaves conspicuously white1 The tree at Dropmore, raised from seed, sown in the mid-winter of 1827, was planted out in 1829, and has shownthe following growth :Measured 1837 1843 1846 1851 1853 1860 1862 1867 1868 1871Height in feet 18 40 48^ 62^ 65 78 85 93 95 100Cf. Card. Chron. 1 843, p. 808 ; 1846, p. 661 ; 1851, p. 246 ; 1853, P- 343 > ' 860, p. 854 ; 1867, p. 808 ; 1868, p. 465 ;1871, p. 1360.a Lord Powis had this tree measured in 1908, L>y a man climbing, as 127 feet by 12 feet i inch. I cannot account forthe difference.3 I measured this tree again in July 1908, and having found a spot from which I could see the top, am confident thatit is more than 130 feet high. It had increased 5 inches in girth in two years.Pseudotsuga 829beneath. At Fulmodestone, Norfolk, Sir Hugh Beevor measured, in 1904, a tree98 feet high by 8^ feet in girth. Henry saw a tree there in 1905, which was 82 feetby 9 feet 4 inches.Many other trees which approach if they do not exceed 100 feet in height, maybe found in the southern and western counties.At Endsleigh, in Devonshire, which was visited by the English ArboriculturalSociety 1 in August 1906, there is a very fine plantation of Douglas fir in GunoakWood, of which careful measurements were made by Mr. R. G. Forbes, forester tothe Duke of Bedford, in November 1906, from which it appears that the threelargest trees in this plantation measure as follows:No. 16. 120 feet high by n inches quarter-girth = 100 cubic feet.No. 23. 100 13 = II7No. 30. no ., 13 = I29Mr. E. C. Rundle, agent for the property, writes to me as follows : " The forestersays that the trees must not be taken as a full crop, for there is space on the £ acrefor forty trees instead of thirty-two. As to their age I believe they must be overfifty years, probably fifty-five, though an old man remembers their being planted.The quarter-girth was taken over bark at half the length of the tree, and an inch tothe foot would be sufficient allowance. They are growing in an exposed position,but in the middle of a wood on high ground, and the soil is not at all good." Thetotal contents of the thirty-two trees is 2857 cubic feet, an average of rather over89 feet per tree. If 357 feet is deducted from the total for bark and small tops, itwill leave a result of 10,000 feet per acre.At Woburn, in a plantation called "The Evergreens," on a very light sandysoil, Mr. Mitchell, forester to the Duke of Bedford, showed me a plantation made in1882, well sheltered by surrounding trees, and wrote me the following particulars :"The number of trees planted was 160, of which 132 are now left. I thinnedthem a few years ago, taking out only dead and suppressed trees. The area of landis as nearly as possible 2 chains square, and includes a few old Scotch and spruce fir.I measured the trees in three classes, as follows :72 trees : 50 feet by 6 inches quarter-girth = 900 cubic feet.4 .. 55 .. 6i .. > 64520 5 .. 4 n ii inDeduct for bark at 8 per centTotal contents of timber .16561361520This works out at 3800 cubic feet per acre at twenty-six years after planting," saythirty years from seed. I may add that though these trees were planted close enoughto kill all their lower branches, yet none of these had fallen, and a good many of thestems showed the same want of straightness which is so often evident in this1 Quarterly Journal ofForestry, \. 64 (1907).

830 The Trees of Great Britain and Irelandspecies in England, and which I attribute to the unripeness of their sappy leadersbefore winter.In 1904 the Earl of Ducie showed me a plantation of Douglas fir on a steepbank called Ironmill Wood near Tortworth, which, though of insufficient area togive the best results, is a good illustration of the growth of this tree on sandstone (Plate 231). The plantation was made in 1868, and was therefore thirty-sixyears old when I saw it. The area, as measured by Mr. Harle, agent to LordDucie, was i acre 28 perches ; the number of trees standing was 238; their averageheight was about 80 feet; and their average cubic contents I estimated at slightlyover 20 feet, making a total of about 5000 feet per acre. Mr. A. P. Grenfell, whovisited the same place in the same year, made a more careful estimate based on themeasurements of the trees standing on ^ of an acre, and came to the conclusionthat the total volume, with allowance for bark, was 5250 feet, which gives an annualaverage increase of 150 cubic feet per acre, no allowance being made for thinnings.Mr. G. F. Luttrell of Dunster Castle, Somersetshire, showed me, in August1906, a plantation of Douglas fir which he made in 1880 on a piece of waste land,which was growing only furze, on gravelly soil close to the rock, which is on the OldRed Sandstone formation. In the following December he had this carefullymeasured, with the following result: Broom Ball Wood, area 3 roods 10 perches,planted entirely with Douglas fir at about 10 feet apart. Number of trees nowstanding, 264. Total contents, allowing half an inch for bark, 2491 cubic feet. Ofthese, 158 trees contain less than 10 cubic feet each, and only 7 contain above 20cubic feet, the largest tree measuring 42 feet timber length and 10 inches quartergirth,equal to a volume of 29 cubic feet. The actual height of the tallest was 73feet, of the shortest 48 feet.The trees are valued as timber by Mr. Luttrell's forester at 6d. a foot, whichamounts on the estimated quantity to £62:65., equal to ^76:13:6 per acre.Deducting from this sum, the expenses of planting and fencing, £6 an acre in 1880,equivalent in 1906, at 4 per cent, interest, to . . £ 16 12 6and the annual deferred rent at 53. an acre, from 1880 to 1906,equivalent to . . . . . . 11 o othe balance, ^49 : is., represents the actual profit per acre^27 12 649 i o£76 1 3 6It seems to me that the price of 6d. per foot for trees of this size is somewhatexcessive, as those of less than 10 feet are hardly fit for anything but pitwood orrough fencing; but the value of the trees over 10 feet might be somewhat higher.From the appearance of this plantation, in which many of the smaller trees werealready suppressed and not likely to grow much more, it seemed to me that either aheavy thinning or clean felling was the proper thing to do, but this must depend onthe local demand for timber of this size and quality. And if the small area,exposed position, and inferior agricultural quality of the land be taken intoPseud otsuga 831consideration, there can be no doubt that this has been an unusually profitableinvestment, and one which would fully justify planting Douglas fir on a largescale in this district. Mr. Luttrell states that where there is sufficient room and lightthe trees reproduce freely from seed.In Scotland there are many Douglas firs exceeding 100 feet in height, butwe cannot say which is actually the largest; and if we did, it would not holdgood for many years to come. The tallest recorded at the Conifer Conference* in1891 was at Lynedoch, on Lord Mansfield's property in Perthshire, which was thenreported to be 92 feet by 12 feet, but had a fork at 60 feet from the ground 2(Plate 230). Another tree at the same place, is the parent of the seedlings plantedat Scone and Taymount, and was only 72^ feet by n feet 2 inches, though plantedin 1834.One of the oldest trees is in the grounds of Scone Palace, and bears theinscription "raised from the first seed, brought by David Douglas in 1827,planted 1834." In 1850 it was transplanted to its present position, and this hasdoubtless checked its growth. It measured, in 1904, 96 feet high by 10 feet in girth.Its foliage is conspicuously white beneath.At Drumlanrig, in Dumfriesshire, there is also an original tree, which was sentby David Douglas to his brother, who was clerk of works at Drumlanrig about 1832.It is growing in shallow gravelly soil near the top of a hill, overlooking a glen, andin 1904 measured 90 feet high by n feet 4 inches in girth.Mr. R. Macleod of Cadboll sends me the measurements of four trees taken in1907 by Mr. C. E. Cranstoun at Corehouse, near Lanark, as follows :No. i.2... 3-4-Height.70^ feet.83 -85 -92Girth at 5 feet.12 feet 5 inches.10127 ..He adds that these were raised from the first seed sent to Scotland by Douglas;and that he finds by repeated measurements of several trees, that their rate of girthincrease is about 2 inches per annum.At Durris, in Kincardineshire, the original and largest tree, planted aboutseventy-two years ago, has now reached 114 feet by \ 2\ feet, and contains over 300feet of timber. At Buchanan Castle, Stirlingshire, Mr. Renwick measured 3 in 1900an original tree 85 feet by 13 feet 2 inches. He informs us that the girth in 1908 is14 feet 2^ inches.At Murthly Castle there are probably more large trees of this species thananywhere in Scotland, the plantation below the castle being especially fine, and alsothe avenue called the Dolphin Walk, where the trees, planted about 8 yards apart,1 Joarn. Roy. Hart. Sec. xiv. 537 (1892).2 An accurate measurement of this tree, made in January 1908 by Mr. A. T. Kinnear, makes it 108 feet high by 13 feet9 inches at 5 feet. The main stem up to the fork, 60 feet from the ground, contains 415 feet, and the two tops together, 48feet, making the whole 463 cubic feet. 3 Trams. Nat. Hist. Soc. Glasgow, vi. 256 (1900).746

832 The Trees of Great Britain and Irelandaverage about 90 feet by 8 feet, and grow at the foot of a bank, in deep sand withpebbles in it, which looks like an old bank of the Tay, which is not far off. In TheGarden for igth May 1900, some particulars are given of the trees here. One,planted in 1847, measured on nth August 1892, 86^ feet by 8 feet 10 inches, andon 24th March 1900, 97 feet 4 inches by 9 feet 10 inches. A great many others wereof about the same size. This proves the diminishing rate of increase, both in heightand girth, after forty to fifty years' growth, even when the lower branches remain.Mr. Fothringham states that all these measurements were taken by sending menor boys up the trees, and not with a dendrometer. He adds that the temperature 1in February 1895 was for several days below zero, and on one night went downto - 11 .There is probably no plantation in Great Britain about which so much hasbeen written as the Taymount plantation on the estate of the Earl of Mansfield, inPerthshire. It lies about seven miles north of Perth, one mile from Stanley Station,and may be seen from the Highland Railway, which passes close to the east of it.The plantation covers eight acres of flat land, which is locally known as "till," twofeet of light loam over red clay, and which may be worth for agricultural purposesI2S. to 155. per acre. This plantation was first fully described in the GardenersChronicle of loth, i7th, and 24th November 1888, by Dr. Schlich, than whom therecan be no higher authority. It was planted by the late W. M'Corquodale in thespring of 1860, with Douglas firs, two-year seedlings, two years transplanted, at9 feet by 9 feet apart, with larch four years old, between every two firs, and anadditional line of larch between every two rows, so that the trees stood 4^ feet apart,and each acre contained 538 Douglas and 1613 larch. The latter were graduallythinned out, and were all removed by 1880. The first thinning of Douglas tookplace in 1887, when about half the trees had already disappeared, 277 per acre onlyremaining. Of these 75 per acre were cut, leaving 202 per acre.Dr. Schlich made a careful estimate of a sample plot measuring ^ of anacre of average appearance, and had a tree felled to ascertain its actual contents ;and from these data came to the conclusion that the total per acre was 3738 cubicfeet of wood over 3 inches diameter, exclusive of top and branches, which gives anannual increment of 133 cubic feet per acre. But this estimate being the grossvolume, when reduced by about one-fourth, makes the quarter-girth measurement, asadopted in English practice, to be 2934 cubic feet.After inspecting a sample area of Scots pine in the same district, Dr. Schlichgoes on to say, " If grown in a well-stocked, overcrowded wood, and in localitiesof equal quality, Douglas fir is not likely to produce more solid wood, during thefirst thirty or forty years, than the larch, and probably also not more than Scotchpine." He then goes into careful estimates of the probable future increase of theDouglas, based on data taken from America, where Dr. Mayr found that in the mostfavourable localities in the Cascade Mountains the average height of mature DouglasAt Balmoral, where there are 25,000 to 30,000 trees, planted in the 'eighties, on a northern aspect at looo to 1200feet altitude, Mr. Michie informs me in a letter that this severe frost, when the temperature fell to - 17^ , did no harm to theDouglas fir.Pseudotsuga 833firs on the best soil was 2 1 3 feet, with a diameter of 6j feet, whilst in Montana itonly reached an average height of 148 feet, with a diameter of 2.6 feet, thus showingwhat an immense influence the soil and rainfall have on the growth of this tree.From a cross section of Douglas fir grown in Washington and then in themuseum at Cooper's Hill, Dr. Schlich remarks " that the rate of growth indicatedin this section, up to thirty years old, resembles that of an average tree in theTaymount plantation in a striking degree, as follows : diameter of average tree atTaymount at 4^ feet, 1 2 inches ; diameter of thirty years' growth on the section fromAmerica, 11.9 inches.After visiting a second growth area of pure Douglas fir on Ladds farm, aboutfour miles from Portland, Oregon, which was believed to be of about fifty years'growth ; I came to precisely the same conclusion, and though I had not then seenDr. Schlich's article, I wrote in my journal at the time, that the trees in Oregonwere very similar in density to those at Taymount, but decidedly cleaner and bettergrown, and having regard to their greater age and better soil, they might average100 feet by 4 feet, and I estimated their cubic contents at something like 6000 feetper acre.When I first visited Taymount, in April 1904, I determined to estimate it formyself, without regard to what others had done. I therefore paced an area of 100yards long by 50 yards wide in what I thought a fair average of the whole plantation, and found that there were on it ninety-nine trees of the first size, and fifty treesof the second. I did not reckon a number of other trees, which were so small,crooked, or poor, that they could not have been sold profitably with the better ones ;and, judging from a fallen tree which I was able to measure accurately, which was55 feet long by 10 inches quarter-girth, equal to 3 8 cubic feet, came to the conclusionthat the total volume of saleable timber at forty-four years after planting, or fortyeightyears from the seed, did not much exceed 5000 feet per acre.Sir Hugh Beevor visited Taymount in the autumn after I was there, and madean estimate in a different way by taking three different areas of £ acre each, andmeasuring everything on those areas. He found 96 trees of 12 inches quarter-girthand upwards at six feet from ground; 44 of 10 and n inches; 44 of below 10inches ; and estimated the total contents per acre at 6226 cubic feet.I revisited Taymount in September 1906 in order to compare it again withwhat I had seen since in America and in England. I measured twenty trees inthe fifth row and twenty in the tenth row from the bank on the east side of theplantation nearest to the high road. I found that their average girth over bark at5 feet was slightly under 4 feet, the largest being 7 feet 10 inches and the smallest2 feet 3 inches. I estimated the average timber length of these trees at 60 feet,and the quarter-girth, under bark at half this length, at 8 inches. If this isapproximately correct, their average contents would be 26 feet 8 inches, and theirtotal per acre something like 5400 feet, which very closely agrees with my previousestimate, allowing for the increase of two years.A very different estimate was made by Dr. Somerville in a paper on " ExoticConifers in Britain," which was printed in the Journal of the Board of Agricul-IV

834 The Trees of Great Britain and Irelandture, December 1903, and of which an abstract appeared in Transactions ofRoyal Scottish Arboricultural Society, xvii. 269. This was based on measurementsmade in June 1903, by the late Mr. Pitcaithley, forester to the late Earl of Mansfield,who selected two typical areas of ^ acre each, on which he counted and measuredthe trees, of which he found eighteen on one and twenty-five on the other area, andaccurately measured the cubic contents of two trees, one of which contained 46.76cubic feet and the other 39.49 cubic feet measured down to 3 inches diameter.Dr. Somerville, assuming Mr. Pitcaithley's measured trees to be average ones,brings out the total cubic contents per acre by quarter-girth measure as 7977 cubicfeet, and comparing this with Dr. Schlich's estimate of 2934 cubic feet made fifteenyears previously, comes to the conclusion that the average increase per acre in thatperiod was no less than 336 cubic feet per annum.This in my opinion is a mistaken calculation, and if compared with the annualincrement of 150 feet per acre per annum in Lord Ducie's plantation and the resultsof the measurement of Gunoak wood, both on better land than that at Taymount,we must hesitate to accept it as even approximately correct. 1The important point to consider is how long these trees will continue to maintain their rapidity of growth, and what will be the value of the timber ? My ownbelief is that they fall off in their rate of increase; that the larger ones willcontinue to suppress and starve out the weaker ones, as they have already doneto a great extent; and that the timber of Douglas fir grown in the country willnever compare in quality or value with the imported timber, which, it must always beremembered, is from 200 to 300 years old, and selected both in the forest and themill from a very much larger quantity.Dr. Schlich writes me as follows: " As to the quality of the Douglas firtimber, I merely quoted what the late Mr. M'Corquodale told me. Since then Ihave paid some attention to the subject and noticed that in timber from youngDouglas firs there is a considerable difference between spring wood and summerwood; hence I am sure, and in this I agree with you, that only trees of considerableage will yield timber equal in quality to that of larch, if at all."There are other causes, which tend to make the production of clean, straighttimber difficult, in many situations and on many soils in this country, and whichshould be considered by all who contemplate planting this species largely for profit.The first is its tendency to form large and spreading branches, which it showsin a very marked degree. In order to prevent this, the trees must be crowded to anextent which is only possible with success on soils of unusual depth, or on slopescomposed of rock which is sufficiently disintegrated to allow the roots to penetratedeeply; in which case they may clean themselves when they attain a height of 60to 80 feet; though I have never seen any in England which have naturally cleaned1 After this was in print I sent it to the Earl of Mansfield for his opinion, and am informed that in 1908 a carefulmeasurement was made by his forester, Mr. A. T. Kinnear, of the whole of Taymount plantation, which now contains 1536Douglas firs on the whole area= 192 trees to the acre. These contain 51,456 cubic feet (under bark) or 6432 cubic feet peracre, being an increase of 3498 cubic feet per acre since it was measured by Dr. Schlich in 1888, equal to about 134 feet peracre per annum since it was planted, the rate of increase from 1888 to 1908 being about 175 feet. The largest tree is 93 feethigh, and contains 118 cubic feet.Pseudotsuga 835their trunks. In default of these conditions recourse must be had to pruning, whichentails considerable expense, and must be repeated at frequent intervals.The second is the tendency which I have observed in so many places for thetrees to ripen their leading shoot prematurely in dry summers, and to make a freshstart in the autumn when wet warm weather sets in. The result is that the secondshoot is weak, immature, and usually becomes crooked either from frost, wind, or thesettling of birds on it. A double lead is then often produced, and the result is seenin many plantations, in the more or less crooked stems, 1 or in forks, which mustseriously depreciate the value of the timber when brought to the sawmill.A third is the effect of gales on the leading shoots, which owing to their greatlength and weakness, seems greater than on any other conifer, especially as owingto the rapid growth of the tree it overtops other species with which it may bemixed. Even if the tops are not broken they become crooked, and often forked, inplaces exposed to wind, and the taller the trees become the more they are liable tothis source of injury.For these reasons it seems to me that the most profitable way of utilisingDouglas fir, is to cut it at a comparatively early age, and utilise the wood for pittimber and estate purposes, for which purposes I am disposed to class it as superiorto spruce or silver fir and inferior to larch.In Ireland the Douglas fir grows very fast, and has attained in many places alarge size. The late Lord Powerscourt planted at Powerscourt in 1865, with his ownhands a tree which measured in 1904 100 feet in height and 9^ feet in girth.There are good trees at Fota, Queenstown, 84 feet by g\ feet in 1903 ; at Carton,81 feet by 7^ feet in the same year; at Stradbally Hall, Queen's County, 86 feet by8 feet 3 inches in 1907 ; at Coollattin, Wicklow, 85 feet by 9 feet in 1906. AtCoollattin there are a few natural seedlings, 2 and several trees bear cones profusely;but the forester has not been able to raise plants from their seeds, doubtless owingto the cones being attacked by the insect which has done so much damage at Durrisin Kincardineshire. At Castlewellan, Co. Down, there are fine trees, about 80 feetin height, which I could not measure on account of heavy rain when I was there in1908. One measured by the Earl of Annesley in August 1908 is 79 feet by 10 feet,but lost 12 feet of its top in a gale in 1902.The late Mr. John Booth of Berlin was a great admirer of this tree, and for manyyears advocated its planting in Germany, where it is now beginning to be lookedupon as one of the most valuable forest trees. The result 3 of an experiment madeby the late Prince Bismarck, on his estates at Sachsenwald near Hamburg, was sentme by Mr. Booth just before his death, and may be summarised as follows :An area of 1.16 acre, the soil being a coarse, somewhat loamy, diluvial sand,was planted in 1881, half with four-year old Douglas, 5 feet apart, and halfwith spruce, 4 feet apart. In 1906, the Douglas plot consisted of 869 trees,measuring 3300 cubic feet of timber; while the spruce plot, 1335 trees, only1 This defect is clearly seen in the Taymount plantation.2 Natural seedlings were seen by Henry, also at Dereen and at Powerscourt.3 Published in detail in Zeitsch? iff fiir Font- undJagdviesen, 1 906, p. 8, of which a translation appeared in Tram: Kny.Scot. Arbor. Soc. xx. 104 (1907).

836 The Trees of Great Britain and Irelandmeasured 1700 cubic feet. The market value of the timber, which could be used forpoles and pit-props, worked out at about 1000 marks for the Douglas, and about 360marks for the spruce. Thus, growing on the same soil, the Douglas, as comparedwith the spruce, had yielded about twice the amount of timber, with about threetimes the value. I visited this plantation in August 1908, and measured two of thelargest trees, which were 74 feet high by 3 feet 8 inches, and 2 feet 7 inches ; butthe average was considerably less. I noticed that the lower branches, though deadfor several years, were not falling off; and that many of the trees showed the sameirregularity in straightness that I have noticed elsewhere. My impression was thatunless heavily thinned, a large proportion of the trees would soon be suppressed bytheir more vigorous neighbours, and that such close planting was neither economicnor desirable.TIMBERI have said so much about the timber of this tree in its own country that it onlyremains to speak of its probable future value here, and as this subject has beenably dealt with in a recent paper by Mr. J. D. Crozier, 1 I cannot do better thansummarise his opinions.He agrees with me that we cannot hope to compete with the imported timberin size, age, or quality, and thinks that in a young state it is not so dense in fibre orso tough as larch of the same age. " For standing in contact with soil, and for suchpurposes as gate-making, fencing, etc., where the ability to stand wear and tear is adesideratum, it is inferior to larch, but there are many other purposes for which it isinfinitely superior, and for the supply of which an infinitely greater volume of timberis required. For constructive purposes of all kinds it is especially suited, and owingto the beauty of its grain and the ease with which it can be worked, it is valuable forthe finished work of interiors. The timber stains well, and when varnished, takeson and retains a beautiful gloss. Outlying and badly-grown trees, when sawn upare liable to warp, but this defect is not apparent when dealing with trees of cleanstraight growth; and with home timber more freedom may be used in regard tonailing. In a younger state it has been tried and found useful as curing-barrelstaves and headings, and for box wood, for which in this locality there is an unlimiteddemand."" What the most profitable length of rotation may be is a question which willhave to be determined by trade demands, but to provide timber of a class fittedfor house construction, any period short of 100 years need not, I feel convinced, becontemplated, and on deep rich soils, probably other ten or twenty years will requireto be added to that period.""As a pitwood tree the Douglas fir is well adapted, and is deserving of consideration wherever crops cultivated for that purpose are found to pay. Crowdedtogether in pure plantation, by the time they have reached their thirtieth year, theywill be found capable of yielding an amount of pitwood almost incredible to thosewho have not seen the tree so grown. For this purpose the planting should not bedone at more than 3 feet apart." (H. J. E.)1 Traits. Hoy. Scot. Art. See. xxi. 31 (1908).CASTANEACas/anea, Adanson, Fain. PL ii. 375 (1763); Bentham et Hooker, Gen. PI. iii. 409 (1880); Dode,Bull. Sot. Dendr. France, 1 908, p. 140.Fagus, Linnaeus, Gen. PL 2 92 (in part) (1737).Casanophorum, Necker, Elcm. Bot. iii. 257 (1790).TREES or shrubs, belonging to the order Fagacese. Bark furrowed. Buds all axillary,no true terminal bud being formed, as the tip of the branchlet falls off in early summer,leaving a small circular scar close to the uppermost axillary bud, which prolongs thebranchlet in the following year. Buds alternate, arranged on the long shoots intwo ranks ; scales numerous, two or three of which are visible externally, lowestpair lateral and each composed of two connate stipules, next pair each corresponding to a stipule and with or without a leaf-rudiment, following pairs of singlestipules each covering a young leaf; all the single stipules accrescent and marking intheir fall the base of the shoot with ring-like scars.Leaves deciduous, alternate, simple, stalked, dentate with slender glandular teeth,penninerved, each lateral nerve ending in a tooth. Stipules ovate or lanceolate,scarious, deciduous, their scars visible in winter on each side of the leaf-scars, whichshow three groups of bundle-dots, and are placed on prominent pulvini, from whichdecurrent lines descend along the branchlet.Flowers moncecious, strong-smelling, 1 fertilised by the wind, unisexual, in slenderelongated erect catkins, of which those arising in the axils of the lower leaves of thebranchlet open early and are entirely composed of male flowers, while the catkinsarising in the axils of the upper leaves are shorter and bear female flowers at theirbase and male flowers on their upper part, the latter not opening until after thefemale flowers have been fertilised. Staminate flowers three to seven, in a cyme inthe axil of a bract, and surrounded by minute bracteoles ; calyx campanulate, deeplydivided into usually six segments, stamens twice or thrice as many as the calyxlobes; filaments filiform ; anthers two-celled, dehiscing longitudinally; ovary aborted.Pistillate flowers, sessile, solitary or two to three together, placed within an involucreof closely imbricated scales, subtended by a bract and two bracteoles; calyx-tubeurn-shaped, divided above into six short lobes; ovary adnate to the calyx tube,six-celled, each cell containing two ovules, surmounted by six simple styles, whichare exserted out of the involucre.i Cf Kerner Nat Hist. Plants, Eng. trans. ii. 200 (1898), concerning the nature of ihe odour of the flowers.837

838 The Trees of Great Britain and IrelandFruit, ripening in one season, one to three nuts, ovoid, plano-convex or compressed, enclosed in an involucre, which is tomentose within and is covered externallywith branched spines fascicled between deciduous scales, the nuts escaping by theultimate splitting of the involucre above into two to four valves. Nut crowned bythe styles, marked with a scar at the base, its shell lined with tomentum. Seedusually solitary, occasionally two to three in each nut, the aborted ovules, two toeleven in number, remaining at the apex of the seed. Albumen absent. Cotyledonsthick, fleshy, undulate, sweet, farinaceous, remaining under ground on germination.The genus : is confined to the warmer parts of the northern temperate zone, andmuch difference of opinion exists as to the various forms 2 which are met with.Formerly only two species were recognised, viz. C. sativa and C.pumila; but theformer, widely spread over North America, Europe, and Asia, exists in certain wellmarkedgeographical forms, which it is convenient to treat as distinct species. Asmall shrub, occurring in North America, near the coast in the South Atlantic statesand in Louisiana and Arkansas, is considered by American botanists to be anotherdistinct species, Castanea alnifolia, Nuttall, and will not be further alluded to.Four species have been introduced into cultivation and are distinguished asfollows:I. Leaves without stellate tomentum, acute at the base.1. Castanea dentata, Borkhausen. N. America. See p. 856.Leaves tapering at the base, long acuminate at the apex, green and glabrousbeneath, pendulous. Petiole glabrous.II. Leaves with stellate tomentum, rounded or cordate at the base.2. Castanea sativa, Miller. Europe, N. Africa, Asia Minor, Caucasus, Persia.See p. 839.Leaves green beneath, always showing some trace at least of tomentum,not pendulous, coarsely serrate. Petiole and young shoots scurfy pubescent.3. Castanea crenata, Siebold et Zuccarini. China, Japan. See p. 854.Leaves green beneath, tomentum variable in quantity, shallowly andcrenately serrate, the teeth often reduced to bristle-like points. Petiole, youngshoots, and midrib densely pubescent with short hairs.4. Castaneapumila, Miller. N. America. See p. 857.Leaves silvery white and always tomentose beneath. Petiole and youngshoots strongly pubescent.1 In Castanea, the leaves are deciduous, no terminal bud is formed, and the fruits ripen in one season. In Castanopsisthe leaves are persistent, a terminal bud is present, and the fruits ripen at the end of the second season.2 Dode enumerates twelve species, some of which are alluded to in our accounts of C. crenata and C. pumila.Castanea 839CASTANEA SATIVA, SPANISH OR SWEET CHESTNUTCastanea sativa, Miller, Diet. ed. 8, No. i. (1768).Castanea vulgaris, Lamarck, Diet. i. 708 (1783); Willkomm, Forstliche Flora, 4 28 (1887);Mathieu, Flore Forestilre, 3 25 (1897).Castanea vesca, Gaertner, Fruct. i. 181, t. 37 (1788); London, Arb. et Frut. Brit. iii. 1983 (1838).Castanea Castanea, Karsten, Pharm. Med. Bot. 495 (1882).Fagus Castanea, Linnaeus, Sp. PL 997 (1753).A tree, attaining over 100 feet in height and an immense girth. Bark ofvery young stems smooth and olive green, soon becoming greyish white, afterfifteen to twenty years gradually changing into a thick brown bark, which is deeplyand longitudinally fissured. Young branchlets green, covered with a minutescattered pubescence above, and with longer hairs near the base; in the second yeargrey, glabrous.Leaves (Plate 202, Fig. n) not pendulous, oblong-lanceolate; broad, unequal,rounded and often auricled at the base ; acuminate at the apex; with about twentypairs of parallel nerves, raised on the under surface of the blade, each ending in atriangular tooth, which is prolonged into a long fine point; upper surface dark green,shining, covered with minute scattered pubescence ; lower surface lighter green, withdense appressed stellate pubescence. 1 Petiole scurfy pubescent, £ to i inch long.Stipules | inch long.Nut, variable in size, abruptly and shortly acuminate at the apex, usually threein each involucre, in wild trees.An elaborate description of the fruit is given by Lubbock.2 The cotyledons arefleshy, occupying nearly the whole of the seed, undulate, and interlocking witheach other at the margins. When sown, the pericarp, owing to the swelling of thecotyledons, splits in the soil at the apex, so that the shoot and rootlet emerge, thecotyledons remaining enclosed in the pericarp and being gradually absorbed. Thegermination thus resembles that of the oak; and the young stem similarly bearsseveral scales (two to six in number) below the primary leaves, which resemble inshape those of the adult plant and bear deciduous stipules.IDENTIFICATIONIn summer the leaves are unmistakable and can only possibly be confused withcertain species of oak, like Quercus serrata and Q. casfaneafolia, which have, however, very different buds. From the other species of the genus, it is distinguishedby the characters given in the key.In winter the following characters (Plate 200, Fig. i) are available : Twigsstout, reddish brown or olive green, shining, conspicuously angled, glabrous for themost part but showing remains of glands and pubescence towards the base, whichis conspicuously ringed by the fall of the previous season's bud-scales. Leaf-scars1 This pubescence often wears off, so that the leaves are glabrescent or even glabrous, when gathered in summer.1 Seedlings, ii. 537 (1892).

840 The Trees of Great Britain and Irelandobcordate or semicircular, with three groups of bundle-dots, and set parallel to thetwig on prominent pulvini, distichous on the long shoots. Stipule-scars long,linear. Buds ovoid, slightly rounded and not acute at the apex, those nearest theapex of the twig the largest; three scales visible externally, first scale small andshort, second scale longer, both glabrous and ciliate; third scale clothed withappressed pubescence and appearing at the apex of the bud.The twigs and buds of the chestnut resemble those of the lime tree. Thepith affords a good mark of distinction, being greenish and five-rayed in Castanea,and whitish and round in Tilia.In France, single trees have been noticed 1 in several localities, which borecatkins entirely formed of pistillate flowers. Such trees, according to Dode, 2 bear alarge quantity of fruit; but the presence in the neighbourhood of a tree withstaminate flowers is necessary for fertilisation. Mr. Lynch informs me that anisolated tree in a garden at Cambridge never bore fruit, until branches, withstaminate flowers from another tree, were laid upon it; but it is uncertain whetherthis tree bore only pistillate flowers, or whether its own pollen was ineffective.Dode also mentions 2 a tree in the department of the Loire which never bore fruit,as its catkins were entirely composed of staminate flowers.The number of seeds in the nut is also variable ; and a single chestnut withthree seeds has been known to germinate and produce three plants.1VARIETIESThe chestnut varies very little in the wild state, though the amount of pubescencewhich occurs on the leaf is remarkably different in many specimens. At the ScientificCommittee meeting of the Royal Horticultural Society on 6th November 1900, someremarkable leaves were shown, consisting of but little more than the midribs, whichhad issued from the stump of a tree that had been felled ; and it is possible that someof the narrow-leaved varieties originated in this way.Schelle 3 enumerates nineteen varieties, which have been obtained in cultivation.Seven of these are forms with variously coloured and variegated leaves, viz.argentea, marginata, argenteo-marginata, argenteo-variegata, aureo-maculata, aureomarginata,and aureo-variegata. These are sufficiently explained by their names;and of those we have seen aureo-tnarginata is the best. 4Var. keteropkylla. 5 Leaves variable in shape, some with irregularly-shaped teethand occasional deep sinuses, others repand in margin and with few teeth.Var. aspleniifolia (var. laciniatd). Leaves with long narrow teeth, ending inlong subulate points.1 Clos, in Bull. Sec. Bot. France, xiii. 96 (1866).2 Dode, in Bull. Sac. Dendr. France, 1908, p. 147. 3 Laub/iols-benenniing, 63 (1903).4 There are small trees of the silver and golden variegated forms at Aldenham which are very handsome and well worthgrowing. A curious purple-leaved variety is described on p. 852. (H. J. E.)6 At Verrieres, near Paris, there is a tree, 28 feet high and 5 feet in girth, which has a few branches with normalfoliage, all the others bearing leaves deeply and irregularly lobed. These two different kinds of branches bear fruit, whichreproduces, when sown, seedlings with the form of foliage from which the nuts have been derived. Cf. ffortus Vilmorinianus,56 (1906). There is a fine specimen of this variety at Murthly Castle ; and Mr. Renwick has sent us specimens from alarge tree at Finlayson, Renfrewshire, a few of the leaves of which are of the heterophylla type.Castanea 841Var. cochlcata. Leaves small, irregularly cut, hollow or with swellings in themiddle.Var. prolifera. Some of the leaves, usually the uppermost ones, remainingwhitish tomentose beneath.Var. glabra. Leaves thin and shining.Var. rotundifolia. Leaves small, not exceeding 2^ inches in length, oval in shape.Var. pendulifolia? Branches pendulous.Many varieties of the fruit, which are propagated by grafting, are cultivated inFrance and Italy. In France, the name marron is given to the best varieties, inwhich the fruit is large, globular, broader than long, and usually single in theinvolucre. According to de Candolle,2 the Romans in Pliny's time already distinguished eight varieties, but it is impossible to discover from the text of thisauthor whether they possessed the variety with a single kernel. Olivier de Serres sin the sixteenth century praises the chestnuts, Sardonne and Tuscane, which produced the single-kernelled fruit called the marron de Lyons. H e considered that thesevarieties came from Italy; and Targioni * states that the name marrone or maronewas employed in that country in the Middle Ages (1170). In England, the cultivation 6 of special varieties of the chestnut for its fruit is so little in vogue that it is noteven mentioned in a late and comprehensive book on fruit culture, The Fruit Garden?by Bunyard and Thomas. 7DISTRIBUTIONThe chestnut occurs wild throughout the whole of southern Europe, in Algeria,Tunis, 8 Asia Minor, the Caucasus, and northern Persia. It has not been found inthe Himalayas where there are several species of Castanopsis occasionally knownin India" as chestnuts, and is replaced in China and Japan by a closely allied species.10Its northern limit in Europe is difficult to trace with accuracy, as the originalarea of distribution has been much extended by cultivation since the time of theRomans; and it has become naturalised in many parts. According to Willkommthe northern limit runs along the edge of the Jura and is continued through Switzerland to the south Tyrol, Carinthia, Styria, and Hungary, where it reaches Pressburg1 Lavallee, Arb. Segrez. 1 13. t. 33 (1885). " Origin of Cultivated Plants, 3 53 (1886).3 Theatre de FAgric. p. 114. * Cenni Storici, p. 180.6 Hogg, in Fruit Manual, 224 (1875), says that the chestnuts produced even in the southern counties are so inferior tothose imported from Spain and the south of France, that no one would think of planting the chestnut for its fruit alone. Hementions two varieties, Devonshire Prolific and Dmviitcn, which succeed in hot seasons. Lord Ducie, however, informed SirW. Thiselton-Dyer that he had once sent a sack of chestnuts to Covent Garden market, which realised £3 ; and was askedto send more, as they were the first on the market. 6 In Country Life Libraiy, 1904.7 W. A. Taylor, in Bailey, Cycl. Amer. Hort. \. 296 (1900), enumerates and describes seventeen varieties of theEuropean chestnut which are in cultivation in the United States.8 Battandier et Traliut, Flore de fAlglrie, 8 19 (1888); wild in forests of Edough near Bone in Algeria, and in Tunisianear Ain-Drahm. Though cultivated near Tangiers and Tetuan it has not yet been found wild in Morocco. Cf. Ball, inJoani. Linn. Sec. (Bot.) xvi. 666 (1878).9 The chestnut has been planted at Bashahr, in the Punjab, where trees fifteen years old are 30 feet high and 4 feet ingirth. Kew Bull. 1 897, p. 113.10 The chestnut has been erroneously supposed, mainly on philological grounds, not to be a native of E',urope, but to havebeen introduced at an early period from Asia Minor. The best discussion on this subject is by De Candolle, in Geog. Bot. ii.688 (1855). A learned paper on the classical names of the oak and chestnut by H. L. Long appeared in Loudon, Card. Mag.1839, pp. 9-20. Dr. Bettelini's excellent account of this tree in Flora Legnosa dsl Sottoceneri, pp. 83-112 (1904), should alsobe consulted. He describes sixteen varieties, cultivated for their fruit in Switzerland and Italy.IV S

842 The Trees of Great Britain and Irelandin the west and the Bihar Mountains in Transylvania. According, however, toFliche, 1 it is not truly wild in any part of France nor even in Corsica, as it neverforms part of the real forests and is generally found either as coppice or as isolatedtrees planted by man rather than as a true forest tree.In France it is common in Provence, Dauphine, the Cevennes, Perigord,Limousin, and all the central plateau, and it fruits abundantly in the environs of Paris.As in England, it was long supposed that there were large forests of chestnut inancient times, and it is popularly believed that the severe winter of 1709 caused theirdestruction in the region of the Loire. This is, however, an error, and the wood supposedto be chestnut, occurring in ancient churches and other buildings at Troyes, Reims,Sens, Chartres, and in Notre Dame at Paris has been conclusively proved to be oak.2The chestnut in France is rarely cultivated in high forest, as the timber isvery liable to shake and to rot at the heart, so that sound pieces of considerablesize are rarely obtained. It is, however, often cultivated as coppice, for useas vine props and hoops for casks. Mathieu mentions a tree growing nearSancerre in the department of Cher, which is 30 feet in girth and appears to beperfectly sound. Mr. Chaumette 3 saw a chestnut in 1851 near Evian in Savoy,which had a girth of 54 feet, was 85 feet high, wide-spreading and well-shaped, butthe trunk was perfectly hollow.The chestnut is truly wild in Spain, 4 and appears to attain there a greaterdevelopment than in any other country. In the north, as in the provinces of Galicia,Asturias, and Biscaya, it constitutes forests of great extent, growing in company withQiiercus Toza, Q. sessiliflora and Q. pedunculata or occasionally with the beech,and ascends from sea-level up to 2500 feet. It abounds in the mountains nearAvila; and between Banos and Bejar there are vast woods in which it occursmixed with Qwrcus Suber. I t also occurs in the mountains of Toledo and ofEstremadura and in the Sierra Morena. In the northern parts of Navarre andAragon, it ascends in the Pyrenees to 3000 feet. In the extreme south of Spain, itno longer descends to sea-level, but forms a zone between 2 700 and 5400 feet altitudein the Serrania de Ronda and the Sierra Nevada ; and small woods also occur onthe Alpujarras. The chestnut is also common in Portugal, and various localitiesare mentioned for it by Colmeiro.In Italy the chestnut occurs throughout the Apennines and also in Sicily,rising to 3000 or 4000 feet elevation ; and pure woods are found, especially inTuscany. The most celebrated tree of this species, the Castagno di Cento Cavalli,growing on Mount Etna, was visited by Brydone 5 in 1770, who found it to be ahollow shell, which looked rather like a group of five trees growing together than asingle tree. Brydone made its girth 204 feet. This ruin has lately been seen byMr. Druce 6 of Oxford, who found four distinct parts still remaining, three of which1 Cf. Fliche, in Bull. Sec. Bet. France, liv. 132 (1907), concerning the recent discovery of chestnut charcoal at aprehistoric station in the department of Dordogne. Dr. Christ, in Flare de la Suisse, snppl. 48 (1907), discusses the questionof the distribution of the chestnut, and now agrees with Engler (Ber. Schweiz. Bat. Ges. xi. 81) that it is not truly wild inSwitzerland, either in the Jura or in the Alps.2 Mathieu, Flare Forestiere, 328, 329 (1897). 3 Phytologist, iv. 71 (1851).4 Cf. Willkomm et Lange, Prod. Fl. Hispanica, i. 246 (1861); and Willkomm, Forstliche Flare, loc. cit.6 Brydone, A Tour through Sicily and Malta, i. 119 (1790). 6 Cf. Pliarmac. Journ. Feb. 27, 1904, p. 258.Castanea 843looked like mighty trees, though not over 70 feet in height. It still fruits freely,and bears on its branches several bunches of the southern species 1 of mistletoe.Besides this great tree, there are four other enormous trees on Mount Etna,mentioned by Parlatore and Tornabene, viz. the Castagno della Nave, 2 2 metresin girth; the C. della Navota, 1 8.7 metres; and the two C. di Santa Agata, 2 2.6and 26.3 metres, all sound and much more beautiful than the C. di Cento Cavalh.The chestnut forms a part of the forests in the south of Germany, but is notindigenous, being introduced, it is supposed, by the Romans, as in Alsace, where itforms large woods, ascending to 2000 feet, on the slopes of the Vosges, and in theplain, as around Sulzmatt and Rohrbach. Along the foot of the Vosges in Alsace,chestnut coppice, treated on a fifteen years' rotation, is very common, the wood beingused for vine-props. The chestnut is cultivated largely in southern Germany as afruit tree, and as an ornamental tree in parts of north Germany, where in favourablesituations, as near Brunswick and at Blankenburg, it ripens its fruit perfectly.It is planted in southern Sweden and on the coast of Norway between Christianiaand Christiansand, and occasionally ripens its fruit. According to Schiibeler, itexists in Norway as a bush as far north on the coast as lat. 63 .In Austria it is commonly planted, as in Bohemia and Moravia, while farthersouth it is supposed to be often wild. There is a remarkable wood of chestnut, onthe domain of Mokritz in lower Carniola, which lies between 500 and 1500 feetelevation. In Carinthia, the chestnut constitutes 10 per cent of the mixed forest onthe Neuhaus estate, ascending to 1800 feet; and at Bleiburg it is still a fine tree at3100 feet elevation.On the eastern side of the Adriatic,2 from Fiume to Castelnuova, the chestnutforms a part of the forest, which is composed mainly of oak and laurel; while in theinterior it is a considerable element in the oak forests of western Bosnia andCroatia. It occurs also mixed with the beech in Croatia, Bosnia, Herzegovina, andMontenegro. Wilkonim speaks of grand woods of chestnut in southern Hungary,Slavonia, Croatia, and Dalmatia; and mentions large wild forests in the Etsch valleyin the Tyrol. Velenovsky 8 states that in the western Balkans, not far from thetown of Berkovitza, there are extensive woods of chestnut, which are apparentlywild, and have an undergrowth of the common hazel. Elsewhere in Bulgaria thechestnut appears to be planted, and is not a common tree.The chestnut 4 is very common in the mountains of Greece, and is met withalso in the islands of Keos, Naxos, and Crete. It occurs either solitary orgregariously, and in some parts of the mountains forms extensive woods.In Macedonia, 6 Thrace, Albania, and Bithynia, the chestnut often forms the lowerborder of the deciduous forest, at 1200 to 3000 feet, occurring above the region of evergreen shrubs ; but here and there it descends to sea-level. Chestnut woods occur onOlympus, in the peninsula of Mount Athos, and on Mount Kortiach near Salonica.1 This appears to be, judging from an imperfect specimen kindly sent by Mr. llruce, Viscum laxitm, Hoissier et Reuter.Cf. Nyman, Consp. Fl. Eurof. i. 320(1878).3 Cf. Beck von Mannagetta, Veg. Verh. Ittyrischen Landern, 147, etc. (1901).3 Flora Bulgarua, Suppl. i. 254 (1898). 4 Halacsy, Consp. / '/. Grace, iii. 125 (1904).6 Grisebach, Fl. Kumelica, \. 339 (1843).

844 The Trees of Great Britain and IrelandIn Asia Minor the chestnut has been found wild in northern and westernAnatolia; but it appears to be absent from the Taurus and Lebanon. In theCaucasus 1 it is found throughout the whole territory, and also in the Talysch, upto 6000 feet elevation ; and it extends into north Persia.A large chestnut grew in Madeira, on the estate of Count Carvalhal, at Achada,23 kilometres from Funchal, and was reported by M. Joly 2 to have been about160 feet in height with a girth at 3 feet 4 inches from the ground of 38 feet 8 inches.It was burnt down three years ago, and no trace of it now exists. The chestnut isnot indigenous 3 in Madeira, although formerly many large planted woods existedthere, most of which have disappeared.The chestnut was probably introduced into England by the Romans. Charcoal,supposed to be of chestnut, was discovered by Mr. H. N. Ridley * associated withpalaeolithic implements and the bones of the rhinoceros in a brick-earth pit betweenErith and Crayford in Kent. Mr. Clement Reid 5 has not found any evidencecorroborating the possibility of the tree being a native of Britain in prehistoric times,and Mr. Ridley's specimen may be capable of some other explanation.The tree 0 is mentioned in Anglo-Saxon literature as the cisten or cyst-beam.The modern name chestnut is a shortened form of ckesten-nut, the fruit of the chesten,the early English name of the tree, representing the old French ckastaigne, from theLatin castanea. King Henry II., in a grant to the Abbey of Flaxley in the Forestof Dean, says : 7 " de eadem foresta dedi eis decimam castanearum mearum " ; and it isprobable that the chestnuts here referred to were cultivated at this early time fortheir fruit and not for their timber.Natural seedlings 8 are common in the southern counties, as in Kent, Surrey,Sussex, and Hampshire; and the chestnut may be considered to be naturalised insome places. Briggs states that it is naturalised in Cotehele wood near Plymouth ;but as Bromfield remarks, it does not spread over waste places in the way that oakand pine commonly do. (A. H.)CULTIVATIONThe Sweet or Spanish Chestnut, as it is usually called, is on soils and situationswhich suit it one of the largest trees in England, and both from an ornamental andan economic point of view one of the most important of exotic hard-woods.It is most at home in the southern counties, for though hardy in almost anypart of Great Britain, it loves a warm soil and a warm summer climate, butwill grow to a large size where the rainfall is as much as 60 inches per annum.1 Radde, Pflanzenverbreit. Kaukasusland, 1 82 (1899).s Cf. Vahl, in Engler, Bot. Jahrb. 1 905, p. 307.6 Origin British Flora, 1 46 (1899).2 Note stir un Chataignier Colossal.4 Journ. Bot. 1 885, p. 253.0 Cf. Murray, New English Dictionary, ii. 329 (1893). The village of Cheshunt does not take its name, as has beensupposed by Ducarel and others, from the chestnut. Skeat, in Place Names of Hertfordshire, 3 7 (1904), proves that Cheshuntis a corruption of Cestrehunt, derived from Anglo-Saxon ceaster, a camp, and Jmnta, a huntsman.7 Ducarel, in Phil. Trans. 1 771.8 There are numerous natural seedlings in Windsor Park, especially amongst the tall pines near Virginia Water. Theyare also common in Norfolk, at Fulmodestone and at Ilargham.Castanea 845With regard to soil, the chestnut is rather fastidious, as, though it will exist for atime, it rarely thrives on soils of a chalky or limy * nature, and will not grow in stiffclay or in peaty soil.All the largest I have seen are on greensand or old red sandstone; and whencultivated for coppice-wood, which is probably its best economic use, it requires abetter soil and climate than any other tree usually so treated. It is propagated byseed, which ripens in the southern counties abundantly in good seasons, though thefruit is inferior in size and quality to what is imported from Spain and France. Thelargest nuts should be chosen and kept dry in sand until spring, as they are devouredby mice, and if sown in autumn are liable to rot if exposed to much frost and wet.They should be transplanted when one year old and kept rather crowded in thenursery until they are 5 to 6 feet high, as they are liable to become very bushy ifthey have room to spread. They are not difficult to transplant, if grown in lightsoil, but must not be left more than two years before transplanting. 8A remarkable instance of the grafting of the chestnut on the oak s was shown mein the Botanic Garden of Dijon in France by M. Genty, the professor of botanythere. The history of this tree is given in full by M. A. Baudot, in a pamphletpublished at Dijon in 1907, from which I gather that in 1835 some acorns of thepedunculate oak were sown by M. Meline, five of which were grafted in 1839 withscions from the chestnut. Three of the grafts failed to take, another was injured bywind, the fifth pushed a shoot in the first year about 4 feet long, and grew sovigorously that it is now nearly 40 feet high with a girth of 4^ feet. The treebore small fruit in 1852 ; and in 1903 some were sown, which germinated andproduced three young plants, of which two are now planted out in the garden atDijon, and a third was sent to M. M. de Vilmorin at Les Barres.The varieties of the chestnut grown for fruit are usually grafted in Frenchnurseries, but are rarely planted in England at present so far as I have seen.As coppice-wood the chestnut is principally found in the hop-growing districtsof Kent, Sussex, and Hants, where, until wiring was introduced, it was one of themost valuable products of English woodland, being cut at intervals of 8 to 12 years andrealising frequently £2 to £$ per acre per annum. But now, though still morevaluable than ash or hazel, it has fallen so much in price that these coppices are notas carefully managed as they used to be ; and the split poles, which are so largely usedfor fencing, are said to be imported from France. In such coppices the stools areat 5 to 6 feet apart, because the thinner a hop pole is in proportion to its height the1 Fliche and Grandeau (Ann. Chimie et Physique, 1 874, p. 354) proved by experiments, that the presence of a considerable amount of lime in the soil causes the chestnut to languish or to die, as too little iron is absorbed by the tree, and thenormal function of the chlorophyll is deleteriously affected. Alphonse de Candolle, in N-uovo Ciorn. Bot. Ital. x. 228(1878), states that the chestnut is never found growing in Switzerland on limestone, and that in places where it is believed tooccur on limestone, careful examination shows that the roots are surrounded by siliceous soil. However, he brings forwardevidence to show that in the climate of south-eastern Europe, as in Hungary and Istria, the chestnut is occasionally foundthriving on pure limestone.2 Sir H. Maxwell recommends sowing the best foreign nuts, but these produce seedlings which in my nursery are muchmore tender when young, than those raised from smaller English-grown seed, and when required for timber trees I shouldprefer the latter.3 M. Trabut, in his pamphlet, Le ckataignier en Algirie, published as bulletin 37, by the Department of Agriculture inAlgeria, states that he saw at the Villa Thuret in Antibes, a fine chestnut, which had been grafted on Qvercus Mirbeckii.

846 The Trees of Great Britain and Irelandmore valuable it is, the young hop shoots, according to Cobbett, disliking a thick poleto twine round.At Welbeck the chestnut is considered by Mr. Michie, 1 forester to the Dukeof Portland, to be the most profitable tree to grow on sandy soil, as it growsmuch faster than oak and realises about is. 2d. per foot at a much earlier period.He showed me a plantation on Tressless Hill thirty-eight years old in 1903, in whichthe trees averaged about 65 feet high by 3 feet in girth, and stood about 150 tothe acre. He said that they should not be grown without underwood, because insevere winters the unprotected trunks were liable to be cracked by frost near theground. 2We have no exact records of the amount of timber per acre that may be producedby the chestnut when grown for timber in England, but I think that in the south ongood land it would probably be greater than that of any other tree. One veryremarkable case is a grove of 34 chestnuts and 9 oaks by the drive leading toBicton House, Devonshire, which average about 100 feet high, by 6 to 7 feet ingirth in the middle of the grove, and 9 to 12 feet on the outside (Plate 232).[ estimated that this area was about half an acre, and the cubic contents of thetimber on it about 5000 feet. At my request the late Mr. Mark Rolle had it carefully measured and wrote me on December 19, 1903, that the exact area on whichthe trunks stood was i rood 32 poles, though, of course, the branches extendedover much more. The cubic contents were 7300 feet and the age of the treesabout 150 years. We may therefore take at least 1 0,000 feet per acre as the resulthere.Another very striking instance of the same character is a grove called "TheChestnut Tole" 3 in Mr. Ashley Dodd's park at Godinton, Kent, where a greatnumber of fine trees, having clean boles of 5010 70 feet high by 8 to 10 feet in girth,grow mixed with ash. One of the chestnut trees was 86 feet to the point where thebranches began, and I think that the timber in this grove would produce as great1 Mr. Michie has sent the following note :"Sow seed in March, collected from sound, healthy, straight-growing trees, forty-five to fifty-five years of age, as Ifind that seeds from trees of that age produce stronger seedlings thaii seed from younger or older trees, or than foreignseed. At one year old I lift the seedlings, shorten the tap-root, and plant in nursery lines. Care must be taken to plant infresh, sweet soil, as the root is very liable to malformation if in contact with fresh manure. In the following year cutthem down to within one inch of the ground, which will cause them to throw out a strong and straight stem from 2 to3 feet long ; after which, at three years old, they can be planted out with safety. Without this treatment before plantingout, they generally require cutting off close to the surface, which is not always desirable in the planted area, owing to rankgrass, bracken, etc., which smothers the young shoots.1 am greatly in favour of pure chestnut woods, very little thinning, and the encouraging of as much undergrowth aspossible, especially on the outsides of plantations, to prevent cold and frosty winds blowing through. At sixty years of agethe trees should stand no more than 16 feet apart, which equals 170 per acre, and taking them at the low average of 50cubic feet per tree, means .£425 (at is. per foot) when the crop is realised.The above crop can be grown on a sandy soil, which is of little value for ordinary agricultural purposes; for instance,in Birklands Wood, adjoining Budly Forest, where the soil is very sandy and light, oaks covering an area of about 100acres, although from sixty to eighty years of age, are long and slender, and contain on an average not more than 6 cubicfeet of timber each; whereas some Spanish chestnuts, planted less than sixty years ago, contain fully eight times as muchtimber as the oaks.On this estate the timber is used for making gates, gate-posts, and all kinds of fencing; also for window-sills of faimbuildings, etc. Timber merchants buy it to supply the Sheffield trade (strickle handles, etc.), and also to put in the inside ofthreshing machines, for coffin boards, etc. The timber should be slowly and thoroughly dried before being used.2 The same thing has occurred at Kew; and, as Sir W. Thiselton-Dyer pointed out, the cracks occur on the south side,and are the result of too rapid thawing by the sun.3 Tole seems to be a local name for a clump of trees standing on the crown of a hill.Castaneaa quantity per acre and of better quality, than the grove at Bicton. But, howeverattractive such plantations may be from an ornamental point of view, there is nodoubt that the timber is worth much more if cut young; and, as a matter of fact,most of the old chestnut trees in the south of England are so shaky that a greatpart of their timber is only fit for firewood or fencing.The chestnut is a good avenue tree in those parts of England where the soiland climate suit it, and there are fine avenues at several places. One of the bestknown to me is at Cowdray Park where there is an avenue about a mile long,commencing at the bottom of the hill, where the trees are very large, andrunning up to an elevation of 500 feet or more. According to Loudon this avenuecontained 300 trees. Another very fine one at Thoresby is supposed to have beenplanted by Evelyn, many of the trees in which are about 20 feet in girth. I noticedhere that the spiral twist in the trunk of the chestnut is variable in direction. Ofthree trees standing together in this avenue, one was twisted from left to right, onefrom right to left, and one had no twist at all; but this twisting of the trunk is commonest on light sandy soil and usually indicates shaky timber.Another fine avenue of chestnuts is at Newhouse Park, on the property of SirRobert Newman, near Mamhead, Exeter. This is 24 yards wide, with the trees 1 2yards apart, which seems to be the correct distance for this tree in an avenue, as itrequires more room than the lime or elm. These trees average about 15 feet in girthand are 70 to 80 feet high. The largest that I measured was 18 feet 8 inches in girth.REMARKABLE TREESThe number of large chestnut trees is so great that it is quite possible wemay omit some of them, but there is no doubt that the most celebrated, andperhaps the oldest planted tree in England, is the Tortworth chestnut, which hasbeen frequently described, and is figured by Strutt, plate xxix., and by Loudon,p. 1988. Strutt says that in 1766 it measured 50 feet in circumference at 5 feetfrom the ground, had a stem 10 feet high to the fork, and had three limbs, one ofwhich was at that time 28^ feet in girth. It was said by Sir R. Atkyns, in hisHistory of Gloucestershire, p. 413, to have been growing in King John's reign, andto have been " 197 yards in compass." It has since been mentioned and describedby almost every writer on trees, but I am informed by Lord Ducie that a good deal ofits history is more or less mythical. At present it is by no means a beautiful tree,and so much of its original trunk is decayed, that no measurement is of much value.I think that no one would recognise the existing tree as having formed the subjectof Strutt's plate; but notwithstanding its age it still produces nuts, from whichseveral trees have been raised and planted.Another very large and celebrated chestnut, also figured by Strutt, plate xiii., andby Loudon, p. 1989, grew at Cobham Hall, 1 Kent, and must have been a finer treethan the one at Tortworth. It measured in 1822, according to Strutt, 29 feet in1 The finest chestnut now existing in this park grows at Ashenbank, and measured, in 1906, 93 feet in height and13 feet 10 inches in girth, with a good bole 40 feet in length.

848 The Trees of Great Britain and Ireland Castanea 849girth at the narrowest part 3 feet from the ground, 33 feet at 12 feet up, and 40 feetat the point where the trunk divided. It was " called the four sisters, from its fourbranching stems closely combined in one massive trunk," though the figure does notshow this clearly. It has now entirely decayed.Another historic tree, the " Monmouth Tree," 1 at White Lackington, inSomerset, was destroyed by the severe storm of Ash Wednesday in 1897. It wasreputed by tradition to have been the tree under which the Duke of Monmouth hada famous banquet in 1680. It was 25 feet in girth with a total height of only 49feet, and had a very venerable appearance. Lord Petre measured in 1758 in WrittlePark, three miles from Ingatestone in Essex, a chestnut 45 feet in girth at 5 feet fromthe ground. 2In Waldershare Park, Kent, the seat of the Earl of Guilford, there are someremarkably fine chestnuts, the largest in girth being 23 feet 3 inches, but not a wellshapedor tall tree. The finest, in my opinion, is a tree 112 feet high with a straightand clean bole 50 feet long by 15 feet 2 inches at 5 feet, and carrying its girth wellup. I estimated the contents of the first length alone at 50 feet by 36 inchesquarter girth, making 450 feet of clean timber.Fredville Park, the seat of H. W. Plumptre, Esq., in the same district of Kent,contains some splendid chestnuts, the largest of which is about 80 feet by 26 feet 3inches. Another is called the Crows' Nest, from the fact of its having a platform,with benches and a table large enough to seat about twenty people, built in thecrown at about 12 feet from the ground and reached by a ladder.An immense but very ill-shaped chestnut tree dividing at 5 feet into threemain limbs grows at Sunninghill Lodge, near Ascot, the seat of Percy Crutchley,Esq., of which a photograph was shown by him at the Lincoln Exhibition of theRoyal Agricultural Society in 1907. This tree was carefully measured in 1816 byT. Luff, who estimated its contents at 716 cubic feet. A measurement made June15, 1907, by M. C. Squires, gives its contents as 1282 feet, an allowance for barkof i^ to 2 inches being made.The finest chestnuts growing near London are those in Kew Gardens, thelargest of which measures 75 feet high, and 20 feet 10 inches in girth. These wereprobably planted early in the eighteenth century.In Herts, there is a large chestnut at Lockleys Park near Welwyn, which theHon. Arthur Bligh informs us is 21 feet in girth; and at Broxbournebury, Mr. H.Clinton Baker measured a tree in 1908, 65 feet by 23 feet 9 inches.At Betchworth Park, part of the Deepdene estate, near Dorking, Surrey, thereare many splendid chestnuts, 8 the finest though not the largest round, being 21 feet5 inches in girth and 90 feet in height. For girth alone I know of few trees inEngland equal to one measured here by Henry, which, though its bole is only 8 feetlong, is 26^ feet in girth at the narrowest point.1 Cf. H. Norris in Proc. Somerset Archaeological Society ( 1897), where a figure of the tree is given.2 Ducarel, Phil. Traits. 1 771.3 An interesting article on the chestnut trees in Betchworth Park appeared in Gardeners' Chronicle, 1 841, p. 4. At thatdate there were about 80 trees, all of large dimensions. Dr. Aikin, in Monthly Magazine for 1798, mentions the rows of oldchestnut trees in this park.At Petworth Park, Sussex, there are several very fine chestnuts, of which onemeasured by Sir Hugh Beevorin 1904, was no less than 118 feet high by 19 feet ingirth, with a trunk clean to about 70 feet, and estimated to contain 800 feet oftimber. It grows on the west side of the drive on the west side of the park, abouttwo miles from the house. Another in a clump close to the house I found to beabout 100 feet high by 21 feet 9 inches in girth.At Steventon, North Devon, there is a very large tree in the garden, which Mr.Barrie measured as follows in 1890: height 86 feet, bole 22 feet 6 inches, girth16 feet ii inches, spread 100 feet in diameter, contents 833 feet.At Tyberton Court, Herefordshire, near the place where the big oak formerlygrew, 1 and on soil heavier than the chestnut usually likes, there is a very fine twintree, which looks as if two stems had started together from the same root. At thebase the two measure 31 feet round and are about 95 feet high, one trunk being20 feet, the other 17 feet 6 inches in girth.At Highnam Court, Gloucestershire, there is a chestnut stool, which girths 32feet at 3 feet from the ground, giving off four great stems 80 to 90 feet in height.At Croft Castle, Herefordshire, there is a row of fourteen trees which weredescribed in the Transactions of the Woolhope Society, 1 871, p. 306, where theirrespective girths are given, and average about 17 feet, the two largest being then20 feet 3 inches and 20 feet 5 inches. They seemed, when I saw them in 1904, tobe long past maturity.Below Warwick Castle, on the banks of the Avon, there is a chestnuthaving a large branch resting on the ground, where it has taken root and thrownup a large vertical stem, the only instance of self-layering 2 I have seen in this tree.The trunk in 1907 measured 16 feet 3 inches in girth. This tree is figured inGardeners Chronicle, 1 873, fig. 222.In Ashridge Park there are many fine chestnuts, one of which has its trunkcovered with great burrs and is 24 feet in girth. At Chatsworth there is a chestnuttree of which Mr. Robertson, forester to the Duke of Devonshire, has been goodenough to send me a photograph. He makes it 86 feet high, with a bole 45 feet by15 feet 10 inches, and the cubic contents about 700 feet.At Harleston, near Althorp, on Lord Spencer's property, are some immensechestnuts growing in a field near the church, on rich red sandy soil, the survivors ofa row of which many were blown down many years ago. The largest measures90 feet by 22 feet 6 inches, and was estimated by Mr. Mitchell, now forester atWoburn, to contain 1200 feet of timber (Plate 233). Another of about the sameheight has a bole 27 feet by 21 feet 6 inches and contains about 887 cubic feet.If the length of clean trunk be considered, I have seen no chestnut equal to oneat Thoresby (Plate 234), which has been drawn up in a thick wood of beech treescalled Osland, and has a clean bole as straight as possible, 70 feet long by 11 feet3 inches in girth, and a total height of about no feet. This was planted about1730, and is on a sandy soil overlying the Bunter beds of New Red Sandstone. 3IV1 Cf. vol. ii. p. 310. 2 Henry saw a self-layering tree at Riccarton. Cf. p. 851.3 For further particulars of this remarkable plantation see our article on the oak, p. 322.

850 The Trees of Great Britain and IrelandAt Euston Park, Suffolk, the property of the Duke of Grafton, Mr. Marshallshowed me, in 1905, in a wood called Barnham Springs, a remarkable growth ofchestnut from a stool cut forty-two years previously. Sixteen straight stems about60 feet high, and z\ feet in girth, had sprung up from the outer edge of thestump, and collectively measured 30 feet in circumference. This growth seemsto show how such trees as the one on Mount Etna have been originally formed,as in another 50 or 100 years these stems will probably seem like one tree. AtMerton Hall, Thetford, Norfolk, a chestnut, planted about 1660, is 87 feet high,with a clean bole, 40 feet in length and 11 feet 4 inches in girth.At Shrubland Park, Ipswich, the property of Ladyde Saumarez, there are somevery large chestnut trees in the grounds. The largest of these, according to Mr.Taylor, measures at ground line 47 feet; at 3 feet 31 feet; at 6 feet 27\ feet.Having had its top blown off some years ago, it is now only 55 feet high.The finest existing chestnut, if height and girth together are considered, thatI have seen, is a tree in a valley called Mackershaw, near Studley Royal, whichseems to be the one figured by Loudon, p. 1986, of which he gives the height as 112feet and the girth at i foot as about 23 feet. When I measured this splendid tree in1904 I made it 112 feet by 20 feet at 5 feet from the ground, and it seemed to be inperfectly sound condition.At Rydal Hall, Westmoreland, there is a very fine tree, girthing 26^ feet at5 feet and 37 feet at the ground, of which the owner, Mr. S. H. le Fleming, haskindly sent me a photograph (Plate 235).In Wales the finest tree I have seen is one which grows just outside the gardenat Dynevor Castle, and measured in 1908 about 113 feet by 16^ feet, with a cleanbole about 30 feet high. A photograph which was taken proved unsuccessful owingto its being surrounded by other trees.Notwithstanding its southern origin, the chestnut grows with great vigour inmany parts of Scotland, and, according to Loudon, 1 who quotes from Walker'sEssays, p. 29, the first exotic tree planted north of the Tweed was a chestnut, ofwhich in 1760 a part of the trunk remained, at Finhaven, an ancient seat of theEarls of Crawford. This was measured in 1744, and, as attested before two justices,was 42 feet 8£ inches in circumference close to the ground.The largest tree I have seen myself is in the Cherry Park, near the stables atInveraray Castle, and measures about 77 feet by 20 feet, with a bole about 16 feetlong. This tree was said in the Old and Remarkable Trees of Scotland to havebeen in 1867 the largest in Scotland, though one at Tyninghame was as tall; andthere are two fine ones, both over 16 feet in girth, at Ardkinglas, in the sameneighbourhood. Lord Kesteven informs us that there is a chestnut 25 feet in girth,growing at Stonefield, near Tarbert, Argyllshire. At Kirkconnell, south of Dumfries, Henry measured in 1904 a fine tree, 73 feet high and 18 feet in girth, with abole of 25 feet. At Kirkmichael House, Ayrshire, a tree measured i8£ feet in girthin 1892 ; and at the Auld House, near Glasgow, two trees, about 60 feet high in 1904,were 16 feet 3 inches and 14 feet n inches in girth respectively.1 Arb. et Frut. Brit. i. 34, 90 (1838).rCastanea 851At Castle Menzies in Perthshire there are several very large trees, perhapsover 300 years old, one of which, in the washing-green, is about 20 feet in girth.Another in the park at Murthly, though not remarkable for height, has a trunkabout 15 feet high and 19 feet 7 inches in girth, twisting from left to right. AtDupplin Castle, Perthshire, there are some very fine trees in a sheltered dell belowthe castle. One of these has a short bole no less than 21 feet 4 inches in girth;another is about 70 feet high by 17 feet 9 inches in girth. Many other largechestnuts in this county are recorded by Hunter; but as a rule they are remarkablerather for their age and girth than for their height, which rarely exceeds 80 feetin Scotland.Sir Herbert Maxwell says 1 that the tallest recorded in Scotland is at MarchmontHouse, Berwickshire, which in 1878 measured 102 feet by 14^ feet, with abole of 32 feet; but Sir Archibald Buchan-Hepburn tells us that a tree at YesterHouse, Haddingtonshire, is 112 feet high by 18 feet 8 inches in girth, according tocareful measurements, taken in 1908 by Lord Tweeddale's forester.The chestnut at Riccarton, near Edinburgh, which was described and figured in1829 by Monteath, System of Draining, 209, as an old tree remarkable for layering,had two stems in 1905, one 17 feet in girth and the other, very decayed, 12 feet ingirth, both giving off branches which had layered and become independent trees.From Castle Leod in Ross-shire Mr. Wotherspoon sends me a photograph of atree, which is probably the largest of the species existing so far north. It is76 feet high and girths 28 feet close to the ground, 21 feet 4 inches at 5 feet, witha bole 14 feet long. 2In Ireland the chestnut thrives remarkably well, and, growing fast, might inmany places be cultivated for its timber. At Fota the chestnuts in a plantationmuch exposed to the strong winds from the sea, withstood without injury the severegale of 1903, when many other species were blown down.The most remarkable chestnut in Ireland is the famous tree at Rossanagh,Wicklow, which was planted, according to Colonel Tighe, who has the family records,in 1718 (Plate 236). This tree is of the large spreading type with a short bole whichdivides into three mighty limbs. The girth of the main stem close to the groundwas in 1903 49 feet, at 3 feet up 27^ feet, and at 5 feet 29^ feet. The height of thetree is about 80 feet, the spread of the branches being 100 feet in diameter. Thethree limbs girth respectively 12 feet 8 inches, n feet 2 inches, and 10 feet.A very fine tree is growing at Powerscourt which was 84 feet high in 1905,with a good trunk carrying its full girth up to 18 feet and giving off the first branchat 20 feet up. It was 28^ feet in girth at the ground, and 22! feet at 5 feet up.At Clonbrock, Co. Galway, there is a tree growing on limestone, planted in1801. It was 8 feet 6 inches in girth at 3 feet up in 1871, and 12 feet 9 inchesin 1904. The chestnut grows at Clonbrock, where rhododendrons refuse to grow;and in the case of the tree just mentioned there is undoubtedly a large proportionof lime in the forest soil on which it stands. At Shannongrove, near Limerick,1 Green's Encycl. Agric. i. 373 (1907).2 This is probably the same tree which Loudon mentions, p. 2001, as growing at Castle Send (sic) in Cromarty.

852, The Trees of Great Britain and Irelandthere are some large trees, more remarkable for girth than for height, one being24^ feet round at two feet from the ground.At Rostrevor House, Co. Down, the seat of Colonel Sir J. Ross of Bladensburg,a chestnut, about 25 feet high, is remarkable for the large size and colour ofthe young leaves, which were purplish when I saw them in July, and are said to turncopper colour in autumn. This variety is of unknown origin, and I have seennothing like it at Kew or elsewhere.TIMBER, MISCELLANEOUS PRODUCTSA great deal has been written as to the use of chestnut wood for the beamsand roofs of ancient buildings, both in England l and France, but it is now prettygenerally admitted that most of the supposed chestnut wood is really that ofthe oak, which it slightly resembles. 2 This subject has been so well discussedby Loudon (pp. 1787, 1989, and 1992) that I need not further allude to it; butthe properties and uses of the wood were apparently much better known formerlythan now, and Mr. N. Kent, in 1792, wrote an excellent paper on the subject fromwhich Loudon quotes largely (p. 1993). The pith of it all agrees with what I havebeen able to learn from various practical men that the wood when young is as goodor better than oak (because it has much less sapwood) for fencing, gate-posts, piles,and hop-poles; but that if allowed to become more than 3 to 4 feet in girth it is soapt to be shaky, that its value rapidly diminishes, and very old trees are usually onlyfit for firewood.The timber in some cases remains quite sound to a great age and becomes mottledand streaked with dark brown like brown oak. I found the butt of an old treeof this nature, in a small timber yard in Wilts, where it had been lying seventeenyears without any use being found for it. I had it cut into boards, from whichthe stiles and rails of an overmantel, and the frames of some doors have been made ;and these, when polished with oil, were both in grain and colour of remarkablebeauty. But even after this long period the wood was not dry, and shrank considerably after it was cut up, so that care must be taken not to put such woodtogether in a hurry.Mr. T. Roberts, forester to the Earl of Egmont, at Cowdray, informs me thatchestnut is used on that estate for joists, window-sills, door-jambs, and otherpurposes, and is found to last quite as long as oak and to be much easier to workup; he also thinks it less liable to insect attacks than oak (presumably sappy oak).But the trees when a hundred years old are all more or less inclined to be shaky,1 Sir George Birdwood in Reports on the Cultivation of the Spanish Chestnut, p. 9, note (India Office, I2th March 1892),states that the late Mr. T. Blashill, who was architect to the London County Council, pointed out in a letter to the Times,that the only instance he knew of chestnut wood in English medkeval carpentry is that of the chancel screen of the church,formerly of the Knights of St. John, at Rodmersham in Kent. The Rev. A. II. J. Massey, Vicar of Rodmersham, tells me,however, that the chancel screen is a modern one of oak, with portions of an ancient screen of chestnut wood worked into it;but the screen separating the Lady Chapel from the chancel, is composed entirely of chestnut wood.8 Mr. Blashill, in Sessional Papers of the Royal Institute of Architects, No. 12 (1877-78), has finally settled any lingeringdoubts which may exist. On the question of oak or chestnut in old timber roofs, he says that in some specimens of Englishoak, particularly in the variety called sessiliftora, the medullary plates are very thin and wide apart, and such specimens areoften mistaken for chestnut, but a very clean transverse section will always render the plates visible. Though usually lighterthan the rest of the wood, they are often dark, and such specimens have also been mistaken for chestnut. He goes on to saythat the clean grain and pleasant working of chestnut make it very suitable for joinery, and there is no fear of its durabilityCastanea 853which prevents their being cut into small scantlings. Mr. Weale tells me that it isused extensively in London for making coffins instead of oak. For making hoops,poles of chestnut are considered the best; and the wood is also largely used formaking wine casks in France and Spain.A section from the butt of a chestnut tree, said to be two hundred years old, wasshown by the Marquis of Exeter at the Forestry Exhibition of the Royal Agricultural Society at Lincoln in 1907. This tree measured 7 feet in diameter at thebutt and was fairly sound and free from shakes. I am informed by Mr. Dansonthat this tree was grown on a clay soil overlying ironstone, with a north-westexposure, about 230 feet above sea level.At Shobdon Court, Herefordshire, the seat of Lord Bateman, I saw the trunkof a large chestnut, measuring 19^ feet in girth, lying on the ground. It was quitesound, with the exception of two small ring-shakes, and by counting the rings Ifound that it was 207 years old.Such poles as are too thick for hop-poles make, on account of their durability,one of the best forms of park fencing that I know, of which many instances arequoted by Loudon. It is said that a park fence, erected in 1772 by Mr. Windhamof Felbrigg, of oak and chestnut thinnings, was taken down in 1792, when thechestnut was found as sound as when put down, while the oak was so much wastedat the ground level that it could not be used again without support.The Earl of Ducie exhibited at the Stroud show of the Gloucestershire Agricultural Society in June 1907 specimens of fencing posts made from chestnut, plantedby himself in 1855, and cut in 1885, which had been in use for twenty-two years, andwere still quite sound.The walking- and umbrella-sticks, which are known in the trade as "Congosticks," are saplings of the chestnut, which are easily manipulated when growing,the knots or markings for which these sticks are valued being produced by laceratingthe bark through to the wood. They were formerly obtained from the north ofFrance, but are now almost exclusively produced near Carlstadt in Croatia. 1The fruit of the chestnut is so well known that I need say little about it, andthough in the colder parts of England it is often so small as to be of little use forhuman food, it is eaten by pheasants and deer. The large chestnuts eaten at dessertare imported, and are known under the name of " marrons " in France where theyare preserved in sugar and form a very favourite sweetmeat. 2being equal, and probably superior to that of any wood (presumably he meant English wood) except oak. He spoke of a largebridge having been built about 1858 of chestnut timber, over the river Wye at Hoarwithy near Hereford. The bridge after nineteenyears was taken down in a crippled condition, which he attributed partly to the design of the bridge, and partly to the decayof the timber at the numerous joints where water could lodge. Yet the great bulk of the wood was perfectly sound ; andseemed to show that for ordinary work not subject to damp, the timber may be very useful. Although he could not admit itsoccurrence in ancient roofs, it might be very suitably used in preference to deal or pitch pine ; and in church furniture itwould probably, in course of time, take a colour which would be far better than that of the stained woods now so much used.1 Keui Bulletin, 1899, p. 53.2 In some parts of Spain and Italy, and in the south of France, chestnuts are ground into flour; and in the form ofcakes, soup, and porridge, form a considerable part of the food of the poorer classes during winter. Specimens of chestnutflour and cakes are exhibited in the museum at Kew ; and in Kew Bulletin, 1 890, p. 173, an analysis of the flour is given byProfessor Church, who considers that it is easily digestible and probably useful as food for children. Further interestingparticulars concerning the use of the chestnut are given in Reports on the Cultivation of the Spanish Chestnut (India Office,I2th March 1892).

854 The Trees of Great Britain and IrelandA tanning material, 1 extracted from chestnut bark, is prepared near St. Male inFrance, and is largely exported to Belgium and to Glasgow. It is said to be used tomodify the colour produced by hemlock extract obtained from Tsuga canadensis?(H. J. E.)CASTANEA CRENATA, JAPANESE CHESTNUTCastanea crenata, Siebold et Zuccarini, Alh. Akad. Mnench. IV. iii. 224 (1846); Schneider,Laubhohkunde, i. 804 (1906).Castanea japonica, Blume, Mus. Hot. Lugd. Bat. \. 2 84 (1850).Castanea vesca, Gaertner, var. pubinervis, Hasskarl, Cat. Hort. Bog. Alt. 7 3 (1844).Castanea vulgaris, Lamarck, var. japonica, A. DC. Prod. xvi. 2, p. 115 (1864); Shirasawa, IconForest. Japan, text 63, t. xxxiv. ff. 14-25 (1900).Castanea vulgaris, Lamarck, var. yitnnanensis, Franchet, Journ. de Bot. 1 899, p. 196.Castanea saliva, Miller, var. actiminalissima, von Seeman, in Diels, flora von Central China, 2 87(1901).Castaneapubinervis, Schneider, Laubholzkunde, i. 158 (1904).A tree, usually smaller in size than the European species, but occasionallyattaining large dimensions. It is probably only a geographical form of that species,but can readily be distinguished and may be kept separate, as it probably differs ingrowth and in cultural requirements.The leaves are borne on shorter petioles, but resemble those of the commonchestnut in shape, being rounded or cordate at the base and having about twentypairs of nerves; but they are smaller in size and have much shallower serrations,with very long and fine spine - like points. The main difference lies in thepubescence,8 which is short and dense on the young branchlets, on the petioles, andon the midrib of both sides of the leaf. In the common chestnut this very distinctpubescence is either absent or replaced by a scurf, very different in appearance. Thecatkins of the eastern tree are more slender and the fruits of wild trees smallerthan in the common species. Castanea crenata also comes into flower, when stillvery young, and often bears fruit when quite a small shrub.In China Castanea crenata* occurs wild, mainly in the mountains of the centralprovinces, as a tree about 40 feet in height; and is nowhere abundant, and so far asI have seen never forms woods of any extent.1 -V similar extract, prepared from the wood of the chestnut, is largely manufactured in Corsica. Mr. Southwell, Vice-Consul at Bastia, gave me some interesting particulars about this industry, when I visited Corsica in December 1906. Thereare four factories near Bastia, which produce about 25,000 tons of extract annually. The bark is not employed in Corsica, asthe dark colour of the extract produced from it is objectionable. Four tons of wood yield about one ton of extract. Thewood is cut into chips, which are soaked under pressure in hot water, which extracts all the tannin and some of the colouringmatter. The resulting liquor is concentrated in vacua. Practically the whole of this extract is used in England and Germanyfor sole-leather. Mr. Southwell informed me that certain trees in "Corsica had brown-coloured wood, which produced anunsaleable extract. He had found by experiment that this brown colour in the wood is due to the presence of iron in thesoil. (A. II.) 2 Kew Bulletin, 1 893, p. 229.3 The pubescence over the lower surface of the leaf is similar to that of the European tree, and is very variable in quantityand persistence.4 The large chestnut tree occurring wild in China is considered by Dode to be distinct from the Japanese tree, and hasbeen named by him C. Dudouxii and C. Fargesii, in Bull. Soc. Dendr. France, 1 908, pp. 150, 158.Castanea 855The low grassy hills of the Yangtse valley and the hills in Chekiang are oftencovered in places with a scrubby growth of chestnut bushes, scarcely ever over5 feet in height. This is a distinct species, 1 and corresponds in many respects toC. pumila of America, the branchlets and petioles being covered with a dense, bristlypubescence, and the fruits extremely small, usually three in each involucre. Thishas been supposed to be Castanea mollissima, Blume, 2 an imperfectly known species.The Chinese have distinguished from the most ancient times two kinds ofchestnut, classically known as the K and the erh. The former, now known as thepan-li is the cultivated tree, the latter, known as the mao-li, is the wild form of thespecies, which produces remarkably sweet small fruit. These have been noticed bymany observers, as by Abel 8 at Tatung on the Yangtze, by Pere David 4 at Kiukiang,and by Fortune 5 near Ningpo, who introduced the small-fruited chestnut intoEngland 6 in 1853; but we are unacquainted with any trees raised at that time.Similar small-fruited chestnuts are known in Japan, and were exhibited in London 7in 1873. (A. H.)The chestnut is widely distributed in Japan where it is called " kuri," from Kiusiuand Shikoku, through the greater part of the mountain forests of Hondo, and inthe plains as far north as central Hokkaido. It is usually mixed with other deciduoustrees, but in some places forms pure forests of small area. Its wood is preferred forrailway sleepers to any other timber, but is not much valued for building purposes.Though, according to Sargent, 8 it does not attain any great size, yet I measured anold tree in the Atera valley which was 15 feet in girth (Plate 237).The tree is commonly seen on dry and barren hillsides in the form of coppice,which is cut every few years for firewood. It is also cultivated for its fruit, andseveral large-fruited varieties are grown which Sargent 8 says are equal in size to thebest in southern Europe, and are largely consumed as food in the towns, and alsoexported from Kobe to San Francisco. These varieties are more precocious than theEuropean tree, bearing abundant fruit when only 10 or 12 feet high, and he recommends their introduction from Aomori in the north of Hondo, as being more likelyto endure cold winters than the French or Kobe varieties.The Japanese chestnut was introduced into the United States 0 about 1891, andRehder 10 states that it is shrubby and usually begins to fruit when about six years old.It has proved hardy as far north as Massachusetts. So far as we know it has not yetbeen introduced into England. (H. J. E.)1 The shrubby chestnut of China is considered by Dode, in Bull. Soc. Dendr. France, 1 908, pp. 151, 152, 153, toconstitute three new species, C. hufehensis, C. Seguinii, and C. Davidii.2 Mus. Bot. Lugd. Bat. i. 286 (1850). Cf. Diels, Flora tion Central China, 288 (1901).3 Narrative of a Journey in China, 1 65 (1818). * Planta Davidiana, i. 277 (1884).6 Residence among the Chinese, 5 1, 144 (1857). e Card. Chron. 1860, p. 170.1 Ibid. 1 875, p. 270. 3 Forest Flora of Japan, 69 (1894).9 Cf. Bailey in Amer. Garden, May 1891, who gives a description and figure of the tree ; and Garden and Forest, viii.460 (1895). W. A. Taylor, in Cycl. Amer. Hort. i. 294 (1900), who enumerates nineteen varieties of the Japanese chestnut,which have been introduced of late years into North America, gives the date of the first introduction as 1876.1 Cycl. Amer. Hort. i. 257 (1900).

856 The Trees of Great Britain and IrelandCAST AN E A DENTATA, AMERICAN CHESTNUTCastanea dentata, Borkhausen, Handb. Forstbot. i. 741 (1800); Sargent, Silva N. Amer. ix. 13, tt.440, 441 (1896), and Man. Trees, N. Amer. 2 20 (1905).Castanea vesca americana, Michaux, Fl. Bar. Amer. ii. 193 (1803); Loudon, Arb. et Frut. Brit. iii.1984 (1838).Castanea americana, Rafinesque, New PL iii. 82 (1836).Castanea vulgaris, y. americana, A. De Candolle, Prod. xvi. 2, p. 114 (1864).Castanea saliva, var. americana, Sargent, Garden and Forest, ii. 484 (1889).Fagus Castanea dentata, Marshall, Arbust. Amer. 46 (1785).A tree attaining in America 100 feet in height. Bark dark brown, and dividedby shallow irregular fissures into broad flat ridges, separating on the surface intosmall thin appressed scales. Young branchlets with minute scurfy pubescence above,and with long hairs near the base; glabrous and grey in the second year.Leaves (Plate 202, Fig. 13), pendulous, oblong-lanceolate, gradually taperingand unequal at the base, long acuminate at the apex, with about twenty pairs ofparallel nerves, raised on the under surface, each ending in a triangular tooth, whichis prolonged into a fine point; upper surface dull, dark green, glabrous ; lower surfacelighter green, glabrous, or with minute scattered hairs, thin but firm in texture.Petiole, \ to f inch long, glabrous. Stipules, ovate-lanceolate, acute, puberulous,about \ inch long.Nut, 1 usually much compressed, \ to i inch wide, gradually acuminate at theapex; two to three fruits together in each involucre.This species is distinguished from the European one by the leaves being alwayscuneate and never cordate at the base, and never having any stellate tomentum, theunder surface being either glabrous or covered with minute glandular hairs.In winter it is readily distinguishable from C. sativa by the glabrous twigs andthe more pointed ovoid buds, which have glabrous ciliate scales as in that species.The buds are smaller than in C. sativa, being only about ^ inch long. In thespecimens seen the twigs are much more slender, with very minute lenticels and smallsemicircular leaf-scars. (A. H.)In America the chestnut is a common tree, and has a wide range from NewEngland and southern Ontario southward along the Alleghany Mountains to centralAlabama and Mississippi, and westward to Michigan, Indiana, central Kentucky,and Tennessee. So far as I have seen it does not attain so large a size as theEuropean species, though Sargent says it occasionally reaches 100 feet in height. 2The largest I saw was a fine old tree on the lawn of Mr. Nathaniel Thayer'shouse at Lancaster, Mass., which was 80 feet by 13 feet 6 inches, and thoughrather decayed at the top, where its branches were supported by iron stays, hadproduced suckers from the root, 40 feet high.1 The seedling of this species is described and figured by Rowlee and Hastings, in Bot. Gazette, xxvi. 351, fig. 18 ([898).2 In U.S. Forest Service, Circ. 7 1 (1907), a leaflet on the cultivation of this species, it is stated that the tree has beenknown in the region of its best development to reach a height of 120 feet. Throughout the greatest part of its range, it ismuch smaller, with an average height of 80 to 100 feet, and a diameter of from 2 to 4 feet.Castanea 857Emerson mentions a tree at Bolton, Mass., which in 1840 was 15^ feet in girthat 6 feet, with an unbranched trunk 24 feet long; and another, on the road toSheffield, which was 21 feet in girth at 4 feet from the ground. He states thatthough near the coast it does not ripen fruit so well, yet that in the interior whengrowing in sunny places it yields abundance of sweet and delicious nuts; andaccording to Sargent these, though smaller than European chestnuts, are superiorto them in sweetness and flavour, and are sold for food in the eastern cities.In Garden and Forest 1 there are several pictures of the chestnut in America,one representing a large tree at Dauphin in Pennsylvania, which is about 6 feet indiameter. Another represents a young forest in West Virginia about forty years old,showing good natural reproduction. A tree on a farm belonging to D. M. Ridgely,near Dover in Delaware, is noted for its excellent fruit, and it has been propagated,the chestnuts being known as Ridgely or Dupont chestnuts.In the United States Bureau of Forestry Bulletin No. 5 3 (1904). R- Zon givesan interesting account of the chestnut tree in Maryland, where it is an importanttimber tree, being used for railway sleepers, telegraph poles, and fencing. It isusually coppiced, and Zon states that the sprouts usually come from the root collar,only 10 per cent, arising from the top of the stump. He has never seen any suckershoots. The capacity of sprouting from the stool is retained to an advanced age,over 100 years. The tree in America usually becomes unsound at about 100years old.The American chestnut has rarely been tried in cultivation in Europe, andthough not likely to succeed so well as the common species, there are thrivingyoung trees at Kew.Emerson states that the timber is one of the best native woods on which to laymahogany veneers ; and Mr. Weale informs me that it is now imported into Englandboth in logs and boards ; but the demand is not very great. It is used by buildersas a substitute for oak, and by cabinetmakers. It carves well, and as it fumes readily,is a favourite wood with makers of antique furniture. In the log its value is fromis. 6d. to is. gd. per cube foot in Liverpool. In the board it is worth from 2S. to2S. 6d. After conversion it cannot be distinguished from the English-grownchestnut. Hough states that the wood is rich in tannin, which is extracted andused for tanning purposes. (H. J. E.)CASTANEA PUMILA, CHINQUAPINCastanea pumila, Miller, Diet., ed. 8, No. 2 (1768); Loudon, Arb. et Frut. Brit. iii. 2002 (1838);Sargent, Silva N. Amer., ix. 17, tt. 442, 443 (1896), and Trees N. Amer., 2 21 (1905).Fagus pumila, Linnaeus, Sp. PI. 9 98 (1753).A tree, rarely attaining in America 50 feet in height and 9 feet in girth,usually much smaller. Bark light-brown, slightly furrowed and broken on the surface1 Garden mid Forest, ix. 114, f. 12, 234, f. 34 (1896), and vii. 484 (1894). Cf. also Ibid. x. 372, f. 48 (1897).IV U

858 The Trees of Great Britain and Irelandinto loose plate-like scales. Young branchlets covered with numerous long erecthairs, becoming grey and glabrous in the second year.Leaves (Plate 202, Fig. 12) smaller than in C. sativa, rarely as much as 5 incheslong, not pendulous, oblong-oval, base unequal and rounded or tapering, apex acute,with about 15 pairs of nerves, which end in triangular serrations, tipped by shortspine-like points ; upper surface dull dark green, minutely pubescent; lower surfacegreyish white and densely tomentose. Petioles short, \ inch long, pubescent.Stipules about \ inch long, pubescent, those of the two lowest leaves broad, ovate,acute, on the middle leaves ovate-lanceolate, towards the top of the branch linear.Nut ovoid, rounded at the slightly narrowed base, gradually narrowed andpointed at the apex, f to i inch long, \ inch broad ; only one fruit in each involucre,which opens generally by two or three valves. The fruit, 1 which is ripe in Americain September, is delicious in flavour, and is occasionally gathered for market.Castanea pumila is distinguished from the other species by its smaller leaves,which remain densely whitish tomentose underneath and have fewer nerves. Inwinter it is distinguished from the common chestnut by the twigs being slender andhaving a scattered loose pubescence, especially marked towards their apex. Thebuds are ovoid, not acute at the apex, minute, about ^ to ^ inch long, with boththe first and second scales appressed-pubescent and ciliate. The leaf-scars andstipule-scars are smaller than in C. sativa.Castanea pumila? according to Sargent, occurs on dry, sandy ridges, rich hillsides, and the borders of swamps, from southern Pennsylvania to northern Floridaand the valley of the Neches River, Texas. It is usually shrubby east of theAlleghany Mountains, becoming a tree west of the Mississippi River, and is mostabundant and largest in size in southern Arkansas and eastern Texas.8 The wood *is similar to that of C. crenata, with very thin sapwood, and is used for fences, posts,railway sleepers, etc.According to Loudon it was introduced in 1699 by the Duchess of Beaufort,but it is extremely rare in cultivation, the only specimens which we have seen beingsmall shrubs at Kew, which, however, seem perfectly hardy.There are two specimens at Verrieres, 5 near Paris, the smaller of which hasa curiously twisted stem and resembles in appearance a dwarf Japanese tree. Theother has two stems, each about 28 inches in girth and about 18 feet high, andproduces fruit regularly and often in great abundance. 6 (A. H.)1 Hough, Trees of N. States and Canada, 1 37 (1907)." According to Taylor, in Bailey, Cycl. Anier. Hort. \. 295 (1900), this species commonly throws up root-suckers.3 Castanea neglecta, Dode, in Bull. Soc. Dmdr. France, 1 908, p. 155, said by Dode to occur in the eastern part ofthe United States, is apparently only distinguishable from C. pumila by its larger and less pubescent leaves. It is possibly,as this author points out, a hybrid between C. dentata and C. ftimila.4 Hough, loc. fit. 6 Hortus Vilmoriamis, 5 5 (1906).6 Since tins article was corrected for the press, a leaflet has been issued by the U.S. Forest Service, on Chestnut Barkdisease, which is caused by a fungus, known as Diaporthe parasitica or Valsonectria parasitica. This has recently destroyedan immense number of trees in the north-eastern states, spreading with great rapidity. As the disease, if once introduced,may be equally destructive in Europe, we think it well to warn arboriculturists against importing American chestnuts atpresent.FRAXINUSFraxinus, Linnaeus, Gen. PI. 3 18 (1737); Bentham et Hooker, Gen. PI. ii. 676 (1876) ; Wenzig,in Engler, Bot. Jahrb., iv. 165 (1883) ; Lingelsheim, in Engler, Bot. Jahrb. xl. 185 (1907).TREES or shrubs, belonging to the natural order Oleaceae; leaves opposite, compound, unequally pinnate, rarely reduced to a single leaflet; stipules absent. Buds,large terminal and small axillary, the former usually with four scales visible externally, the latter with two outer scales; these scales are rudimentary leaf-stalks,often showing at their apex traces of the pinnate leaf, and increase in size after thebud opens, falling off eventually and leaving ring-like scars at the base of theshoots.Flowers polygamous or dioecious, in panicles or fascicled racemes, terminal onleafy shoots of the year, or developed from separate buds either in the axils of theleaf-scars of the previous year, or at the base of the young branchlets. Calyx absentin some species ; when present, campanulate and four-lobed. Corolla absent in manyspecies; when present, of two to four (rarely five to six) petals, free or connate inpairs at the base. Stamens two, rarely three or four, affixed to the base of thepetals or hypogynous. Ovary, with a style divided above into a two-lobed stigma,two-celled, each cell containing two pendulous ovules. Fruit, a samara, indehiscent,convex or compressed below, with a dry pericarp produced into an elongatedterminal and more or less decurrent wing, 1 usually one-celled and one-seeded. Seedpendulous ; embryo erect in a fleshy albumen; cotyledons flat.The genus Fraxinus is widely distributed over the temperate regions of thenorthern hemisphere, three 2 species, however, occurring within the tropics in Cubaand the Philippines, and south of the equator in Java. The genus consists ofnearly sixty species, many of which are imperfectly known and require furtherstudy in the field. Even in the case of the Mediterranean species, authorities are atvariance. The present account deals only with the species which have been seenin the living state.The genus is divided into five sections :I. Ornus, Persoon, Syn. PI. ii. 605 (1807).Calyx and corolla both present, the calyx persisting under the samara.Panicles terminal on leafy shoots or axillary on the branchlets of the currentyear. About eighteen species.1 Abnormal fruit with three wings, has been observed in several species, as f. americana, f. caroliniana, f. Berlandieriana.2 F. caroliniana, a native of the United States, is met with in Cuba. f . Eedenii, Boerl et Koord, occurs in Java ; andF. philippinensis, Merrill, in the Philippine Islands.859

860 The Trees of Great Britain and Ireland Fraxinus 861II. Ornaster, Koehne and Lingelsheim, Mitt. Deut. Dendr. Gesell. 1 906, p. 66.Calyx present, persistent under the samara. Corolla absent. Flowers interminal panicles, appearing with the leaves. Seven species.III. Sciadanthus, Cosson et Durieu, Bull. Soc. Bot. France, ii. 367 (1855).Calyx present, persistent under the samara. Corolla absent. Flowers indense fascicled cymes, axillary on the preceding year's shoot. Two species.IV. Leptalix, Rafinesque, New Flora, iii. 93 (1836).Calyx present, persistent under the samara. Corolla absent. Flowers inpanicles, axillary on the preceding year's shoot. About fifteen species.V. Fraxinaster, De Candolle, Prod. viii. 276 (in part) (1844).Calyx and corolla both absent. Flowers in panicles or racemes on the preceding year's shoot. About twelve species.These sections, based on the characters of the flowers, are not available inpractice in the determination of living trees, flowering specimens of which are oftennot obtainable; and the following key groups the species according to the charactersof the branchlets and foliage :KEY TO THE SPECIES IN CULTIVATIONI. Leaves simple or -with two to three leaflets.* Branchlets four-angled.1. Fraxinus anomala, Watson. Colorada, Utah, Nevada. Seep. 898.Leaves usually simple, ovate or obovate, glabrous beneath.** Branches terete.2. Fraximis angustifolia, Vahl., var. monophylla. See p. 880.Leaves opposite, simple or two- to three - foliolate, lanceolate, glabrousbeneath.3. Fraxinus excelsior, Linnaeus, var. monophylla. See p. 866.Leaves opposite, simple or two- to three-foliolate, ovate or oval, pubescentbeneath at the base.4. Fraxinus syriaca, Boissier. Western Asia. See p. 883.Leaves in whorls. Leaflets usually three (occasionally five to seven occurringon the same branch), lanceolate, glabrous.II. Leaves with five or more leafletsÂ. Branchlets, leaf-rachis, and leaflets quite glabrous.* Leaflets stalked.5. Fraxinus potamophila, Herder. Turkestan. See p. 885.Leaflets seven to nine, ovate, serrate.1 Cf. F. syriaca (No. 4), which has occasionally five to seven leaflets.* * Leaflets sessile.6. Fraxinus angustifolia, Vahl. S. France, Spain, Portugal, N. Africa. See p. 879.Leaflets seven to thirteen, lanceolate. Leaf-rachis strongly winged, thewings meeting above ; groove interrupted.7. Fraxinus Willdenowiana, Koehne. Origin unknown. See p. 884.Leaflets, seven to eleven, ovate or lanceolate, increasing markedly in sizefrom the base to the apex of the leaf. Leaf-rachis with a continuous opengroove.B. Branchlets glabrous ; leaflets pubescent on part of the lower surface.* Leaf-rachis strongly winged on the upper side, the wings meetingin part above, forming an interrupted open groove.8 to 11. Leaf-rachis not conspicuously bearded at the nodes.8. Fraxinus oxycarpa, Willdenow. Italy, S.E. Europe, Asia Minor, Caucasus.See p. 882.Leaflets seven to thirteen, ovate or lanceolate; serrations few, ending in longincurved points.9. Fraxinus excelsior, Linnaeus. Europe, Caucasus. See p. 864.Leaflets, five to eleven, oblong-lanceolate; serrations crenate, numerous,exceeding in number the lateral nerves.10. Fraxinus excelsior, Linnaeus, var. rotundifolia. See p. 866.Leaflets nine to thirteen, i^ to 2^ inches long, ovate, oval, or orbicular,coarsely bi-serrate.n. Fraxinus Elonza, Dippel. 1 Origin unknown. Seep. 883.Leaflets, eleven to thirteen, small, less than 2\ inches long, irregularly serrate,oblong, lanceolate or oval; under surface with brown tomentum near the base.12, 13. Leaf-rachis with conspicuous tufts of brownish-red tomentumat the nodes.12. Fraxinus nigra, Marshall. N. America. Seep. 898.Leaflets, seven to eleven, oblong-lanceolate, rounded or broadly cuneate atthe base, sessile.13. Fraxinus mandshurica, Ruprecht. Eastern Asia. See p. 893.Leaflets seven to thirteen, oblong-lanceolate, gradually tapering at the base,sub-sessile.* * Leaf-rachis with a continuous open groove on its iipper side, whichis sometimes almost obsolete.f Some or all of the leaflets distinctly stalked.14, 15. Leaflets white beneath.14. Fraxinus americana, Linnaeus. N. America. See p. 901.Leaflets seven to nine, 4 to 6 inches long, long-acuminate, dull light greenabove ; rachis with an extremely slight groove.1 The groove on the leaf-rachis is variable in this species, sometimes being open its whole length.

862 The Trees of Great Britain and Ireland15. Fraxinus texensis, Sargent. Texas. See p. 907.Leaflets five to nine, z\ to 3^ inches long, shortly acuminate, shiningbluish-green above; rachis with a very slight groove.16. Leaflets green beneath; rackis glabrous?16. Fraxinus caroliniana, M iller. 2 S.E. United States, Cuba. See p. 904.Leaflets, five to seven, about 3 inches long, shortly acuminate; rachis with awell-defined but shallow groove.17 to 23. Leaflets green beneath ; rachis slightly pubescent at the nodes.17. Fraxinus rhynchophylla, Hance. N. China, Manchuria. Seep. 892.Leaflets five to seven, 3 to 4 inches long, coriaceous, terminating in an obtusetippedacumen, entire or very obscurely serrate.18. Fraxinus chinensis, Roxburgh. Central and Southern China. See p. 895.Leaflets seven to nine, 3 to 4 inches long, coriaceous, shortly cuspidate at theapex, crenately serrate.19. Fraxinus obovata, Blume. Japan. See p. 895.Leaflets five to seven, 2 to 3 inches long, membranous, variable in shape,irregularly serrate, with minute curved bristles on the lower surface and petiolules,which are also present on the rachis of the leaf.20. Fraxinus longicuspis, Blume. Japan. See p. 897.Leaflets five, 3 to 4 inches long, membranous, very pale beneath, abruptlycontracted into a long cuspidate apex, crenately serrate.21. Fraxinus Ornus, Linnaeus. S. Europe, Asia Minor. See p. 887.Leaflets five to nine, 2 to 3 inches long, membranous, shortly acuminate,serrate.22. Fraxinus floribunda, Wallich. Himalayas, Upper Burma. See p. 890.Leaflets seven to nine, 4 to 6 inches long, membranous, apex long-acuminate,serrate; lateral nerves prominent and numerous.23. Fraxinus quadrangulata, Michaux. N. America. See p. 900.Branchlets quadrangular and four-winged. Leaflets seven to nine, 3 to 5 incheslong.tt Leaflets sessile or sub-sessile?24. Fraxinus Spaethiana, Lingelsheim. Japan. See p. 897.Leaflets seven to nine, 4 to 6 inches long, coriaceous, lanceolate, longacuminate,irregularly and often crenately serrate. Distinguished from allother species in cultivation by the dilated swollen base of the leaf-stalk.25. Fraxinus lanceolata, Borkhausen. N. America. See p. 906.Leaflets seven to nine, 3 to 6 inches long, lanceolate, long-acuminate. Rachisgrooved.26. Fraxinus Berlandieriana, De Candolle. Texas, Mexico. See p. 907, note i.Leaflets, five to seven, 2 to 3 inches long, oval or obovate. Rachis grooved.1 Fraxinus lanceolata (cf. No. 21), has occasionally the leaflets distinctly stalked, and might on that account be soughtfor here.2 Cf. No. 31 A. Two forms of this species occur in cultivation, differing in the absence or presence of pubescence on thebranchlets and leaf-rachis.3 Fraxinus Elonza (cf. No. II), sometimes having an open continuous groove on the rachis, might be sought for here.Fraxinus 86327. Fraxinus dimorpha, Cosson et Durieu. N. Africa. See p. 884.Leaflets, seven to nine, about f inch long, ovate. Rachis with wide-spreadingwings.C. Branchlets minutely pubescent; leaflets glabrous.28. Fraxinus Mariesii, J . D. Hooker. Central China. See p. 892.Leaflets five, coriaceous, about 2^ inches long, oval, stalked, crenately serrate.29. Fraximis Bungeana, De Candolle. North China, See p. 891.Leaflets five to seven, membranous, about i^ inch long, mostly stalked, ovalor rhomboid, long-acuminate, irregularly serrate.30. Fraxinus raibocarpa, Regel. Turkestan. See p. 886.Leaflets five to seven, oval, entire, about i^ inch long; upper leaflets subsessile,lower leaflets stalked.D. Branchlets, leaf-rachis, and leaflets pubescent.* Leaflets distinctly stalked.31. Fraxinus Biltmoreana, Beadle. United States. Seep. 905.Leaflets seven to nine, oval, about 4 inches long, white beneath ; rachis veryslightly grooved.31 A. Fraxinus caroliniana, Miller. 1 S.E. United States, Cuba. See p. 912.Leaflets five to seven, oval, about 3 inches long, green beneath, rounded orbroadly cuneate at the base ; rachis with a well-defined but shallow groove.* * Leaflets stalked, sessile, or sub-sessile.32. Fraxinuspennsylvanica, Marshall. N. America. See p. 908.Leaflets, seven to nine, lanceolate, 4 to 5 inches long, green beneath, pubescenton both surfaces, long-acuminate, tapering at the base; rachis densely whitepubescent,and with a narrow, shallow groove. Buds reddish pubescent.33. Fraxinuspubinervis, Blume. Japan. See p. 896.Leaflets five to seven, lanceolate, 3 to 4 inches long, glabrous above, greenand pubescent beneath, acuminate, tapering at the base; rachis grooved, withpubescence densest at the nodes. Buds greyish pubescent.* * * Leaflets sessile or sub-sessile.34. Fraxinus oregona, Nuttall. Western United States. See p. 910.Leaflets seven to nine, oval, 3 to 4 inches long, green beneath, shortlyacuminate, entire or obscurely crenate in margin.35. Fraxinus velutina, Torrey. Texas to California. See p. 912.Leaflets three to five, about i£ inch long; lateral leaflets variable in shapeand serration, terminal leaflet obovate.36. Fraxinus xanthoxyloides, Wallich. Baluchistan, Afghanistan, N.W. Himalayas.See p. 885.Leaflets, five to nine, about f inch long, ovate. Rachis with wide-spreadingwings.» Cf. No. 16.

864 The Trees of Great Britain and Ireland37. Fraxinus holotricha, Koehne. Origin unknown. See p. 887.Leaflets nine to thirteen, lanceolate or ovate-lanceolate, about 2 inches long,sharply serrate.FRAXINUS EXCELSIOR, COMMON ASH(A. H.)Fraxinus excelsior, Linnaeus, Sp. PI. 1 057 (1753); Loudon, Arb. et Frut. Brit. ii. 1214 (1838);Willkomm, Forstliche Flora, 6 58 (1897); Mathieu, Flore Forestiere, 2 41 (1897).A large tree, attaining 140 feet in height. Bark smooth and greyish whenyoung, becoming rough and fissured in old trees. Branchlets glabrous. Leaflets(Plate 262, Fig. 4), 9 to 15, sessile and articulate, oval- or oblong-lanceolate,acuminate at the apex, tapering at the base, where the margin is entire, elsewherecrenately serrate, the serrations more numerous than the lateral nerves; uppersurface glabrous and green; lower surface paler with pubescence on the midrib,extending over the basal part of the leaflet; venation pinnate, the lateral nervesforming loops near the margin. Rachis glabrous or pubescent, strongly winged,the wings meeting above, 1 except opposite the insertion of the leaflets where thereis an open channel, and below the leaflets where the rachis is flattened or broadlygrooved.Flowers,2 opening before the leaves appear, fertilised by the wind, in denseaxillary panicles, polygamous or occasionally dioecious, without calyx or corolla.Male flowers with two stamens more or less connate below. Female flowers witha two-celled superior ovary, the style being dilated above into two thick stigmas.Perfect flowers with an ovary and two stamens.3Fruit, of two carpels, joined together to form the body of the samara, whichis compressed at right angles to the partition and is produced in front into aveined membranous wing. The samarae are very variable in shape, but are usuallylinear-oblong or elliptic, obtuse at both ends, and notched at the tip. They hang inracemes on long stalks, and, ripening in autumn, generally remain on the tree till thefollowing spring; and are ultimately carried by the wind a short distance away fromthe parent tree.SEEDLING 4The young plant on appearing raises the samara out of the soil, the twocotyledons being united together at first by a cap formed of the albumen. The1 Rain collecting on the leaflets drains into the ducts thus formed, inside of which are hairs and peltate groups of cellsthat gradually absorb the water, which is retained for several days after a fall of rain. See Kemer, Nat. Hist. Plants, Eng.Transl. i. 231, fig. 54 (1898).2 Section Fraxiiiaster.3 Schulz, in Ber. Deutsch. Bat. Ges. x. 401 (1892), has shown that trees of the common ash greatly vary in the kind offlowers which they produce. Trees bearing only male flowers are common ; while those with only female flowers or withonly perfect flowers are rare. In many cases two of the three kinds of flowers are borne on the same tree ; and what is veryremarkable, a tree is not necessarily of the same sex in successive years. Ash trees do not flower, as a rule, regularly everyyear ; and fruit is much more abundant in some years than in others.* Figured in Lubbock, Seedlings, ii. 214, fig. 512 (1892).Fraxinuscotyledons, when developed, are about f to i^ inch long, oblong, obtuse, entire,glabrous, pale beneath, tapering at the base into a very short winged petiole. Caulicleterete, i to 3 inches long, ending in a long yellow, fleshy, flexuose tap-root. Youngstem, green and glabrous, terete below, angled above. First pair of leaves, arising£ to i inch above the cotyledons, simple, ovate, acuminate or acute, irregularlyserrate and ciliate, minutely pubescent, on a winged petiole about ^ inch long.Second pair of leaves, three-foliolate, on a petiole about an inch long, the terminalleaflet the largest. Third pair with either three or five leaflets.IDENTIFICATIONThe common ash is only liable to be confused with species like F. angustifoliaand F. oxycarpa ; but is readily distinguished by its black buds, and the crenateserrations more numerous than the lateral nerves in the leaflets.In winter, the twigs are stout, shining-grey or olive green, compressed towardsthe tip, swollen at the nodes. Leaf-scars, opposite, obliquely set on projectingpulvini, semicircular or almost orbicular, often with lateral projecting horns, andshowing an almost circular row of bundle-dots. Terminal buds black, conical,quadrate, with four scales visible externally, but consisting altogether of seven toeight pairs of scales. Lateral buds smaller, given off at a wide angle, with two orthree external scales.VARIETIESThe common ash, though distributed over a wide area, varies little in the wildstate; and such varieties, as have been based on the form of the fruit, cannot beconsidered as well established. Near Perpignan a form with small leaves has beencollected, which is var. australis, Godron et Grenier, Flore de France, ii. 471. Inthe province of Talysch in the Caucasus, a remarkable form occurs with largeleaflets, velvety pubescent underneath; and the shoots, buds, and leaf-rachis aredensely pubescent. This variety, which was described by Scheele 1 as a distinctspecies (F. coriarcefolia), is said by Koch 2 to be met with occasionally in cultivationin gardens, where it is known as Fraxinus expansa.A curious variation in the common ash was observed by A. D. Richardson 3 inthe case of four young plants, found growing in a clump of several hundreds, onthe banks of the Boyne near Navan in Co. Meath. The leader shoots had theleaves alternate in a 2/5 spiral arrangement, instead of the normal opposite anddecussate one.Numerous varieties have been obtained as seedlings in nurseries or as isolatedspecimens growing wild.1 Linniza, xvii. 350 (1843). 2 Dendrologie, ii. I, 243 (1872). 3 Card. Chron. xxxvi. 133, fig. 55 (1904).IV

866 The Trees of Great Britain and Ireland1. Var. monophylla, O. Kuntze, Flora von Leipzig, 8 2 (1867).Fraxinns monophylla, Desfontaines, Tab. de FJScole de Sot. 5 2 (1804).Fraxinus heterophylla, Vahl. Enum. PI. i. 53 (1804); Loudon, Arl>. et Frut. Brit. ii. 1228 (1838).Fraxinus simplirifolia, Willdenow, Sp. PI. iv. 1098 (1805).Fraxinus excelsior, var. diversifolia, Alton, Hort. Kew. iii. 445 (1789).This remarkable variety, which is usually known as the laurel-leaved or simpleleavedash, is met with in the wild state in the forests near Nancy in France, andalso rarely in England and Ireland. It appears 1 occasionally when a quantity of ashseeds is sown, and intermediate forms are found with three, five, and seven leaflets.The shoots and buds are identical with those of the common ash. In the ordinaryform of the variety, the leaf (Plate 262, Fig. 2) is simple, not being divided intoleaflets. Occasionally there is a large leaflet, with one or two small leaflets at itsbase; and this form is known as var. monophylla laciniata. The simple - leavedform or the terminal leaflet in the two- to three-leaflet form, has a stalk about halfas long as the blade or a little longer, and is variable in shape, being usually oval inoutline with an obtuse, acute, or acuminate apex; margin coarsely serrate; lowersurface pubescent except towards the apex; petiole widely grooved on its upperside. A form of the single - leaved ash with variegated leaves, was discovered,according to Loudon, in 1830 at Eglantine, near Hillsborough, Co. Down, Ireland.The simple-leaved ash is very distinct in appearance and thrives well in towns.It is usually propagated by grafting.At Beauport, Sussex, there is a tree of this kind, 70 feet by 5 feet 9 inches;and self-sown seedlings reproducing the variety have been observed by us there.Other large specimens occur : at Syon, a tree 2 84 feet by 7 feet 6 inches ; at Sidmouth,measured by Miss Woolward in 1904, two trees 9 feet 4 inches each in girth, thetaller being 86 feet high ; also three good trees in the grounds at Woburn growingbeside the lake. Lord Kesteven reports one at Stubton Hall, Newark, which was 67feet high by 8 feet i inch in girth in 1906. Elwes has seen others from 50 to 70 feethigh at Scampston Hall, Yorkshire; at Sharpham, near Totnes; and at DodingtonPark, Gloucestershire. A tree at Oxford, near the east end of the broad walk inChristchurch meadow, mentioned by Walker 3 in 1833, is, according to Mr. Druce,about 65 feet high and 4 feet 7 inches in girth. It is crowded by other trees and isnot thriving.On the Pitfour estate near Mintlaw in Aberdeenshire, a tree 55 feet high by7 feet 9 inches is reported by Mr. Ainslie; and Elwes saw one at Gordon Castle,which in July 1907 was covered with fruit and measured about 60 feet by 9feet 2 inches in girth.There is a very good specimen in Stephen's Green, Dublin; one at BeauparcHouse in Co. Meath measured, in 1905, 40 feet high by 6 feet 2 inches in girth;and another at Curraghmore, Co. Waterford, was 50 feet by 5 feet in 1907.2. Var. rotundifolia. A tree growing in a wood at Strete Ralegh, near Exeter,1 Cf. Mathieu, Flare Forestitre, 241 (1897), and Allgem. Garten-Zeitung, iii. 6 (cited by Loudon).2 This tree in 1849 was 50 feet high by 4 feet in girth; it is now beginning to decay at the top.3 Flora of Oxfordshire, 3 (1833). Cf. also Dyer, injourn. Bot. ix. 147 (1871).Fraxinus 867the seat of H. Imbert Terry, Esq., who has sent specimens, is remarkably distinctin the shape of the foliage from any ash known to me; but is probably onlya variety of the common ash, with which it agrees in bark and in buds. In theabsence of flowers and fruit, this identification is not quite certain, and on thataccount a full description is now given: Leaflets (Plate 266, Fig. 32), nine tothirteen, i^ to 2^ inches long; terminal leaflet stalked, with a long cuneate base;lateral leaflets, sessile, broadly oval or ovate, unequal at the usually cuneate, butoccasionally broad and rounded base, acute or slightly acuminate at the apex,coarsely bi-serrate, slightly scabrous with scattered stiff hairs on the upper surface,pale beneath with dense woolly pubescence on the sides of the midrib and lateralnerves near the base. Leaf-rachis, strongly winged, the wings meeting above in itsapical half, but forming a wide open groove towards the base; pubescent on thedorsal side with scattered stiff hairs, densest at the nodes.This ash resembles in foliage the figure of F. rotundifolia, Aiton, 1 which isgiven by Willdenow.2 The latter species, according to Aiton, Willdenow, andLoudon, 3 is a small tree of Italy, with flowers and buds like F. Ornus; and theStrete Ralegh tree cannot be identified with it, as in all essential characters 4 itresembles the common ash.Nothing is known of the origin of the tree at Strete Ralegh, which MissWoolward found in 1905 to be about 75 feet in height, the bole dividing near theground into two stems, 3 feet i inch and 2 feet 7 inches in girth respectively.3. Var. angustifolia, Schelle. A variety 5 with small narrow leaves (Plate 262,Fig- 5)i which differs in no essential character from the common ash, of which ithas the buds and the characteristic serrations and pubescence of the leaflets; and inthis way can be readily distinguished from such species as F. angiistifolia, Vahl, andF. oxycarpa, Willd.4. Var. crispa, Loudon (also known in gardens as var. atrovirens and var.cucullatd). Leaflets dark green, curled and twisted. Plant usually rigid andstunted, of very slow growth.5. Var. nana, Loudon (also known in gardens as var. pokmoniifolia and var.globosa). A compact slow-growing dwarf form, with very small leaves.6. Var. aurea, Loudon. With yellow branches. A pendulous form of this isknown.7. Var. asplenifolia, Koch. Leaflets very narrow, almost linear.8. Var. fungosa, Loddiges. Bark remarkably wrinkled, with corky ridges.9. Var. verticillata, Loudon. Leaves whorled, not opposite as in the commonform.10. Var. monstrosa, Koch. Young branches fasciated.11. Var. erosa, Persoon. Leaflets incised.1 Hort. fCcw. iii. 445 (1789). Cf. our remarks on this species under F. Ornus, p . 888.2 Berlin. Baumzucht, 1 16, fig. vi. 1 (1796). 3 Art. et Frut. Brit. ii. 1244 (1838).4 Bark and buds especially. The pale uuder surface of the leaf, which is thin in texture, is seen in common ash seedlingsand in some forms of var. nwnophylla. The strongly-winged rachis of the leaf is characteristic of F. excelsior and itsnear allies.6 Var. elegantissima, in cultivation at Aldenham, obtained from Simon-Louis, is scarcely to be distinguished from thisvariety.

868 The Trees of Great Britain and Ireland12. Var. verrucosa, Desfontaines. Branchlets warty.13. Certain variegated forms are known, as var. albo-marginata, leaflets edgedwith white, and var. albo-variegata, in which the white colour appears as blotches onthe leaflets.Most of the foregoing varieties are of little or no beauty or interest, and do not,so far as we know, become large or shapely trees. (A. H.)14. Vsx.pendula, Aiton. The weeping ash in some form or other is found inalmost every garden, but rarely as a large tree. Loudon describes several forms ofit, and says that the original tree was discovered near Wimpole in Cambridgeshire150 years or more ago, and was decaying in 1835.Another form, the Cowpen ash, which grew near Morpeth, is figured byLoudon; * and I have seen two trees which have naturally assumed a very similarhabit. One stands by the road in the village of Ollerton, Notts. The other is in afield at Marsden, in the parish of Rendcombe, Gloucestershire.A third form, called by Loudon the Kincairney Ash, grew in the parish ofCaputh, near Dunkeld, Perthshire, and was distinguished by its alternately pendulousand upright branches. It was propagated at the Perth Nurseries, but I have notnoticed any of this variety now in cultivation.To make an effective tree, the weeping ash should be grafted on a very tallstock, and if the branches are attended to, may be trained into a shady arbour likea great umbrella. But if the stock is also allowed to grow as well as the graft, theeffect will be more curious than beautiful; and the weeping ash is not so muchadmired or planted as it was formerly when trained and clipped trees were morein fashion. At Heanton Satchville, in North Devon, the seat of Lord Clinton, Isaw it trained in combination with a trellis of living ashes which were planted allround the central weeping tree, and had their stems woven together when youngso as to form the walls of the arbour; but in the course of time this had becomeragged ; and as the ash does not bear clipping like the beech or hornbeam, I shouldprefer either of those trees for such a purpose.By far the finest grafted weeping ash that we know of is growing in the gardensof Elvaston Castle, near Derby, the seat of the Earl of Harrington. It was reported 2in 1905 to be 98 feet high, with long weeping branches hanging vertically from thesummit of the tree, one of them descending to about 20 feet from the ground;but when I saw it in 1906 I did not think it was more than about 90 feet high,the bole, 6 to 7 feet in girth, being straight and clean. (Plate 238.) This treewas grafted by Barren about 1848. Another larger tree also exists here, which hasa bole 50 feet by 12^ feet, and was apparently grafted with weeping ash at thesame time, but in this case the branches of the stock have outgrown the grafts.In Ireland there is a very handsome and well shaped weeping ash at Castlewellan,41 feet high, with a trunk 5 feet in girth with branches hanging to the groundall round it. (H. J. E.)1 Arb. et Frut. Brit. ii. 1216, f. 1045 (1838). 1 Garden, 1 905, Ixviii. 400, with figure.Fraxinus 869DISTRIBUTIONThe common ash is spread through almost all Europe, and occurs also in theCaucasus.The northern limit in Europe passes from the Trondhjem fjord in Norway,about lat. 63 40', through Sweden at about lat. 61 , and in Finland extends to lat.62 , descending from there through Russia in a S.E. direction to Riazan, whence itcontinues N.N.E. to Kazan, its extreme eastern point in Europe. From Kazan, theeastern limit descends in a S.W. direction through Penza, Saratof, and Voronej toKharkof, and then passes by Ekaterinoslav to the Crimea. In the Caucasus 1 theash does not occur north of the rivers Kuban and Terek. The southern limit extendsfrom the Transcaucasian province of Talysch through Asia Minor and Turkeyto Dalmatia, and across Italy and southern France to the Pyrenees. In the Iberianpeninsula 2 it is met with, according to Willkomm, in the mountains of Catalonia,Aragon, Burgos, Santander, Leon, Asturias, Galicia, and northern Portugal. Thewestern limit takes in the western coast of France and the British Isles. Outsidethe range mentioned above, it occurs as small scrub in rare situations, as inNorway at Troinso (lat. 69 40'), and in the government of St. Petersburg.An ash occurs in the western Himalayan region which, according to most of theauthorities, is Fraxinus excelsior. It has been distinguished as a distinct species byWenzig, 3 and, so far as I can judge from dried specimens, is very different indeedfrom the common ash. Sir George Watt informs me that this ash is always aninsignificant tree, never attaining more than 30 feet in height and a foot in girth.The ash is generally met with growing wild as isolated trees or in small groupsin the continental forests, but pure woods of some extent occur in moist situations, asin river valleys subject to flooding, in Hungary, Slavonia, Poland, and Russia. Innorthern regions it is rather a tree of the plains and valleys than of the mountains;but in southern Europe it is only met with in the mountains. It ascends in southTyrol to 4000 feet, and in Switzerland to about 4400 feet.The ash is a true native of the British Isles, and has been found in a fossil statein the interglacial beds at Hitchin in Hertfordshire, and in neolithic deposits atCrossness in Essex. 4It may be said to occur wild in every part of the British Isles, except in thenorthern part of Scotland, where, however, it bears the climate in plantations. InYorkshire 6 it ascends to 1250 feet elevation. In Braemar, H. B. Watt 6 observed itup to 1200 feet. In Ireland 7 it is frequent in woods, hedges, and rocky places ; andascends in Donegal to 800 feet, in Down to 1000 feet, and in Wicklow to 1300 feet.1 It ascends in the Caucasus to 6000 feet, according to ^&A

870 The Trees of Great Britain and IrelandAsh woods, supposed to be wild, occur on limestone in hilly districts inYorkshire, Derbyshire, and Somerset. In the latter county, they are very numerousin the Mendip Hills, and have been mapped by Mr. C. E. Moss, 1 who gives an interesting account of their distribution and peculiar features. The ash is often pure, witha dense undergrowth of hazel, or it is mixed with yew and whitebeam. Mr. Mossnotices the prevalence of dog's mercury and wood garlic 2 in many of these ash woods.The Gaelic name of the ash, according to Sir H. Maxwell,3 is uinnse (inshy), andbecomes Inshawhill in Wigtonshire, and the plural uinnsean (inshan) takes thepeculiar form of Inshanks, the name of two places in that county, and Inshewan,near Kirriemuir; while the common alternative form uinnseog (inshog) remains asInshock in Forfarshire, Inshaig in Argyllshire, Inshog near Nairn. Analogous forms, 4with the initial letter^! appear in names of places in the south and west of Ireland,as the river Puncheon in Co. Cork. (A. H.)CULTIVATIONThough the oak will always be looked on as the premier tree of Great Britain,yet now that its most important use has passed away, the ash must be considered asthe most economically valuable of all our native trees, and is perhaps the onlyhardwood from which a quick and certain return can be expected by the planter.It is almost the only tree whose value has not fallen in consequence of foreigncompetition, and, though a good deal of American and some Hungarian and Japaneseash is now imported, yet the timber of these is not considered equal for toughness,strength, and elasticity to the best English ash, for which no foreign wood forms anefficient substitute. And as the tree can be grown over all parts of our islands, andattains a great size wherever suitable soil is found, it should be planted morelargely in all favourable situations, where it produces timber of good quality.In considering the requirements of the ash, one must always remember that it isa bad neighbour both to other trees and to crops, and that it is far more valuable astimber when grown in woods where it can be drawn up to a good height, than inhedgerows where it produces many branches. It likes a deep, rich soil, neither toowet nor too dry, and grows very well on limestone formations, even on a shallowsoil, if the rock is sufficiently disintegrated for the roots to penetrate the crevices.It is short-lived on wet or swampy soils, and the timber is inferior on sandy orpeaty land. The finest trees are generally in sheltered situations, but though itis the latest of our native trees to come into leaf, none suffers more from latefrosts; and therefore when planted in low situations it is often severely injuredwhen young. It will grow up to a great elevation, and in the most exposed situations, though here it becomes stunted and branchy. No hardwood except thatof the chestnut becomes valuable at so early an age, but the wood of old trees,even when sound, usually becomes discoloured or " black-hearted," and ash is never1 Geog. Distribution of Vegetation in Somerset, 41 (Roy. Geog. Soc., 1907).2 Mr. A. C. Forbes, in Eng. Estate Forestry, 7 2 (1904) says : "We have always noticed that the existence of the wildgarlic, Allium ursinum, is an almost certain indication of good ash ground."3 Scottish Land Names, 109 (1894). 4 Cf. Joyce, Irish Names of Places, 488 (1870).Fraxinus 871more valuable than when it is from 3 to 6 feet in girth, with a clean stem, a sizeoften attained at fifty to sixty years of age.In plantations ash is often mixed with other trees, and if allowed to take thelead will do them more harm than oak, but a few ash should be introduced inthe best soil of larch or other plantations, because the seedlings, which spring upabundantly, will, when the conifers are cut, renew the plantation naturally, and theparent trees will throw up vigorous shoots from the stool after felling.In the midland counties, ash is the commonest, and by far the most profitableunderwood, being cut at intervals of twelve to twenty years; when the poles aremuch in demand for many purposes, especially for sheep hurdles. But in mostplaces during the last twenty years ash poles have fallen in value, though largertimber has increased in price; and so much damage has been done to the stools byrabbits that large areas are now becoming very thin, and the crop inferior. No treeexcept beech suffers more from rabbits than ash, and where they are allowed toincrease, and are not killed before winter, the bark of old trees as well as of underwood is sure to be peeled, and the natural reproduction from seed checked. I believethat where the soil is stiff, young ash will pay for some cultivation when young, astheir shade is not dense enough to keep down grass and weeds, and if they becomestunted, as they often do after planting, it is better, and, indeed, necessary, to cutthem down to the ground two to five years later.Self-sown ash seem to grow more vigorously than planted ones, if not toocrowded, and their rate of growth is sometimes extraordinary. An ash self-sownin my nursery, at three years old was 7 to 8 feet high, whilst the transplantedseedlings on the same ground were only 3 feet high at the same age. I have seenshoots 6 to 7 feet long the first year from strong healthy stools, and poles worth ^15to ^"20 per acre at sixteen years old, on land which for agricultural purposes wasnot worth 8s. an acre. The stools, however, often become worn out and hollow atthe base after five or six cuttings, and these should be replaced with seedling plantsevery time the crop is cut.Some years ago, when ash coppice began to fall in price, I left the strongest andstraightest pole on every stool at the rate of about 160 per acre, with the object ofconverting the coppice into timber trees. But though, where the soil is good andthe stools young, these poles are likely to make useful trees at fifty to sixty years fromthe last cutting over, yet where the land is poor they have increased but little, andhave a hidebound appearance, owing, no doubt, to the want of shelter and theexposure to the sun and wind. I should advise all intending planters of ash toexamine carefully the best local ash plantations, and inquire into the probabledemand for poles before adopting this course.The ash is always raised from seed, except in the case of varieties which aregrafted on stocks of the common ash. The keys are ripe in late autumn, and oftenhang on the tree till the following summer, especially when they are not mature.The best-ripened seed, I believe, usually falls first, and should be gathered beforewinter, and put in a shallow layer mixed and covered with earth or sand, and keptfairly dry until the following winter, when it should be sown. It is advised in books

872 The Trees of Great Britain and Irelandthat they should be turned over several times during the year before sowing, but 1have not found this necessary; and with regard to the time of sowing, it should beput off as long as possible, because the natural germination of the seed takes placesix to eight weeks before the tree comes into leaf, and the tender seedlings are thusoften injured and killed by late frost. Therefore I advise storing them in a cold place,and not sowing until they begin to germinate. If they come up too thickly andsurvive the first spring, they may be transplanted in the following March or April,which will tend to check their early leafing, but if thin on the ground they may beallowed to stand two years before transplanting into rows. At three, or at most four,years old, they will be fit to go out permanently, the stronger side branches anddouble leads, if any, being first pruned. If intended for copsewood, they must becut over in the month of April, two or three years after planting, and any pruningnecessary to older trees should be done in summer or early autumn, so that thewounds may heal as soon as possible. The tree makes an abundance of fibrousroots, and unless these are allowed to become dry, the proportion of loss from transplanting should be very small, and transplanting may be done later than in the caseof most hard woods.For ash coppice, 4 or 5 feet apart is the right distance; for timber trees, theymay be alternated with spruce or larch, which will keep them from becomingbranchy. The cutting of the stools must be done with a sharp knife or axe as nearthe ground as possible, and with an upward cut, and the poles removed at latest bythe middle of May, as much harm is caused by getting the poles away after thestools have begun to push new growth.One of the best examples of copse-grown ash that I have seen in England isthe Walk Copse near Buckhold, Berks, where a number of tall, slender, clean poles,believed by Dr. Watney to be about sixty years old, have originally sprung up fromseed, in a plantation once largely composed of silver firs. Though the soil is a flintyclay, now of little agricultural value, the majority of these trees are 90 to 100 feethigh, by 3 to 4 feet in girth, and quite clean to 50 or 60 feet. One of the best wasquite straight and clean to 65 feet high, but only about 3 feet in girth. Such polesas these are much sought for by agricultural implement and coach makers, and areworth from 2S. 6d. to 35. per foot.REMARKABLE TREESI do not know of any ash at present alive in England which equals in size a treementioned by Loudon as growing near Moccas Court, Herefordshire, on the edge ofa dingle. This had immensely large roots, running on the surface for 50 feet ormore down the steep hillside, and a clear trunk of 30 feet long, 7 feet in diameterat 15 feet from the ground. This, including three large limbs, was estimatedto contain 1003 cubic feet of timber. This ash is remembered as a marvelloustree, though quite decayed, by the Rev. Sir George Cornewall, who told me thatnot a vestige of it now remains.The tallest living ash trees I have seen or heard of are in a grove near theFraxinusold heronry at Cobham Hall, Kent, the seat of the Earl of Darnley, a placewhich contains taller and finer ash trees and hornbeams, and more of them, than anythat I have seen in England. Strange to say, Strutt, who figured several treesat Cobham, overlooked these; but in Francis Thynne's continuation of Holinshed'sChronicles, p. 1512, I find the following, which shows that Cobham was renownedfor its trees more than three centuries ago. Speaking of William, the last LordCobham but one, he says :" Besides which, owerpassing his goodlie buildings at the Blackfriers in London,in the year of Christ 1582, and since that the statelie augmenting of his house atCobham Hall, with the rare garden there ; in which no varietie of strange flowers andtrees do want, which praize or price maie obtaine from the farthest part of Europe,or from other strange countries, whereby it is not inferior to the garden of Semiramis."The largest ash here, described by Loudon, was a tree 120 feet high, witha trunk 6 feet 8 inches in diameter, straight, and without a branch for a great height.This was perhaps the same whose trunk I saw in July 1905 lying on the ground,where it had fallen several years ago. But those which remain are not only thetallest ash trees, but the tallest trees of any sort with one exception that I havemeasured in England, and there are so many of them that I can well believe thatI did not measure the tallest. The tallest tree, measured in April 1907, by LordDarnley, is 146 feet high by 12 feet in girth. Another growing by the side of adrive, which he christened Queen Elizabeth's ash, I measured 143 feet high by12 feet 7 inches in girth. In the grove near it are several, very nearly if notquite, as tall, one of which I made 141 feet by 13 feet i inch, with a bole 50 feethigh, and a roughly estimated contents of 700 to 800 feet. (Plate 239.) Another,140 feet by 12 feet 9 inches, with a bole of 48 feet, which, judging from the largemass of fungus growing on its root, is probably decaying. There are many othertrees in this grove which are 125 to 130 feet high, and stand pretty close together,growing in a sheltered situation, 1 on what appeared to be a deep but rather sandyloam.Other remarkable ash trees at Cobham are the Twisted Ash, whose trunk isspiral, and measures 116 feet by 17 feet 9 inches. (Plate 240.) The View Ash,a tree nearer the house, is only about 80 feet high by 17 feet 9 inches at 5 feet,but is 29 feet in girth at the base, and has its trunk and most of its branchescovered with green and healthy twigs.Next to Cobham in respect of its great ash trees is Knole Park, also in Kent,where, in a sheltered valley near the gate from the Sevenoaks Road, called " TheHole in the Wall," are a number of very fine sound trees, from 125 to 130 feet highor more, and from 13^ to 16 feet in girth, one of which has a bole 35 feet long, andprobably contains over 700 feet of timber. Here again the soil is a deep sandyloam, which grows splendid beech, oak, and chestnut, but I cannot guess the age ofthe ash, though they are probably over 200 years.One of the most perfect examples, from a timber point of view, is a treegrowing in a wood called Poultridge, just outside Ashridge Park (Plate 241), whichIV1 This is about 330 feet'above sea-level, and is situated between the Medway and the Thames.

874 The Trees of Great Britain and Irelandis about 125, perhaps 130 feet high, but difficult to measure on account of thesurrounding trees. It had, in 1906, an absolutely straight, clean stem, about 75feet in height by 11 feet 10 inches in girth, and looked as if it would square 27 to 28inches halfway up, in which case it contains about 400 feet of faultless timber in onelength. It is surrounded by other trees and underwood, and a photograph could nothave been taken if Mr. Liberty, forester to Earl Brownlow, had not been obligingenough to clear away the intervening brushwood.At Chilham Castle, in Kent, the seat of C. S. Hardy, Esq., there are somesplendid ash trees. The best of these was recently cut down in the heronry, whereI saw its stump in 1907, and counted about 185 rings, the diameter being about4 feet. The soil here does not appear to be deep, and is on a chalk subsoil, as atAshridge, but the tree grew in a very sheltered position, was drawn up to a heightof 132 feet, and nourished by the beech which surrounded it. The first length contained 236 cubic feet. A tree of very similar character was growing in the park notfar off, and measured about 115 feet by 13 feet 6 inches, with a straight, clean boleabout 45 feet long. I estimated its contents at 280 feet in the first length.At Godinton, Kent, there are many very fine tall ash trees in what is called the"Tole," a splendid clump mainly composed of chestnut, which contains as large aquantity of clean, fine timber as I ever saw on the same area in England. Imeasured an ash here about no feet high, and only 6 feet 5 inches in girth, whichhardly fell off at all in thickness up to 50 or 60 feet.At Woburn there was in Strutt's time an immense ash which he figures onPlate 22, and gives the height as 90 feet, the girth at 3 feet, 15 feet 3 inches, thebole 28 feet high, and the diameter of the branches 113 feet. The estimatedcontents were 872 feet. This tree was still healthy in Loudon's time but I can findno trace of it at present.At Arley Castle there is a fine ash tree 120 feet by 14 feet as measured in 1904by Mr. R. Woodward ; and at Althorp there was a very fine ash which is now muchdamaged by wind and old age. When measured in 1890 by Mr. Mitchell, nowforester at Woburn Abbey, 1 it had a bole 36 feet high with immense limbs spreadingon all sides, was 17 feet 3 inches in girth at 3 feet, and had a cubic content of 800 feet.At Hatfield there is also a very fine ash tree growing near the big elm in a hollow,which I made 104 feet by 15 feet 8 inches. An immense ash butt was bought byMr. Miles of Stamford in February 1894, and hauled into his yard by eleven horses.I am informed by Mr. C. Richardson that it measured 20 feet long by 50 inchesquarter-girth, equal to 435 cubic feet of timber, and that it was the largest hehad ever seen or heard of. Mr. A. B. Jackson in 1908 saw some tall trees atKedleston Park, Derby, one measuring 120 feet by 10 feet 2 inches, and another125 feet by 8 feet 10 inches. He measured also a tree at Elvaston Hall, no feetby 12 feet 7 inches.At Studley Royal, Yorkshire, there are some tall, straight, and clean ash, the bestthat I measured being 119 feet by 10 feet 6 inches; and at Castle Howard andother places in the same county, and in parts of Lincolnshire, the ash is a more1 Trans. Roy. Scott. Arb. Soc. xiii. 88 (1891).Fraxinus 875profitable timber tree than any other. The largest that I have seen in the northis one at Lowther Castle, known as Adam's Ash, which is 21 feet 10 inches in girth.Lees in Gardeners Chronicle, November 7, 1874, figures several curiously splitand distorted relics of ash trees which existed in the district he knew so well roundWorcester; but does not mention any of extraordinary size. He suggests that suchwrecks may often have escaped being converted into firewood, owing to thesuperstition which formerly prevailed, that rickety children might be cured bypassing them through a fissured ash tree; and relates an instance known to him ofthe curious superstition, that a similar passage through a cleft ash would inducefertility in barren women.White, in his Natural History of Selborne (Ed. Alien, 1900, p. 266), speaks of arow of pollard ashes at Selborne which had been in former times cleft and held openby wedges while ruptured children were pushed through the aperture, in thebelief that they would ,be cured of their infirmity. He also states that there werethen several people living in his parish, who in childhood were supposed to be socured. He describes an old pollard ash which for ages had been looked on with nosmall veneration as a " shrew-ash," and whose twigs were, when applied to the limbsof cattle, supposed to cure the pains caused by a shrew-mouse having run over theaffected part. A " shrew-ash " was made by boring a hole in the trunk and puttinga live shrew into the hole, where it was plugged up with several quaint incantationsnow long forgotten.In Wales I have not seen or heard of any larger than a tree by the slaughterhouse in Dynevor Park. Though not very well shaped and somewhat past itsprime, I found it in 1908 to be 104 feet by 22 feet 9 inches.In Scotland there are also many great ash trees, of which perhaps the largestrecorded anywhere has long ago completely decayed. It was described inWalker's Essays, p. 17, as growing in a deep, rich soil, in the churchyard ofKilmalie, and was considered to be the largest and most remarkable tree in Scotland ;and said to measure in 1764 no less than 58 feet in girth at the ground.Another celebrated old tree is in the hotel garden at Logierait, Perthshire, whichHunter 1 described in 1883 as measuring 47 feet 7 inches at i foot from theground, and 32 feet 5 inches at 5 feet. It was then completely hollow andcovered with ivy, with an opening 5 feet 9 inches wide on one side, in which asummer-house had been made. But the late Sir R. Menzies informed me in 1903that this tree is now very much decayed.Strutt, Sylva Scotica, Plate 8, figures a very fine ash tree, at Carnock,Stirlingshire, then in perfect vigour, and said to have been planted about 1596 bySir T. Nicholson of Carnock. He gives its measurement as 90 feet high by 19 feet3 inches at 5 feet. This tree, however, died and was broken up about 40 yearsago. The tallest ash that I have myself measured in Scotland is at Gordon Castle,which is a fine healthy tree in the home park, 101 feet high with a bole of 30 feet,which in 1904 girthed 12 feet 6 inches. J. Webster records one at the same place1 Woods, Forests, and Estates of Perthshire, 545 (1883).

876 The Trees of Great Britain and Ireland70 feet by 15 feet 8 inches in 1881. Sir Archibald Buchan-Hepburn reports a treeat Smeaton-Hepburn 124 feet by n feet in 1908.But these are much exceeded in height by a tree west of the Beech Walkat Mountstuart in the island of Bute, which James Kay 1 states to have beenin 1879 no less than 134 feet high by 9 feet 4 inches, with a bole of 35 feet6 inches, and if this measurement was correct it must have been the tallesthardwood tree in Scotland. It was estimated to contain 273 feet of timber. Icould not, however, identify this tree when I visited Mountstuart in 1906, and fearthat, like some of the splendid beech trees which grew there, it has fallen.At Ochtertyre, Perthshire, Hunter 2 records an ash supposed to be about 400years old, which measured in 1881 34 feet 10 inches at i foot and 20 feet 8inches at 5 feet, and I am informed by the widow of the late gardener, Mr.Conacher, that the tree still remains in very good condition. At Keir, near theBridge of Allan, there is a remarkable ash stool, from which four stems proceed,averaging 6 feet 4 inches in girth and 103 feet in height. A tape 16 feet roinches long, girths the four stems at 5 feet from the ground. At Dupplin Castle Imeasured a fine tree about roo feet high with a stem clean to 45 feet and 10 feet 7inches in girth. At Dalswinton, Dumfriesshire, Henry measured in 1904, a tree nofeet by 8 feet 3 inches, with a fine clean stem ; and another 93 feet by 13 feet 3 inches.Near Cawdor Castle, Nairnshire, the property of the Earl of Cawdor, and oneof the most beautifully situated of the really old inhabited castles in Scotland, thereis a very large, though branchy and ill-shaped ash no less than 21 feet in girth.At Brodie Castle, Morayshire, there is a very large tree, 18 feet 8 inches in girth, ofwhich the owner has kindly sent me a photograph; and at Darnaway, in the samecounty, an immense tree of great age, much damaged by storms, existed in 1881.Even as far north as Conon House, Ross-shire, the seat of Sir Kenneth MacKenzieof Gairloch, the ash grows extremely well in a low-lying flat. Here I saw a lot ofbeautifully grown, though not very large trees, which would have been a credit toany woodland in the south.On the shore of Loch Fyne, a mile north of Minard Castle, a curious ash growson the beach at high-water mark, which is known as the " Nine Sisters," because ninestems sprang from the same root, the largest of which when I saw them in 1907 were7 to 9 feet in girth.In Ireland the ash thrives exceedingly well; and often attains an immense size.In Co. Meath, where the soil is remarkably fertile, it has in many parts expelled allthe other trees from the hedgerows ; and one may drive long distances on the roadsbetween lines of flourishing ash trees, without seeing a single oak or beech.In the latter part of the eighteenth century several ash trees of enormous sizewere still living, of which Hayes gives an account. 3 He relates, on the authority ofan official of the Dublin Society, that a tree was then standing at Donirey nearClare Castle in Co. Galway, which measured in girth 42 feet at 4 feet and 33 feet at6 feet. The trunk had long been hollow, having been used 25 years before as a1 Tram. Scot. Art. Sec. ix. 7 5 (1879). - Op. fit. 454.3 Practical Treatise on Planting, 1 37, 142, 148 (1794).Fraxinus 8 77school. In 1794, it had only a few branches remaining, which were, however, stillvigorous. At Curraghmore, the seat of the Marquess of Waterford in the county ofthat name, there were many enormous ash trees, one of which was 22^ feet in girthat a height of 13 feet 9 inches, the girth of the butt being 33 feet 9 inches. In 1792Hayes measured the famous ash tree at Leix in Queen's County, which was 40^ feetround at i foot, and 25 feet at 6 feet up, where the girth was least. Marsham in aletter to Gilbert White, dated i2th February 1792, says that a print of this tree wasthen being engraved in London, but we have never been able to see a copy of this,if it was ever published.An ash at Castledurrow, which Hayes says was the finest he ever saw, inOctober 1793 measured 18 feet in girth and nearly as large at 14 feet high, thebranches extending 45 feet from the stem in every direction. Another at KennityChurch, King's County, was 21 feet 10 inches in girth with a bole of 17 feet. Thefuneral parties used to stop and say prayers under this tree, after which they threwa stone at its foot.None of these relics of antiquity now remain; and most of the fine ashes havebeen cut down. Probably the tallest now left in Ireland is one at Woodstock inCo. Kilkenny, in an alluvial flat beside the river Nore and close to the village ofInistioge (Plate 242). Henry, in 1904, made it 127 feet in height by 16 feet 9inches in girth; but the forester's records give it as 136^ feet high in 1901. Thistree was u feet in girth in 1825, n feet 8 inches in 1834, 13 feet 2 inches in 1846,and r 6 feet 9 inches in 1901, according to the same records. Another tree besideit, which was 9 feet 5 inches in girth in 1825, was ir feet 9 inches in 1901 ; anda third tree, 10 feet 5 inches in 1825, had attained 15 feet 3 inches in 1901.At Mitchellstown, in Co. Cork, there was in 1903 a remarkably fine ash, witha tall clean stem, which was i n feet high by 9 feet in girth. Another tree, 91 feethigh, was 27 feet in girth at 2 feet above the ground, but only 15 feet 2 inches at 6feet up. At Kilmacurragh in Co. Wicklow there is a good ash, the height of whichexceeds roo feet, but is impossible to measure on account of the situation of thetree ; the girth in 1903 was 13 feet at 5 feet above the ground. There are fine treesof great girth (17 to 18 feet), but not remarkable for height, at Doneraile Court,Co. Cork. Henry measured in 1904, on Lord Oranmore's property at CastleMacgarrett, Co. Mayo, a tree 119 feet by 14 feet 2 inches, containing 216 cubic feetof timber. At Castlewellan, Co. Down, there are some very fine ash in the park,one of which near the garden gate is figured on Plate 243. It measured in 1908about 70 feet high, and 18 feet in girth.The finest ash that I have seen on the continent is in the Royal park of Jaegersborgnear Copenhagen, and measured in 1908, 125 feet by 11 feet 8 inches.ASH DISEASESThe commonest disease in the ash in my own district is a canker which affectstrees during most of their life without killing them, though the timber is worthlessexcept for firewood. This disease is described and figured by the late Mr. Wilson

878 The Trees of Great Britain and IrelandSaunders, F.R.S. ;* and Mr. Sidney Webb fully explained the manner in which extensive canker resulted from minute wounds, at a meeting of the Scientific Committee ofthe Royal Horticultural Society, February n, 1879. A full account of the diseaseis given by Dr. Masters in the Gardeners' Chronicle, 1 879, p. 208, where it isstated that the injury is originated by the larva of a minute moth called Prayscurtisellus? Plate 244 shows a bad case of this canker in an ash at Staple, nearColesborne; and there is a tree at least 70 feet high by the roadside, close tothe sixth milestone from Cirencester to Cheltenham, which has had this diseasefrom its base to near the summit as long as I can remember.In Trans. Scot. Arb. x. 235, there is a useful paper on the Ash Bark Beetle,Hylesinus Fraxini, a pest which seems to be dangerous only where the ash isalready unhealthy. As the eggs of this insect are laid in spring only under the barkof felled, dead, or sickly trees, wherever this pest is troublesome, all such should beremoved from the neighbourhood of the healthy trees by April, and ash loppingsshould not be left on the ground. A curious malformation occurs in a tree growingclose to Cirencester, on the east side of the Tetbury road nearly opposite the Kennels.Plate 244 shows the remarkable growths on its branches, specimens of which weresent to Kew and found to contain numerous examples of Hylesinus Fraxini.TIMBERFor coach, waggon, and agricultural implement making, and for all purposes inwhich strength, toughness, and durability are required, ash timber has no equal, andno substitute has been found among foreign trees which can be relied on as well.In consequence, it is now the easiest to sell, if not quite the highest priced, of allEnglish timbers ; and its growing scarcity seems to point to a great future for it.It varies much, however, in strength, toughness, and elasticity, according to thesoil on which it is grown, and the age at which it is cut. I am informed by Mr.Clutterbuck of the Gloucester Waggon Works, who has had long experience withEnglish and foreign timber, that there is no better ash in England than that grownon the Cotswold hills; but if left standing too long, it becomes discoloured at theheart, and is probably never worth more per foot than when 60 or 70 years old.It is now perhaps the only wood worth growing as copsewood, and, when established on good land and cut every ten to fifteen years, still makes as much as apound per acre per annum. The poles are used for making hurdles, hoops, crates,and many other purposes, and as hop-poles are only second to chestnut. I havefound that for the rails of light field-gates in a hunting country, on account of theirelasticity nothing is better; and when well made they last thirty years or more. Ashwood takes creosote well, which very much increases its durability; and some sheephurdleswhich I had creosoted thirteen years ago are still sound, though when notso treated, they do not last more than three to five years. Ash, however, soondecays in contact with the soil and is unfit for building purposes, though it was1 Journ. Roy. Hort. Soc. v. 135 (1879).8 Also known as Tinea cartisella, Don. Cf. Schlich's Man. Forestry, iv. 344 (1907), where the moth is figured anddescribed.Fraxinusformerly used for staircases. Loudon says that the staircase at Wroxton Abbey,near Banbury, was made of this wood, and in Hatfield House some of the insidework is made of ash. Slabs from flat-sided trees often show a very ornamentalcurly grain which makes very handsome panelling, and might be used for doorpanels with good effect.The burrs are also cut into beautiful veneers which, when polished, are used incabinetmaking, and which sometimes in Hungary and south Russia are of greatsize and perfectly sound, though in England usually small and faulty. I purchasedin Manchester, under the name of Circassian ash burr, some splendid veneers of thiswood, which measure about 5 feet by 3 feet, and are made up of small, closely crowdedknots, which take a fine polish and are of a greyish white or pinkish grey colour. 1English ash, however, seldom or never assumes the wavy grain which is found inHungary and Russia, and is one of the most beautiful woods I know. Thisis known as fiddle-back ash, because wood of this character, usually maple,is selected for the backs of violins. It varies very much in colour, the most valuablebeing the whitest; and also in the size and character of the figure; but when acombination of small waves with eye-like patches is combined, it is superior to thebest American maple. Such wood was formerly much used for decorating railwaycarriages, and for furniture, but from some reason which I cannot explain is now outof fashion. I believe that the waving rarely extends throughout the tree, the bestfigure being always near the outside, and the causes which produce it are, so far as Iknow, as yet without any scientific explanation. (H. J. E.)FRAXINUS ANGUSTIFOLIA, NARROW-LEAVED ASHFraxinus angustifolia, Vahl, Enum. i. 52 (1804); Loudon, Arb. et Frut. Brit. ii. 1229 (1838).Fraxinus australis, Gay, ex Koch, Dendrologie, ii. I. 247 (1872).A tree attaining 70 to 90 feet in height. Shoots glabrous, green, slender.Leaflets (Plate 262, Fig. 6), seven to thirteen, ij to 3 inches long, smooth andslightly coriaceous, shining above, usually pretty uniform in size, subsessile,lanceolate, base cuneate, apex acuminate, glabrous on both surfaces; coarsely andsharply serrate except near the base; serrations few, spreading, often with incurvedpoints (occasionally deeply serrate with long bristle points). Rachis of leaf glabrous,strongly winged, the wings meeting above and only showing a groove opposite theinsertions of the leaflets. Flowers (section Fraxinaster) without calyx or corolla, fewin erect racemes, arising from the axils of the leaf-scars of the preceding year's shoot.Fruit lanceolate, obliquely truncate and entire at the apex; but apparently variable.This species is distinguished from all forms of the common ash by its absolutelyglabrous leaflets, which have fewer, sharper, and more spreading serrations than inthat species. The terminal buds are also different, being small, dark brown, quadrate,1 I have seen in London a fine old cabinet, supposed to be veneered with Amboyna wood, so like the Circassian ash inpattern, though the colour was yellower, that I much doubt whether the two could be distinguished when made up.

88oThe Trees of Great Britain and Irelandconical, usually glabrous, with four outer scales six external scales, however,occurring in individual trees with leaves in whorls of threes.VARIETIES1. Var. monophylla^ (F. Veltheimi, Dieck). A form in which the leaves(Plate 262, Fig. 3) are simple, unequally two-foliolate, or three-foliolate; terminalleaflet or single leaf lanceolate, acuminate, coarsely serrate or dentate; lateralleaflets, when present, much smaller but similar in shape; petiole with a wide opengroove on the upper side. Shoots green, glabrous, with pink lenticels.This variety can only be confused with Fraxinus excelsior, var. monophylla,from which it is readily distinguishable by the glabrous leaflets, different in textureand usually narrower, being lanceolate and not ovate. This tree appears to do wellin cultivation, but will probably not attain a large size. We know of no trees of thisvariety in England except in the collection at Kew.2. Var. lentiscifolia (F. lentiscifolia, Desfontaines 2 ?). In the typical form ofF. angustifolia, both as it occurs wild and under cultivation, the leaflets are set closeupon a short rachis and point forwards towards its apex. In this variety, theleaflets are set wide apart upon an elongated rachis, from which they spread out atright angles and are not directed forwards; they also differ slightly in colour andtexture from the type.Willdenow 3 considered F. parvifolia, Lamarck, 4 F. tamanscifolia, Vahl, 6 andF. lentiscifolia, Desfontaines, 6 to be identical. I have not been able to followKoehne or Dippel in their treatment of these forms as distinct. It is not in theleast certain what species was intended by Lamarck's parvifolia, a name which hasbeen given by some writers to a small-leaved variety of F. oxycarpa. I n theKew Herbarium, a garden specimen, collected by Bentham in 1854 and labelled"F. lentiscifolia," agrees with our variety of F. angustifolia; and a wild specimenfrom Italy is indistinguishable from it.Both pendulous and dwarf forms of var. lentiscifolia are known in cultivation.DISTRIBUTIONThe narrow-leaved ash, with glabrous leaves, is common in the south of France,and occurs also in Spain, Portugal, Morocco, 7 and Algeria.Captain Widdrington 8 says of this tree : " The ash is extraordinarily rare in thesouth and central regions of Spain. It is not now cultivated, and the only specimensI saw growing wild were in the wilder parts of Estremadura and the Sierra Morenawhere they were generally by the side of water-courses. The only species that Ihave seen in these regions is the lentiscifolia. The first year that I was in the1 Stated in the Kew Hand List of Trees, 545 (1902), to be a cross between F. parvifolia and F. excelsior moncphylla.It shows no evidence of hybrid origin, and is evidently a variety of F. angustifolia.* Table de I'Zcole de Bot. 52 (1804). 3 Sp_ PL i v 2 > , IO, ( lgos) _ 4 Encyci_ ;i. 5 46 ( I7g6).7 £"!""- L 52 (1804). e TaUe de fEcoU de Bot. 5 2 (1804).This ash, referred to F. oxyphylla by Ball, was seen by him in Morocco, between Tangiers and Tetuan, and also at thebase of the Atlas mountains. Cf. Journ. Linn. Soc. (Bet.) xvi. 564 (1878).8 Spain and the Spaniards, i. 390 (1844).Fraxinusinterior of Spain, I picked up the seed of an ash near the Escorial; but the leaveshaving fallen, I did not ascertain the species, but sending them to England theyvegetated, and are now growing in Northumberland. This is the same tree, andI have never seen it farther north than New Castile; at the same time I think itprobably may exist as far as Leon, where, the instant you cross the chain, theFraxinus excelsior, our common species, supplies its place; at least, I could make outno difference. The timber of F. lentiscifolia is heavy and less elastic than that ofour species, but the elegance of the tree, and its perfect hardiness in a dry soil, shouldmake it more common than it is in our ornamental collections."This species * replaces the common ash in Algeria, where it is only found wildin quantity in the forests of the plains and along the banks of streams and rivers ;but it ascends as isolated trees occasionally to 6000 feet in the mountains, and isreported to be common in the Djurdjura range. I saw it growing in fields near theforest of Akfadou, inland from Bougie, where the trees have a mutilated appearance,owing to the annual lopping of their branches by the natives, who feed their cattlewith the leaves. 2 In Algeria 3 the tree attains a large size, and grows in good soilwith great rapidity, reaching a height of 90 feet by 3 feet in diameter at seventyyears old. The wood is similar to that of the common ash, though slightly inferiorin quality. Dr. Trabut informed me that he had sent seed to Australia, where thetree is said to thrive well, succeeding better than the common ash.A tree at Chiswick House measured, in 1903, 75 feet by 7 feet 5 inches. Anotherat Whitton, near Hounslow, in 1905 was 56 feet by 6 feet 2 inches. At WilliamstripPark, Gloucestershire, in 1904 Elwes measured a tree 60 feet by 6 feet 9 inches.The variety lentiscifolia has been identified by us at Syon, 55 feet by 5 feet6 inches in 1905 ; at Hardwick, Bury St. Edmunds, a grafted tree, 72 feet by 6 feet3 inches below the graft and 7 feet 10 inches above it; at Bicton, another graftedtree, over 50 feet by 6 feet; and at Stowe, also a grafted tree, 68 feet by 7 feet8 inches. From the similarity in appearance of the trees at Hardwick, Stowe,and Bicton, there is little doubt that they were all propagated and planted at thesame time. Elwes has also seen at the Hendre, Monmouthshire, a tree of similarappearance, which was 71 feet high by 5 feet 4 inches below, and 6^ feet abovethe graft. There is also a healthy grafted tree at Ware Park, Herts, growing onsandy soil, which Mr. H. Clinton Baker measured as 78 feet by 6 feet 3 inches in1908. Another (Plate 245) on the lawn at Rougham Hall, the seat of F. K. North,Esq., is 76 feet high by 8^ feet above the graft, and 7^ feet below it. A third atRibston Park, Yorkshire, was 68 feet by 6 feet 7 inches.It is hardy as far north as Denmark, where Elwes measured in the park ofCount Friis, in 1908, at Boiler near Horsens, a grafted tree about 60 feet by 4feet 4 inches, which was bearing immature fruit.(A. H.)1 The Algerian tree has been distinguished as var. numidica (F. nnmidica, Dippel), with broader and larger leaflets; butspecimens gathered by me at Akfadou are typical angustifolia.3 M. Maurice L. de Vilmorin, in Bull. Soc. Amis des Arbres, 1 895, states that this ash is much planted around villagesin Kabylia, where its leaves, which are stripped off the tree in September, are an indispensable fodder for cattle, sheep, andgoats at this season when no grass is available. The foliage of a single tree is usually worth 50 francs; and he was shown avery old wide-spreading tree, the owner of which sold its leaves annually for 300 francs.3 Cf. Lefebvre, Forets de FAlglric, 348 (1900).IV

2 The Trees of Great Britain and IrelandFraxinus 883FRAXINUS OXYCARPAFraxinus oxycarpa, Willdenow, Sp. PI. iv. 2, noo (1805); Loudon, Arb. et Frut. Brit. ii. 1230(1838).Fraxinus oxyphylla, M. Bieberstein, Fl. Taur. Cauc. ii. 450 (1808); Boissier, Flora Orientalis, iv.40 (1879).Fraxinus rostrata, Gussone, PI. Rar. 3 74, t. 63 (1826).A tree of moderate size. Shoots green, glabrous; lenticels pink. Leaflets(Plate 263, Fig. n), nine to thirteen, usually small, i^ to 3 inches long, sessile orsubsessile ; lanceolate, oval or ovate ; base tapering, apex acuminate ; serrations few,sharp, spreading and often ending in incurved points; lower surface pubescent onthe midrib and veins towards the base. Rachis of the leaf glabrous, winged, thewings meeting above and only forming an open channel opposite the nodes. Flowers(section Fraxinaster) without calyx or corolla, in short racemes in the axils of theleaf-scars of the previous year's shoot. Fruit broad, oblanceolate, acute or acuminateat the apex.This species closely resembles F. angustifolia, differing in the leaflets beingalways pubescent beneath. The terminal buds are conical, quadrate, long andslender, with four outer narrow scales of equal length, dark brown and pubescent.It is doubtful if the differences in the samarae relied upon for the separation ofthis species from F. angustifolia are really constant. The fruits in this group ofFraxinus are extremely variable, and do not appear to give specific characters.While F. angustifolia seems to be confined to the western part of theMediterranean region, Fraxinus oxycarpa is widely distributed in Italy, Asia Minor,Persia, and the Caucasus.Var. parvifolia, Wenzig. 1 Leaflets small, oval-oblong. Boissier considers thisto be rather a bushy or sterile juvenile form of the species than a distinct variety,and records it from various localities in Asia Minor. An ash identified by Mathieu, 2with F. parvifolia, Lamarck, grows in the neighbourhood of Montpellier in France,as a shrub about 5 to 10 feet high, and belongs to this variety.At Kew, the ash, cultivated as F. parvifolia, Lamarck, is a variety ofF. oxycarpa, distinguished by having leaflets (Plate 263, Fig. 9) shorter than inthe type, broader in proportion to their length, and more closely set upon therachis.Fraximis oxycarpa is much rarer in cultivation in England than F. angustifolia;but small trees are growing at Kew, Woburn, Eastnor Castle, Oxford BotanicGarden, and at Grayswood, Haslemere, where it has produced seed, from whichplants have been raised by Elwes, and are growing at Colesborne. It does not seemin this country to be so vigorous a tree as F. angustifolia. (A. H.)1 In Engler, Bot. Jahrb. iv. 175 (1883). 2 Flore Forestitre, 245 (1897).FRAXINUS SYRIACAFraxinus syriaca^ Boissier, Diag. Ser. I. ii. p. 77 (1849).Fraxinus oxyphylla, M. Bieberstein, var. oligophylla, Boissier, Fl. Orient, iv. 40 (1879).A tree attaining 60 feet in height. Shoots glabrous, green, stout, conspicuouslymarked by very prominent leaf-bases; lenticels white. Leaves (Plate 263, Fig. 10),small, always in whorls of threes or fours. Leaflets usually three, occasionally fiveto seven on some of the branchlets, sessile, lanceolate to ovate, base cuneate, apexacuminate, sharply and coarsely serrate, the serrations with incurved points, glabrouson both surfaces. Rachis of the leaf narrowly winged, the wings not meetingabove, but forming an open groove.Flowers (section Fraxinaster) in short racemes in the axils of the leaf-scars ofthe preceding year's shoot; without calyx or corolla. Fruit ovate-oblong; apexrounded, truncate or acuminate, ending in a mucro.This species occurs in Syria, Kurdistan, Persia, Baluchistan, and Afghanistan.The occurrence always of the leaves in whorls, a phenomenon met with inindividual instances in other species, appears to be constant in this species. Onstrong shoots leaves with five to seven leaflets exceptionally appear ; those withthree leaflets, however, being by far the most common. Small specimens of thistree are growing at Kew, but it does not seem likely to be worth growing inEngland. (A. H.)FRAXINUS ELONZAFraxinus Elonza, Dippel, Laublwlzkunde, i. 87, fig. 46 (1889); Koehne, Deutsche Dendrologie, 5 13(1893)-A small tree. Branchlets green, glabrous; lenticels few, oval, white. Budslaterally compressed and not quadrangular, narrowed and rounded at the apex;external scales four, densely brown pubescent, inner pair longer than the outer pair.Leaflets (Plate 266, Fig. 27), eleven to thirteen, i to 2^ inches long, sessile, ovalor lanceolate, with unequal base and acuminate apex, sharply and irregularly serrate,some of the serrations being often triangular and spreading ; under surface pubescentnear the base with brown tomentum, often occurring only on the inner side of themidrib. Leaf-rachis, with scattered pubescence, densest at the nodes; stronglywinged, the wings meeting above in part of their length. Fruit described as broadlylinear, with almost parallel sides, truncate and emarginate at the apex.The native country of this species is unknown ; and it is possibly a hybrid, as1 Fraxinus sogdiana, Butige, Mini. Sav. Etrang. Acad. PMerslxnirg, vii. 390 (1851), occurring in Turkestan, formerlysupposed to be identical with this species, is considered distinct by Koehne in Gartenflora, 1 899, p. 288, and by Lingelsheim,in Engler, Bot. Jahrb. xl. 222 (1907).

The Trees of Great Britain and IrelandKoehne suggests. It occurs in England in cultivation under the erroneous name ofF. chinensis? the plants at Kew having been obtained from Sir C. W. Strickland,who tells us that it was sent out by the Royal Horticultural Society some years ago.Plants cultivated as F. Elonza are usually F. oxycarpa. (A. H.)FRAXINUS WILLDENOWIANAFraxinus Willdenowiana, Koehne, Deutsche Dendrologie, 5 15 (1893).Fraxinus parvifolia, Willdenow, Berlinische Baumzucht, 1 24, t. 6, f. 2 (1796) (non Lamarck); andSp. PI. iv. noi (1805).A small tree. Shoots glabrous, lenticels white. Leaflets (Plate 265, Fig. 24),seven to eleven, 2 to 3^ inches long, subsessile (except the terminal one, which ismuch the largest and stalked), ovate, base broad and unequally cuneate, apexacuminate; serrations coarse and sharp with minute incurved points; both surfacesglabrous. The leaflets increase in size from the base to the apex of the leaf, therachis of which is winged, the wings usually not meeting on the upper side, butforming an open groove. Fruit unknown.This species was considered by Willdenow to be different from F. parvifolia, withwhich he had first identified it, yet he left it with this name. Koehne has accordinglygiven it a new name. It is sometimes met with in cultivation under the name ofF. rotundifolia. It is readily distinguished by the large terminal leaflet and theopen-grooved rachis from the other glabrous species. Its native home is uncertain.It is perfectly hardy at Kew and has very distinct foliage. (A. H.)FRAXINUS DIMORPHAFraxinus dimorpha, Cosson et Durieu, Bull. Soc. Bot. France, ii. 367 (1855); Mathieu, FloreForestiere, 245 (1897).Fraxinus xanthoxyloides, Wallich, var. dimorpha, Wenzig, in Engler, Bot. Jahrb. iv. 188 (1883).A small tree attaining 40 feet in height. Young shoots purple, slender, glabrous,obscurely quadrangular. Leaves (Plate 262, Fig. i) on barren branchlets withseven to nine small leaflets, which are J to f inch long, sessile or subsessile, ovate oroval, crenulate-serrate, and glabrous except for pubescence on the midrib towardsthe base on the under surface; leaf-rachis usually glabrous, strongly winged, thewings spreading and forming a very open channel. The leaves on flowering shootsare larger, with seven to eleven leaflets, oblong-lanceolate, acute and serrate.Flowers (section Sciadanthus) perfect, without a corolla, but with a calyx whichpersists under the fruit, grouped in fascicled cymes on the previous year's shoot1 F. chinensis, Roxburgh, a native of China, is entirely distinct from F. Elonza, which is closely allied to F. oxycarpa, aMediterranean species. Koch, in Dendrologie, ii. pt. I, 247 (1872), mentions F. Elonza as having been in cultivation someyears, and considered it to be probably a variety of F. angustifolia.Fraxinusin the axils of the leaf-scars. Fruit oblong; body compressed with a longitudinalfurrow on each surface, and many-rayed; wing long and obliquely truncate at theapex. Buds very small, with two outer scales pinnately lobed.This remarkable ash was observed by Sir Joseph Hooker in South Morocco. 1It also occurs in the mountainous regions of Algeria in the valleys at 4000 to 6000feet altitude. In dry situations it remains a bush with very rigid and almost spinybranches, and rarely flowers even when very old. On the banks of streams it growsto be a small tree and produces flowers and fruit. The wood is very hard andheavy, with a satin-like lustre.This species is rarely seen 2 except in botanical gardens, where, as at Kew, itgrows to be a small tree, remarkable for its diminutive foliage. It has smooth greybark. (A. H.)FRAXINUS XANTHOXYLOIDESFraxinus xanthoxyloides, Wallich, List, 2833 ; C. B. Clarke, in Hooker, Fl. Brit. India, iii. 606(1882); Brandis, Indian Trees, 444 (1906).This species is probably only a pubescent geographical form of Fraxinusdimorpha, which it resembles exactly in habit. The young shoots are covered witha minute dense pubescence. The leaflets only differ from those of F. dimorpha inhaving a scattered pubescence all over the lower surface; the rachis of the leaf isalso pubescent.This species occurs in Baluchistan, Afghanistan, and the north-west Himalayanregion, at altitudes of 3000 to 9000 feet, growing mainly in dry valleys, where it isoften gregarious. It is reported to attain a height of 25 feet.It is rare in cultivation, and forms a small tree, scarcely distinguishable inappearance from F. dimorpha. (A. H.)FRAXINUS POTAMOPHILAFraxinus potamophila, Herder, in Bull. Soc. Imp. Mosc., xli. 65 (1868); Dippel, Laitbholzkunde, i .98, fig. 54 (1889).A small tree; branchlets glabrous. Leaflets (Plate 262, Fig. 8) small, sevento nine, about if to 3 inches long, stalked (petiolules glabrous, £ inch or more inlength), ovate, tapering unequally at the base, acute at the apex, coarsely serrate, theserrations often ending in long points; glabrous on both surfaces. Rachis of theleaf with angled edges on its upper side, enclosing a shallow groove. Flowers1 Hooker and Ball, Tour in Morocco, 1 76 (1878); Ball, njourn. Linn. Soc. (Bot.), xvi. 564 (1878).2 A tree at Coorabe Wood, which had attained almost 25 feet in height, was destroyed in 1907. Mr. A. B. Jacksonhas seen small specimens at Barren's nursery, Elvaston, where they were erroneously named E. lentiscifolia.

886 The Trees of Great Britain and Ireland Fraxinus 887unknown. Fruit described as linear-oblong, nearly two inches long, acute, blunt orobliquely truncate.This ash grows along the banks of rivers in Turkestan and Songaria, occurringin the Hi region at 1000 to 2500 feet elevation. It was introduced into cultivationby the Botanic Garden at St. Petersburg; and small trees are doing fairly well atKew.Fraxinus Rcgelii, Dippel, 1 of which I have seen no authenticated specimen, issaid to be also a native of Turkestan, and was considered by Koehne 2 to be probablyidentical with F. potamophila, Herder. There are young plants in the Kew collection, raised from seed sent in 1900 by M. Scharrer, Director of the Botanic Gardenat Tiflis, and named F. Regclii on his authority, which are remarkably distinct fromany ash known to me, and differ from Dippel's description of F. Regelii in the largersize of the leaflets, which are crenate and not dentate in serration. The Tiflis plantshave the young branchlets glabrous, purplish; leaflets (Plate 265, Fig. 25), five orseven, about 3 inches long, stalked, the base of the leaflet often decurrent on oneside of the petiolule to its insertion; terminal leaflet obovate or rhomboid; lateralleaflets ovate or oval; all shortly acuminate or cuspidate at the apex, unequal atthe base, crenately serrate; bluish green and glabrous on the upper surface; palegreen and slightly pubescent on the sides of the base of the midrib on the lowersurface ; rachis elongated, terete, glabrous, with a shallow groove on its upper side.The identification of these plants with F. Rcgelii must be left uncertain.(A. H.)FRAXINUS RAIBOCARPAFraxinus raibocarpa, Regel, in Act. Hort. Petrop. viii. 685 (1884).A small tree. Branchlets brown, minutely pubescent, glandular. Leaflets(Plate 266, Fig. 29), five, upper subsessile, lower stalked, about ij inch long, oval,unequal and rounded at the base, acute or obtuse at the apex, usually entire inmargin without cilia; under surface glabrous, with a few minute brown glands.Leaf-rachis slightly glandular, with a wide open groove on its upper side. Fruitin leafy panicles, arising on the current year's shoot; samara surrounded at the baseby the persistent calyx, curved, falcate ; body terete and rayed; wing terminal,very broad, spathulate-obovate, obtuse.This species, of which the flowers are unknown, belongs apparently to the sectionOrnus. I t was discovered in 1882 by Regel at elevations of 6000 to 7000 feet inthe mountain valleys of eastern Bokhara and Turkestan; and was introduced intocultivation shortly afterwards by the St. Petersburg Botanic Garden. Small plantsat Kew have grown very slowly, and this species does not seem likely to be worthcultivating in this country. (A. H.)1 Laubliohkimde, i. 97, fig. 53 (1889), described from plants sent out by the St. Petersburg Botanic Garden as F. sogdiana,an entirely different species, referred to on p. 883, note I.2 Deutsdie Dendrologie, 5 15 (1893).FRAXINUS HOLOTRICHAFraxinus holotricha, Koehne, in Mitt. Dent. Dendrol. Ges. 1 906, p. 67.A small tree. Branchlets grey, densely covered with short stiff erect pubescence. Leaflets, nine to thirteen, subsessile, lanceolate or ovate-lanceolate, about2 inches long, \ to f inch wide; apex prolonged into a sharp-pointed and oftencurved acumen; base unequal and tapering; margin ciliate, unequally and sharplyserrate ; both surfaces covered with a scattered grey pubescence. Rachis of the leaf,with pubescence like that of the branchlets, densest at the nodes; narrowly groovedon the upper side.Flowers (section Fraxinaster), without calyx or corolla, in short racemes,perfect in the single specimen seen ; ovary pubescent. Fruit unknown.This species resembles F. angustifolia in the shape and size of the leaflets ; butdiffers in the copious pubescence on the branchlets and leaves. The buds arequadrate, with four dark brown scales, very pubescent at the tips.F. holotricha was discovered by Koehne in Spain's nursery near Berlin, and inthe botanic gardens at Berlin and Dresden, where it had received the erroneousname of F. potamophila. Its native country is unknown. The specimen, from whichI have drawn up the above description, was sent me from Metz by Messrs. Simon-Louis. Three young plants were sent by Spath to Kew in 1908. (A. H.)FRAXINUS ORNUS, FLOWERING ASH, MANNA AsnFraxinus Ornus, Linnaeus, Sp. PI. 1 057 (1753); Bentley and Trimen, Medicinal Plants, iii. 170,fig. 170 (1880); Hanbury, Science Papers, 362-368 (1876).Fraxinuspaniculata, Miller, Diet. No. 4 (1759).Fraxinusflorifem, Scopoli, Fl. Cam. i i. 282 (1772).Ornus europaa, Fersoon, Syn. PI. ii. 605 (1807); Loudon, Arb. et Frut. Brit. ii. 1241 (1838).A tree attaining about 60 feet in height and 6 feet in girth, with smooth ashygreybark. Young branchlets slender with white lenticels, usually glabrous, occasionally glandular pubescent, marked with a ring of brown hairs at the base of the shoot.Leaflets (Plate 265, Fig. 26), five to nine, 2 to 3 inches long, the terminal oneobovate and stalked, the lateral ones with distinct stalklets, which are about £ inchlong and brownish pubescent; ovate to oblong, base rounded or broadly cuneate,apex shortly acuminate, finely and irregularly serrate, under surface glabrous exceptfor brown woolly pubescence on the midrib. Rachis of the leaf grooved, the groovedeepest in the upper part of its length, usually glandular pubescent, with tufts ofbrown hairs opposite the insertions of the leaflets.Flowering branches developed from terminal buds, which contain leaves as wellas flowers, the inflorescences being usually accompanied by two pairs of leaves andconsisting of five panicles, the largest one terminating the branch, four smaller onesarising in the axils of the leaves, the whole forming a drooping compound panicle.Pedicels long and slender. Calyx divided into four triangular acute persistent sepals.

888 The Trees of Great Britain and IrelandCorolla deeply divided into four, strap-shaped wavy wide-spread ing petals. Stamenstwo, hypogynous, the filaments twice as long as the petals. Flowers apparentlyperfect, but functionally behaving as if distinctly staminate and pistillate. Fruitcompressed, with a terminal flat obovate-linear wing, blunt or emarginate at theapex.IDENTIFICATIONIn summer the tree is readily distinguished by its smooth bark and stalkedleaflets showing the characters just enumerated. In winter the twigs show a slightpubescence towards the apex and a ring of hairs at the base of the shoot. Leafscarsparallel to the twig on projecting leaf-cushions, semi-orbicular to crescentic,the ends of the horns truncate, marked on the surface with a curved row of separatebundle scars. Terminal bud large, greyish to greyish brown, ovoid, four-sided,rounded (rarely acute) at the apex, the two outer scales gaping above and denselypubescent. Lateral buds smaller, densely pubescent, arising from the twigs at awide angle.VARIETIESFraxinus Ornus, occurring over a wide area, both as a wild tree in forests, andcultivated in sunny arid regions, as in Sicily, shows considerable variation in thesize, shape, and texture of the leaflets; and several varieties have been established. 1The only one of those which is truly distinct is var. argentea, Grenier et Godron, 2a remarkable form, growing wild in the forests of Corsica and Sardinia, distinguishedby the leaflets being silvery white beneath, firm in texture, crenulate-serrate, usuallysmaller than the type, often subsessile, though stalked leaflets also occur on thesame branch, ovate or oval in outline, occasionally approximating to an orbicularshape. This singular variety has been made a distinct species; 8 but modern Frenchand Italian botanists regard it as only a peculiar geographical variety, which seems tobe sporadic in forests where the type is also met with.Aiton 4 and Willdenow 5 described, as the true manna ash, F. rotundifolia, withbroad ovate or almost orbicular, deeply serrate leaflets. Willdenow's figure of thefoliage corresponds with the ash which I have described (p. 866) as F. excelsior, var.rotundifolia, and it is possible that both he and Aiton were in error in consideringtheir plant to have flowers like those of F. Ornus.Lamarck's 6 Fraxinus rottindifolia, differs, according to the description, fromAiton and Willdenow's species of the same name, and is considered by Wenzig 7 andLingelsheim e to be F. Ornus, var. rotiindifolia ; but is kept up by Koehne 9 as adistinct species. Hanbury 10 states that the manna ash cultivated in Sicily shows1 Fieri et Paoletti, Flora AnaKtica d'ltalia, ii. 341, mention, besides the typical form, var. rotundifolia with broadelliptical leaflets, and var. lanceolata with lanceolate leaflets.2 Flore de France, ii. 473 (1850). 3 Fraxinus argentea, Loiseleur, Flora Gallica, ii. 697 (1806).4 Hart. Kew, iii. 445 (1789).6 Berlin. Baumzucht, 1 16, fig. vi. I (1796). The figure was copied by Loudon, Art. et Frut. Brit. ii. 1244(1838),who merely repeats Willdenow's description, and was probably unacquainted with the tree.6 Emycl. ii. 546 (1786). * In Engler, Bot. JcJirb. iv. 169 (1883). 8 Ibid. xl. 212 (1907).fl Dartschc Dendrologie, 508 (1893). l Science Papers, 368 (1876).Fraxinusgreat variation, but that no special form can be singled out as deserving the namerotundifolia ; and I have seen no specimens of F, Ornus, which could be separated asa var. rotundifolia, much less any which could be separated as a distinct species. 1A variegated form of F. Ornus, of which I have seen no specimens, is said tooccur in cultivation. A simple-leaved form, var. diversifolia, Roch., is of rareoccurrence in the wild state, and has been noticed in the canton of Tessin inSwitzerland. 2DISTRIBUTIONThe manna ash is widely distributed in southern Europe and Asia Minor. InFrance it only occurs wild in the department of the Maritime Alps ; and in Switzerland it is met with in a few places in mountain woods about Lake Lugano in thecanton of Tessin. 2 It grows in the southern Tyrol, where it ascends to 2000 feetelevation, in Carniola, Istria, Dalmatia, Croatia, Slavonia, and Banat, reachingits northern limit in Hungary on the south side of the Carpathian chain. It iscommon in eastern Spain, Corsica, Sardinia, Italy, Servia, Bosnia, Greece, andAsia Minor.The wood of the flowering ash is excellent, and the foliage is used as fodder inthe southern countries of Europe; but its chief economic importance is due to itsbeing the source of manna. The manna of commerce, according to Hanbury,is exclusively collected in Sicily, where the plantations are known as Frassinetti.However, of late years, attempts have been made to cultivate it on a large scale formanna in the southern parts of the Austrian empire. In Sicily trees begin to producemanna when they are about eight years old; and they are tapped in subsequentyears annually until they are about twenty years old, when they are cut down andtheir place taken by coppice shoots from the stools. During July and Augusttransverse incisions are made in the bark, so as just to reach the wood; and themanna exudes as a clear liquid, which solidifies on the stem of the tree or onpieces of straw or wood that are inserted in the incisions. Manna consists mainlyof a peculiar sugar called mannite, which is a mild laxative and is employed as anofficinal drug in many countries. (A. H.)REMARKABLE TREESThe manna ash is said to have been introduced into England by Dr. Uvedaleof Enfield about 1710. It is commonly cultivated as an ornamental tree, on accountof its beautiful appearance when in flower; and it thrives and attains a large size,especially in the southern parts of England. The largest I have measured is an old1 Tenore, in Syll. PI. Fl. Neap. 1 0 (1831) considers Lamarck's plant to be a variety of F. Ornus. Both Tenore andBertolini, Fl. Italica, \. 54 (1833), were of opinion that F. rotundifolia, Willdenow, was a distinct species. Lingelsheim, inEngler, Bot. Jahrb. xl. 212, 213 (1907), retains F. rotundifolia, Tenore, loc. fit., as a distinct species, confined to a smallarea in south Tyrol, liosnia, and Dalmatia; and creates a new species, F. cilicica, occurring in Cilicia in Asia Minor. Thesesupposed species appear to be glabrous forms of F. Ornus.2 Cf. Bettelini, Flora Legnosa del Sottoceneri, 1 45 (1904).IV 2 A

890 The Trees of Great Britain and Irelandtree in the park at Godinton, Kent, which is about 60 feet high, 10 feet 4 inchesin girth below the graft, and 8 feet 8 inches above it.There is a fine tree growing at Escot, Ottery St. Mary, Devonshire, theresidence of Sir John Kennaway, which was in 1905, according to Miss F. H.Woolward, 61 feet in height, with a girth of stem at 5 feet up of 6 feet 9 inches, thecircumference at the base being 11 feet 9 inches. This tree flowers abundantly everyyear. Another at Carclew is 50 feet by 9^ feet.A tree, about 60 feet high, grows near the stables at Tottenham House,Savernake, the seat of the Marquess of Ailesbury ; a photograph of this tree, whichwas taken in June 1908, from the roof of the stables, shows it in full flower.There is a good tree at Kew, close to the North Gallery, which measured, in1907, 42 feet by 5 feet; and another, on the mound, near the Cumberland gate, is60 feet high and 7 feet 6 inches in girth at three feet from the ground, dividingabove into three main stems.At Syon there is a tree 62 feet by 6 feet 6 inches, and good trees occur atBeauport, Sussex, and Whiteaway, Devon. A very fine one is reported by theHon. Vicary Gibbs to be growing at Rook's Nest, near Oxted, Surrey. AtBrocklesby Park, Lincolnshire, a tree was 50 feet by 7 feet i inch in 1904.In Scotland there is a tree at Gordon Castle about 50 feet high.In Ireland, a tree at Fota measures 65 feet by 5 feet 6 inches; and trees about40 feet high are growing at Narrow Water, near Warrenpoint, in Down, and atGlenstal, Limerick.Henry saw specimens in the botanic gardens at Copenhagen and Christiania,about 40 feet high and 3 feet in girth, in 1908. (H. J. E.)FRAXINUS FLORIBUNDAFraxinusfloribttnda, Wallich, List, 2 836, and in Roxburgh, Fl. Ind. i. 150(1820); Hooker, Fl. £r.India, iii. 605 (1882); Gamble, Man. Indian Timbers, 4 71 (1902); Brandis, Indian Trees,443 (ô6 )-Ornttsfloribunda, A. Dietrich, Sp. PI. i. i, 249 (1831).A large tree, attaining in the Himalayas over 100 feet in height. Shoots compressed, purple, glabrous, with scurfy glands; lenticels white and prominent.Leaflets (Plate 264, Fig. 17), seven to nine (rarely five), the upper pair subsessile,the others on glabrous petiolules, 4 to 6 inches long, oblong, except the terminalone which is obovate, base unequal and rounded, apex long-acuminate, midriband principal veins prominent and pubescent beneath, regularly and sharply serrate.Rachis of the leaf winged, the wings enclosing a broad open groove on the upperside, pubescent in the groove and at the insertions of the leaflets.Flowers (section Ornus) in large panicles; corolla lobes, ^ to \ inch, linearoblong.Samaras very narrow, obtuse or emarginate.Fraxinus 891This is the largest and finest of the flowering ashes, and attains a great size inits native home, Sir D. Brandis mentioning trees in the Chenab Valley planted nearvillages and temples, which reach 120 feet in height and 15 feet in girth. It is theonly valuable species of ash in the Himalayas, where it grows on rich moist soils,generally on limestone. It is distributed throughout the Himalayas from the Industo Sikkim, between 5000 and 9000 feet, but is only common locally. It is also metwith in Baluchistan, Afghanistan, and in the Shan Hills of Upper Burma.We have seen no large trees of this species in England; but it appears to dowell at Kew, and probably would succeed as an ornamental tree in at any rate thewarmer parts of the British Isles. (A. H.)FRAXINUS BUNGEANAFraxinus Bungeana, I)e Candolle, 1 Prod. viii. 275 (1844); Franchet, PI. David, i. 203 (1884);Hemsley, Journ. Linn. Sot. (-#

892, The Trees of Great Britain and IrelandFRAXINUS MARIESIIFraxinus Mariesii, J. U. Hooker, Bot. Mag., 6678 (1883); Hemsley, Journ. Linn. Soc. (£ot.\ xxvi.86 (1889).A small tree. Branchlets slender, terete, purple, minutely pubescent, especiallytowards the tip; lenticels white. Leaflets (Plate 264, Fig. 18) five, i£ to 3 incheslong, coriaceous, oval, acute or acuminate at the apex, base rounded or slightlytapering; regularly and crenately serrate, occasionally entire in margin; glabrousbeneath. Petiolules f to f inch, scurfy pubescent. Rachis of the leaf purple, finelypubescent, grooved on its upper side.Flowers very showy in erect panicles from the uppermost axils, about as long asthe leaves. Calyx minute, four-cleft. Petals five or six, £ inch long, linear-oblong.Stamens two to four, as long as the petals.This was discovered by Maries in the Lushan Mountains near Kiukiang incentral China; and it has not been found elsewhere by subsequent collectors.Maries sent home seeds in 1879 from which the plant was raised by Messrs.Veitch. It flowered 1 for the first time, at Coombe Wood, as early as 1882.It appears to be perfectly hardy in England, and is an ornamental small tree orshrub of considerable value, on account of the creamy-white large panicles of flowers,which appear about the end of June, and the bronze tint of the foliage.(A. H.)Fraxinus 893in its upper half, sometimes almost entire; upper surface glabrous; lower surfaceglandular and glabrous except for some pubescence along the midrib towards thebase. Terminal leaflet largest, basal leaflets smallest. Rachis of the leaf with awide and open shallow groove on its upper side, pubescent in its whole length andbearded at the insertions of the leaflets, the pubescence being continued on theupper side of the petiolules.Flowers (section Ornaster) in compact terminal panicles, polygamous, withouta corolla; calyx persistent at the base of the samarae, which are long, narrow,and erect on filiform pedicels.This is the large ash tree 1 which is common in the mountains about Peking;and it also occurs in the adjacent parts of Mongolia, and in Manchuria andin northern Korea. It was discovered 1 by Pere D'lncarville in the eighteenthcentury. Dr. Bretschneider sent seeds to the Arnold Arboretum in 1881, andplants were raised there, which are growing vigorously and promise to become largetrees. They had already in 1893 produced flowers and fertile seed.The species is remarkably distinct, and is very different from F. chinensis,Roxburgh, of which it has been supposed to be a variety. It appears to be scarcelyknown in Europe, the only specimens which I have seen being from Tortworth,where small plants are reported to be growing badly ; and from Aldenham, wherethe foliage of a young tree, growing freely, is remarkable for its large size and glossyappearance. (A. H.)FRAXINUS MANDSHURICAFRAXINUS RHYNCHOPHYLLAFraxinus rhynchophylla, Hance, Journ. Bot., vii. 164 (1869); Franchet, PI. Davidiance,, \. 2 03, t. 17(1884); Sargent, Garden and Forest, vi. 484, fig. 70 (1893); Komarov, Fl. Manshurm, iii.248 (1907).Fraxinus Bungeana, Maximowicz, Mtt. Biol. ix. 396 (1874) (not De Candolle).Fraxinus chinensis, Roxburgh, var. rhynchophylla, Hemsley, Journ. Linn. Soc. (Bot.), xxvi. 86(1889).A large tree. Shoots glabrous, lenticels few and scattered. Terminal budsremarkable, obtuse, conical, somewhat four-sided, with four or six outer scales, whichare ovate, strongly keeled, with acute points directed outwards, pubescent and grey onthe dorsal surface, and densely ferruginous woolly pubescent on their inner surfaceand edges; lateral buds, ovoid to rounded, small, with two to four outer scales,directed outwards at an open angle.Leaflets (Plate 265, Fig. 23) five to seven, coriaceous, stalked or subsessile,ovate to ovate-lanceolate; base cuneate; apex with a long acumen, which is bluntor rounded and tipped with a short mucro; margin remotely and crenulately serrate1 Hortus Veitchii, 367 (1906).Fraxinus mandskurica, Ruprecht, Bull. Phys. Math. Acad. Sc. Pttersb., xv. 371 (1857); Maximowicz,Prim. Fl. Amurensis, 1 94, 390 (1859), and Mel. Biol. ix. 395 (1874)5 Hemsley, Journ. Linn.Soc. (Bot.), xxvi. 86 (1889); Komarov, Fl, Manshurics, iii. 248 (1907).A large tree, attaining 100 feet in height and 12 feet in girth, with bark likethat of the European ash. Branchlets glabrous. Leaflets (Plate 266, Fig. 28), sevento thirteen, 3 to 5 inches long, oblong-lanceolate, sessile, or with very short pubescentstalklets, tapering and unequal at the base, long-acuminate at the apex, sharply andirregularly serrate; glabrescent above; under surface with scattered coarse hairson the sides of the midrib and lateral nerves. Leaf-rachis, with dense tufts of rustybrowntomentum at the nodes, winged, the wings meeting above in part of its length,elsewhere deeply grooved. 2Flowers (section Fraxinaster) dioecious, in panicles in the axils of the leafscarsof the preceding year, without calyx or corolla. Fruit, in loose clusters,oblong-lanceolate, apiculate or emarginate.1 See Bretschneider, European Bot. Discoveries in China, 5 3, 336, 1058 (1898). There are two ashes in the Pekingmountains, one a large tree, F. rhynchophylla; the other a small shrub, F. Btmgeana.2 The rachis of the uppermost two leaves is usually fringed at its insertion, close to the terminal bud of the branchlet,with rusty-brown pubescence.

894 The Trees of Great Britain and IrelandThe Manchurian ash is the representative of the common ash in eastern Asia,and is very similar in appearance to F. nigra? being mainly distinguishable by thelonger points to the leaflets, which are more tapering at the base, often shortlystalked, and usually more sharply serrate. (A. H.)F. mandshurica is widely spread throughout Manchuria, Amurland, Korea,Saghalien, and Japan, and is a large tree, Ruprecht having measured specimensat the mouth of the Ussuri 12 feet in girth.In Japan this fine ash is known as yachidamo, and is one of the commonesttrees in Hokkaido, but is only known to occur in the north of the main island.I did not see it in Aomori or Akita. It seems to grow best in Hokkaido in thedeep rich flats of black alluvial soil which are now being rapidly brought undercultivation by the Japanese settlers, but even here seldom attains the dimensions ofthe European ash; the average size of the trees in the virgin forests being from80 to 100 feet high by 6 to 8 feet in girth, though no doubt, if they had roomto spread they would grow much thicker than this. The general habit of the treeand of its leaves and seed is very similar to that of F. excelsior, and the woodalso seems similar, but it is not apparently used for the same purposes as in Europe.A great quantity of it is made into railway sleepers, which are now being exportedlargely to Korea and China, and which, if I can judge from what I saw on theHokkaido railroads, will not last very long unless creosoted. There is, however,a particularly handsome variety of this wood, which seems to be found only nearthe butt and on the outside of old trees growing in damp places, which goes bythe name of Tamu, and which, if known in Europe would certainly command a highprice for veneers. A large wardrobe, which has been made for me from Japanesewoods, is fronted with veneer cut from a billet of this wood, which I brought homein 1904, and is extremely handsome; showing a figure like that of the bestHungarian ash but of a pale pinkish-brown colour. This wood is much used inJapan for veneering railway carriages, for doors, and for the posts used forsupporting and for fitting the sliding screens of Japanese houses. It is liableto warp, however, and requires very careful seasoning to prevent cracking.F. mandskurica was introduced 2 at Kew in 1891 from the St. PetersburgBotanic Garden; but only one plant now survives, which has a stunted and unhealthyappearance. Like many plants from Manchuria and Amurland, the foliage appearsearly in the spring, and is badly injured by late frosts every year. I have raisedseedlings from seed sent me from Japan by Professor Shirasawa, but they are tooyoung to judge of the probable success of the tree in England.Sargent, however, says 3 that this tree has proved perfectly hardy in the ArnoldArboretum, where it has been introduced for some years. He considers it one ofthe noblest of all the ashes, and one of the most valuable timber trees of EasternAsia. None of the other ashes of Japan, so far as I saw, attain any great size, orare likely to have any economic value in Europe. (H. J. E.)1 Lingclsheim, in Engler, Bot. Jahrb. xl. 223 (1907), unites this species with f. nigra, of which he considers theManchurian tree to be only a geographical variety.2 The St. Petersburg plants were probably raised from seed from Amurland. Cf. Gartenflora, xxvii. 13 (1879).3 Forest Flora of Jaf an, 5 2 (1894).Fraxinus 895FRAXINUS CHINENSISFraxinus chinensis, Roxburgh, Fl. Ind. i. 150, Carey's edition (1820); Hanbury, Science Papers,271, fig. 17 (1876); Hance, Journ. Bot. xxi. 323 (1883) ; Hemsley, Journ. Linn. Soc. (Hot.)xxvi. 85 (1889); Lingelsheim, in Engler, Bot. Jahrb. xl. 216 (1907).A small tree. Branchlets glabrous. Leaflets seven to nine, 2 to 4 inches long,coriaceous, shortly cuspidate at the apex, crenately serrate ; terminal leaflet the largest,obovate-oval, and long-stalked ; lateral leaflets oval or elliptic, subsessile or withshort pubescent winged stalklets, unequal and broadly rounded or tapering at thebase ; dark green and glabrous above ; under surface pale green and pubescent on thesides of the midrib and lateral nerves. Leaf-rachis deeply channelled throughout,with brownish tufts of tomentum on the upper side of the nodes ; base of therachis of the uppermost two leaves fringed with brown hairs.Flowers (section Ornaster) in terminal and lateral glabrous panicles; calyx4-toothed ; corolla absent. Fruit, about i^ inch long, \ inch wide, oblanceolate,acute or rounded at the apex.This species in the wild state is very variable as regards the shape of the leaves,five varieties being distinguished by Lingelsheim. The above description is drawnup from small specimens cultivated at Kew, which were received in 1891 fromSt. Petersburg under the name F. Bungeana. A distinct variety has been incultivation at Kew for some years under the erroneous name F. longicuspis, whichhas five or seven leaflets, obovate-lanceolate or narrow-oblong, 2 to 3^- incheslong, about i inch wide, cuspidate at the apex, cuneate at the base, indistinctlyserrate. It agrees exactly with a dried specimen at Kew, gathered in the Ningpomountains, where the broad leaflet form also occurs.This species is widely spread throughout the central and southern provinces ofChina, and is noteworthy as being one of the trees on which the wax insect lives.It is very rare in cultivation, the only specimens we have seen being those at Kewand a small shrub at the Edinburgh Botanic Garden. (A. H.)FRAXINUS OBOVATAFraxinus obovata, Blume, Mus. Bot. Lugd. Bat. 3 11 (1850); Franchet et Savatier, Enum. PI. Jap.i. 310 (1875).Fraxinus Bungeana, Maximowicz, Mil. Biol. ix. 396 (1873) (not De Candolle); Franchet etSavatier, Enum. PI. Jaf. ii. 434 (1879); IJngelsheim, in Engler, Bot. Jahrb. xl. 214 (1907)(in part).A small tree. Branchlets glabrous. Leaflets five or seven, 2 to 3 inches long,membranous, terminal one largest and long-stalked ; lower pairs shortly-stalked,

896 The Trees of Great Britain and Irelandobovate or oval; cuneate and unequal at the base ; acute, acuminate or rounded atthe apex ; remotely and irregularly serrate ; upper surface shining, smooth, glabrous ;lower surface light green, glabrous, but roughened on the midrib and nerves byminute, curved, stiff bristles, which are also present on the petiolules and on themargin of the blade. Rachis of the leaf, grooved throughout, pubescent on theupper side of the nodes, and armed near the nodes and at the base with minutecurved bristles, which are also occasionally present towards the apex of the branchlet.Flowers (section Ornus), in terminal glabrous panicles, on filiform pedicels; calyxwith five long acuminate teeth. Fruit linear, i^ inch long, ^ inch wide.This species, of which we have seen the type specimen, preserved at Leyden,and a specimen at Kew, lately received from Tokyo, differs from all the ashes knownto me, in the occurrence of characteristic minute curved prickles on the undersurface, margin, and petiolules of the leaflets and on the rachis of the leaf. Blumeconsidered it to be only cultivated in Japan, and possibly an introduction fromChina It appears to be unknown to Japanese botanists. However the onlyliving specimen which we have seen is a small plant at Aldenham, which was raisedfrom seed obtained from the Imperial Garden at Tokyo. (A. H.)FRAXINUS PUBINERVISFraxinus pubinervis, Blume, Mas. Bot. Lugd. Bat. i. 311 (1850); Franchet et Savatier, Enum. PI.Jap. i. 311 (1875), and ii. 435 (1879); Lingelsheim, in Engler, Bot. Jahrb. xl. 214 (1907).Fraxinus Bimgeana, De Candolle, var. pubinervis, Wenzig, in Engler, Bot. Jahrb. iv. 170 (1883).A small tree, with smooth bark. Young branchlets densely covered withgreyish pubescence, disappearing in the second year. Buds conical, densely coveredwith a brownish-grey pubescence. Leaflets, five or seven, coriaceous, 3 to 4 incheslong, i to ij inch wide ; terminal largest, long-stalked and often broadest in itsupper half; lateral, upper pair subsessile, lower pairs shortly stalked ; lanceolate,cuneate at the base, acuminate at the apex, crenately serrate ; upper surface glabrous ;lower surface pale green, with dense whitish pubescence on the side of the midrib,spreading to the lateral nerves, and continued on the petiolules. Rachis of the leaf,with a continuous open groove, pubescent throughout, the pubescence densest at thenodes. Flowers (section Ornus), in large, terminal glabrous panicles, with earlydeciduous petals and long pedicels ; calyx with long acuminate teeth. Fruit linearspatulate,acute, i^ inch long, ^ inch wide.This species is known to the Japanese as toneriko; and in a native book onforest trees, is said to attain 30 feet in height and 3 feet in girth, but no accurateaccount of its distribution or habitat is given. There is a dried specimen at Kew,lately received from Tokyo. The only tree in cultivation known to me, is one atAldenham, about 15 feet high, which was received six years ago from theYokohama Nursery Company, and is growing vigorously.Fraxinus 8 97FRAXINUS SPAETHIANAFraxinus Spaethiana, Lingelsheim, in Engler, Bot. Jahrb. xl. 215 (1907).Fraxinus Sieboldiana, Dippel, Laubhohkunde, i. 63, t. 27 (1889); Koehne, Deutsche Dendrologie,(1893) (not Blume).A small tree. Branchlets glabrous, grey. Leaflets seven to nine, coriaceous,4 to 6 inches long, i^ to if inch wide, sessile or subsessile (terminal leaflet sessileor with a stalk up to ^ inch long); lanceolate-oblong; unequal and tapering at thebase; abruptly contracted at the apex into a long, often curved acumen; marginnon-ciliate, irregularly and often crenately serrate ; lateral nerves fifteen to twentypairs ; glabrous except for slight pubescence along the midrib on the lower surfaceat the base. Rachis of the leaf, glabrous, with a continuous open groove on itsupper side, dilated at its base into a swollen, dark-brown, shining sheath, whichpartly embraces the branchlet and conceals the glabrous dark-brown buds. Flowers,section Ornus. Fruit in large terminal glabrous leafless panicles. Samarae linearspatulate,i^ inch long, \ inch wide in the broadest part, rounded and entire at theapex ; calyx with five short teeth.This species, which has been a considerable time in cultivation, under thegarden name of F. serratifolia, is readily distinguishable from all the other species ]of ash which have been introduced, by the swollen base of the petiole, whichsomewhat resembles that of the plane tree. It is a native of Japan, where it appearsto have been confused with F. Sieboldiana, which we consider to be a form ofF. longicuspis. Specimens lately received at Kew, through the Hon. Vicary Gibbs,from Dr. Fukuba, Director of the Imperial Gardens at Tokyo, enable us to describethe fruit, which has hitherto been unknown.It is a handsome and striking species, represented at Kew by a tree about15 feet high, and at Aldenham by small plants. (A. H.)FRAXINUS LONGICUSPISFraxinus longicuspis, Siebold et Zuccarini, in Abhand. Baier. Acad. Wissen. iv. 3, p. 169 (1846) ;Franchet et Savatier, Enum. PI. Jap. i. 310 (1875); Shirasawa, Icon. Ess. Forest. Japon, text126, t. 81 (1900); Lingelsheim, in Engler, Bot. Jahrb. xl. 214 (1907).Fraxinus Sieboldiana, Blume, in Mus. Lugd. Bat. i. 311 (1850).A tree, attaining, according to Shirasawa, 50 feet in height and 7 feet in girth.Young branchlets grey, glabrous. Leaflets five, occasionally seven ; terminal largest,stalked; lateral, upper pair sessile, lower pair stalked; oblong, oblong-lanceolate, or1 F. platypoda, Oliver, a species, discovered by me in Central China, has a similar swollen base to. the petiole; but ithas never been introduced.IV 2B

The Trees of Great Britain and Irelandovate, cuneate at the base, abruptly contracted above into a long cuspidate apex,crenately serrate; upper surface glabrous; lower surface pale green, slightlypubescent on the side of the midrib, elsewhere glabrous. Rachis of the leaf, groovedon its upper side, pubescent at the nodes, glabrous elsewhere. Buds purplish brown,minutely pubescent. Flowers (section Ornus) in terminal and lateral panicles;petals four, narrowly-linear. Fruit about an inch in length, subtended by a 4-toothedcalyx, oblanceolate, with a long cuneate base and an obtuse emarginate apex.This species is variable as regards the shape of the leaflets; and two varietieshave been noticed, regarded as distinct species by Blume; one characterised bybroad ovate leaflets (terminal leaflet 3 to 3^ inches long and ij inch broad), withslight pubescence along the midrib beneath ; the other with narrow oblong leaflets(terminal leaflet 3^ to 4 inches long and i to i^ inch wide), glabrous on the undersurface; but, as Franchet points out, there are numerous specimens with intermediate characters.Fraxinus longicuspis 1 is a native of the mountainous districts of Japan, attaining, according to Shirasawa, an elevation of 5000 feet in the central chain of Hondo.It is abundant in Nikko, Chichibu, and Kiso ; and has been collected near Hakodateby Maximowicz, in Akita by Elwes, in the mountains of Yamagata by Faurie, and inthe island of Tsu-sima by Wilford. On account of its small size, it is of no economicimportance in Japan. It appears to be extremely rare in cultivation in this country,the only specimen which we have seen being a small plant at Kew, about 2 feethigh, which was raised from seed sent by Sargent in 1894. (A. H.)FRAXINUS NIGRA, BLACK ASHFraxinus nigra, Marshall, Arb. Am. 5 1 (1785); Sargent, SilvaN. Amer. vi. 37, tt. 264, 265 (1894),and Trees JV. Amer. 764 (1905).Fraxinus sambudfolia, Lamarck, Diet. ii. 549 (1786) ; Loudon, Arb. et Frut. Brit. ii. 1234 (1838).A tree, attaining in America 90 feet in height and 5 feet in girth of stem, withbark divided into large irregular scaly plates. Shoots glabrous. Leaflets (Plate 264,Fig. 19), seven to eleven, 3 to 5 inches long, terminal one petiolulate, lateral leafletssessile; oblong to oblong - lanceolate, rounded and unequal at the base, longacuminate at the apex, remotely and finely crenulate - serrate; under surfaceglabrous, except for long reddish hairs along the nerves and midrib, densesttowards the base, where they spread over the surface of the leaflet. Leaf-rachiswinged on the upper side, the wings meeting above and not forming a continuousopen groove; glabrous except opposite the insertions of the leaflets, where it isbearded all round with dense rufous hairs.The terminal buds are blackish, broadly ovate and acute, with six scales visibleexternally, of which the outer pair, slightly puberulous, almost enclose the others.The Japanese botanists recognise two species : F. longicuspis, called in Japanese osliida or aotago; and F, Sieboldiana,known as shioji. The latter name may possibly refer to F. Spaethiana. Cf. p. 897.Fraxinus 899Flowers (section Fraxinaster) polygamous, without calyx or corolla, arising inthe axils of the leaf-scars of the preceding year's shoot. Fruit, linear-oblong, with ashort flattened faintly nerved body, surrounded by a thin wing very emarginateat the apex.This is the American representative * of the common European ash, and it iseasily distinguished from nearly all the other species by the dense ring of rufouspubescence on the leaf-rachis at the nodes. Fraxinus mandshurica, which has therachis similarly bearded, but more deeply grooved, differs in having leaflets with asharp serrate margin and a tapering base. (A. H.)Fraxinus nigra is found, according to Sargent, in deep, cold swamps and on thelow banks of streams and lakes, from southern Newfoundland and the north shoresof the Gulf of St. Lawrence to Lake Winnipeg, and southward to the mountains ofVirginia, southern Illinois, and north-west Arkansas. Macoun says that in Canadait is more widely distributed than the white ash, and more abundant than the latterthroughout its range, from Anticosti to eastern Manitoba, in swamps and riverbottoms. It grows on peat mosses, but remains small in such situations. I saw itin the woods about Ottawa, but of no great size, and Sargent gives 80 to 90 feetas its extreme height. Ridgway found it abundant in Knox County, Indiana,where a tree 83 feet high, of which the bole was 57 feet, only measured \\ footin diameter at 5 feet from the ground. He says that it presents so very close aresemblance to the young Pecan tree (Carya oliv&formts) as not to be readilydistinguished except by experts. 2 In Carya, however, the leaves are alternate. Itswood, according to Emerson, is remarkable for toughness, and on this accountwas preferred to every other by the Indians for making baskets, and is still used forthat purpose in preference to every kind of wood except that of a young white oak.We have seen no specimens of this tree except small plants 3 at Kew andColesborne, although it was probably planted in many places early in the nineteenthcentury; having been introduced, according to Loudon, in 1800. Like many othertrees of the Atlantic side of North America, it is short-lived and does not thrive inour climate. Prof. Sargent informs me in a letter that at the Arnold Arboretum,near Boston, it is one of the most difficult of all trees to grow. At Angers inFrance, it does badly on its own roots, but succeeds when well grafted ; a specimenthere having attained 15 feet high in eight years. (H. J. E.)1 According to Cobbett, Woodlands, Art. 136 (1825), the seeds of this species, like the English ash, do not come upuntil the second year.2 Mr. G. B. Sudworth, however, tells me that the bark of the two is so distinct that they can readily be distinguished.He adds that other species of ash, hickory, oak, Liquidambar, and Nyssa, are now used for basket-making.3 These plants were raised from seed sent from Michigan in 1895, and are not thriving, having been repeatedly injuredby frost. They are now only 2 feet high.

900 The Trees of Great Britain and IrelandFRAXINUS ANOMALA, UTAH ASHFraxinus anomala, Watson, King's Rep. v. 283 (1871); Sargent, Silva N. Amer. vi. 39, t. 266(1894), and Trees N. America, 7 65 (1905).A small tree, attaining about 20 feet in height, with bark shallowly fissured bynarrow ridges. Branchlets quadrangular, slightly four-winged, glabrous, with inconspicuous reddish lenticels. Leaves (Plate 262, Fig. 7) simple (occasionally twotothree-foliolate), ovate or obovate, acute or rounded at the apex, base cuneateor cordate, slightly crenulate or entire in margin, glabrous beneath. Petioleflattened and grooved above, about half the length of the blade of the leaf.Flowers (section Leptalix) in panicles from the axils of the leaf-scars of thepreceding year's shoot, with a calyx, but corolla absent. Fruit oblong, with a roundedwing surrounding the long, flattened, striately nerved body.This curious tree, so remarkable amongst the ashes, in its usually simple leavesand quadrangular stems, occurs in Colorado, Utah, and Nevada.It is in cultivation at Kew, where it is perfectly hardy, and is worthy of a placein collections on account of its peculiarities. (A. H.)Fraxinus 901arboretum of the Experimental Farm. Sargent says that it grows usually on richlimestone hills from southern Michigan to Iowa, central Missouri, and north-easternArkansas, and southward to northern Alabama and east Tennessee on the BigSmoky mountains, where it attains a great size. Usually it is from 60 to 70 feethigh, with a trunk 2 to 3 feet in diameter; but Ridgway says that four freshlycut trees, in the Wabash valley in Illinois, were 116 to 124 feet high, with cleartrunks 51 to 76 feet long, and 2 to z\ feet diameter on the stump. Here it wascommon in rich hilly woods, but I saw none standing of anything like thesedimensions.The tree was discovered by the elder Michaux in 1795 and introduced 1 by himinto France, and his son speaks of the beautiful stocks that were growing in Europe ;but I have seen none of considerable size, the best perhaps in England beinga tree at Tortworth, which was 34 feet by i foot 10 inches in November 1905, andreported by Lord Ducie to be growing freely. Michaux says that the wood of thisspecies in the western states is extensively used for waggon-building and wheels,and also for flooring houses; but it does not now seem to be known to Englishimporters. Mr. G. B. Sudworth informs me, however, that this timber is still foundin considerable quantity in the Ohio valley, and can be obtained in logs as large as24 to 30 inches in diameter. He adds that for farm tool handles it is preferred toany other ash on account of its superior strength and elasticity. (H. J. E.)FRAXINUS QUADRANGULATA, BLUE ASHFraxinus quadrangulata, Michaux, Fl. Bar. Am. ii. 255 (1803); Loudon, Arb. et Frut. Brit. ii.1235 (1838); Sargent, Silva N. Amer. vi. 35, t. 263 (1894), and Trees N. Amer. 7 61 (1905).A large tree, attaining 120 feet in height and 9 feet in girth. Bark separatingirregularly into large thin plates. Branchlets glabrous, stout, quadrangular; withfour wings between the nodes, persisting and becoming corky in the second year;lenticels white. Leaflets (Plate 265, Fig. 22), five to nine, 3 to 5 inches long; onpubescent stalklets, £ to \ inch long; ovate or oval, rounded or broadly cuneate atthe base, acuminate at the apex, regularly serrate, glabrous above; under surfacecovered with scattered whitish tomentum, densest towards the base. Leaf-rachispubescent, with a shallow open channel on its upper side.Flowers (section Leptalix) in panicles from the axils of the leaf-scars of theprevious year's shoot, perfect, calyx obsolete, corolla absent. Fruit oblong; bodylong, flat, with numerous faint rays, surrounded by the base of the broad wing, whichis emarginate at the apex.This species is readily distinguished from all the other species with numerousleaflets by the conspicuous wings on the branchlets. (A. H.)The blue ash though little known outside of botanic gardens in Europe is, nextto the white ash, the largest of its genus found in the United States. It is unknownin New England or Canada, though hardy at Ottawa, where I saw it in theFRAXINUS AMERICANA, WHITE ASHFraxinus americana, Linnseus, Sp. PI. 1 057 (1753); Loudon, Arb. et Frut. Brit. 1 232 (1838);Sargent, Silva N. America, vi. 43, tt. 268, 269 (1894), and Trees N. America, 7 67 (1905).Fraxinus alba, Marshall, Arbust. Am. 5 1 (1785).Fraxinus acuminala, Lamarck, Diet. ii. 547 (1786).Fraxinus juglandifolia, Lamarck, Diet. ii. 548 (1786).Fraxinus epiptera, Michaux, Fl. Bor. Am. ii. 256 (1803).A large tree, attaining in America 120 feet in height and 15 to 20 feet in girth.Bark deeply divided by narrow fissures into broad flattened scaly ridges. Shootsstout, green, glabrous, with white lenticels. Leaflets (Plate 265, Fig. 21), sevento nine, 4 to 6 inches long, distinctly stalked (the petiolules glabrous and i to înch long), lanceolate or oval, rounded or broadly cuneate at the base, acuminate atthe apex; entire, crenulate, or coarsely serrate; under surface whitish and pubescentalong the midrib and nerves, or in some cases throughout. Rachis of the leafterete and apparently not grooved on the upper side; but usually a slight groovecan be made out on close examination.Flowers (section Leptalix) dioecious, in glabrous panicles in the axils of the leafscarsof the preceding year's shoot; corolla absent. Fruit in crowded clusters, surrounded by the persistent calyx at the base, lanceolate or oblong, with a terete rayedoblong body, much shorter than the terminal wing, which is pointed or emarginateat the apex.1 It was introduced, according to Loudon, into England in 1823.

002 The Trees of Great Britain and IrelandVARIETIESA form is known with very small fruit, var. microcarpa. Gray. 1 In regard tothe leaf, two varieties have been described, 2 viz.:Var. acuminata, Wesmael. Leaflets dark green above, very white and almostglabrous beneath, nearly entire in margin. This form is more common in thesouthern states.Var. juglandifolia, Rehder. Leaflets usually broader than in the precedingvariety, more or less pubescent beneath, coarsely serrate at least above the middle.This is the northern form of the species.These two varieties occur in cultivation in England, the leaves of both remainingunchanged in colour until they fall in autumn. A form of var. acuminata occurs,in which the leaflets are narrow at the base, and turn reddish brown before they fall.There is also said to exist a horticultural variety, var. albo-marginata, in whichthe leaflets are edged with white.IDENTIFICATIONAs Fraximts americana may be a valuable tree for economic planting in England,its correct identification is important. Reputed trees of F. americana growingslowly usually turn out on examination to be F. viridis or some other species.In summer, the leaflets white beneath and distinctly stalked, the rachis tereteand practically not grooved, and the glabrous branchlets, will readily distinguishF. americana. The only species which closely agrees with it, F. texensis, has, whenadult, fewer, usually five (rarely seven) leaflets, which are smaller and quitedifferent in shape, being broadly oval with a rounded or acute apex. (A. H.)DISTRIBUTIONThe white ash is one of the best known and most highly valued trees in NewEngland and Canada; where it occurs according to Sargent, from Nova Scotiaand New Brunswick through Ontario to northern Minnesota, southward to northernFlorida, and westward to Kansas, Indian Territory, and eastern Nebraska, and thevalley of the Trinity River, Texas; being less common and smaller west of theMississippi. It attains in the forest a height of 120 feet with a diameter of 4 feetand upwards, and thrives best in a deep loamy soil near the banks of streams,just as the common ash does in England. When standing alone it assumes aspreading habit with large branches.Ridgway measured a sound tree in Wabash County, Illinois, which was 144feet high with a clean stem 83 feet long and 9 feet in girth at the top, and 13 feetat the base, which, according to English measure, would have contained over 500cubic feet; and Dr. C. Schneck measured a tree in the same county 144 feet highwith a stem 90 feet long and 17^ in girth above the swell at the base; and this tree,if it carried its girth up, might have contained 1000 feet of timber.1 This variety is common in the Gulf States, the fruit being less than \ inch long; whereas it usually attains \\ to 2inches. Sargent, however, states that both large and small fruit may occur on the same individual, and even on the samebranch- 2 Cf. Rehder in Cycl. Auier. Horticulture, 607 (1900).Fraxinus 9 3I saw no large specimens of this species in New England or Canada wheremost of the best have been felled. I noticed at Ottawa that in the autumn itsleaves assume a rich purplish colour, which the black and red ashes did not show.Michaux says 1 that on large trees the bark is deeply furrowed, and dividedinto small squares from i to 3 inches in diameter; and that it grows in Maine incompany with the white elm, yellow birch, white maple, hemlock, and black spruce,and in New Jersey with red maple, shellbark hickory, and button-wood (Platanusoccidentalis), in places that are constantly wet or occasionally flooded. Pinchot andAshe 2 figure a splendid tall straight forest tree with small head and rough barkresembling that of the common ash, and say that its average height in North Carolinais 50 to 80 feet with a diameter of 2 feet to 3 feet.Professor Sargent says 8 that the white ash when planted with the common ashin the high regions of central Europe comes into leaf still later than that species,and thus escapes the spring frosts; it is less able to resist drought than the greenash, and is usually found on moist soil, though it does not like wet swamps, like theblack ash. In the forest it sends up a perfectly straight and slender stem toa great height, ash poles 100 feet high and not over a foot in diameter beingoften seen. A photograph taken in Chester County, Pennsylvania, shows a tree inthe open with a rather spreading head and a trunk 15 feet 10 inches in girth.CULTIVATIONFraxinus americana is stated by Aiton * to have been introduced into Englandin 1724 by Catesby. Sargent, however, points out that the ash described by Catesbyis another species.Though very seldom seen, I believe that the American ash will grow in thiscountry, in some places at least, almost as fast as the native ash. My attention wasfirst called to this fact by the very straight, clean, and rapid growth of a youngtree at Kew which stands by the walk not far from the Old Conservatory, now theMuseum of Timbers. The date of planting is unknown. It measured in 190763 feet by 2 feet 5 inches, while F. lanceolata, growing near it is 43 feet by 2 feeti inch, and F. oregona, 44 feet by 2 feet 8 inches.Knowing that the wood is considered better for oars than that of the native ash,and used exclusively for the heavy oars of our navy, I thought it worth trying asa timber tree, and raised a large number of plants from seed sent me by Messrs.Meehan of Philadelphia as F. americana, but which I found out three years later tobe F. lanceolata Later on I raised seedlings of the true white ash and found that atfirst they do not grow nearly so fast as F. lanceolata. They do not ripen the young1 N. Amer. Sylva, iii. 49. 2 Trees of North Carolina, p. 71, plate 6.3 Garden and Forest, vii. 402. * Hort. Kew. iii. 445 (1789).6 Mr. G. B. Sudworth informs me that ten or fifteen years ago Messrs. Meehan became aware that they had been sellingtwo or three species of ash seed as F. americana, and submitted samples to him for identification. To his great surprise hefound that none of it was pure, but contained a mixture off. americana, F. lanceolata, and F. pennsytsanica. Since then theyhave been more careful. He adds that continental tree planters have, to his knowledge, been planting green and red ash inmistake for white ash ; and it is probable that most of the young trees grown nnder that name on the continent are incorrectlynamed.

904 The Trees of Great Britain and Irelandwood well until they attain a certain age, but are perfectly hardy and will probablygrow best in the east and south-east of England.The only large trees I know of in England are two at Kew which grow onthe mound near the Cumberland gate and which measure 85 feet by 8 feet 6inches and 85 feet by 8 feet respectively (Plate 246). There is a tall slender whiteash at Croome Court, Worcestershire, crowded by other trees, which measures about80 feet by 4 feet 6 inches, and another 1 at Arley Castle 60 feet by 5 feet 4inches. There is also a smaller one at Syon. Another at Tortworth was 46 feetby 3 feet in 1907. Loudon says that at St. Anne's Hill in Surrey there was one33 feet high which had been planted thirty-six years, and that near London youngplants are generally injured by spring frost, which I have not found to be the caseat Colesborne.Sir Charles Strickland tells me that he has planted the tree at Hildenley atYorkshire, but does not find it succeed so well as the Oregon ash.A tree at Fota, in the south of I reland, is 50 feet high and 6 feet in girth.Cobbett says : z " This tree grows much faster than ours. I have abundant proof,for the American white ash plants which I have at Kensington, which were not sowntill last April, are now (1825) full as tall again as any of the English ash of the sameage that I ever saw. This, therefore, is above all others the ash which I recommendto be put into plantations in England, whether for ornament, for timber, or for underwood." But Cobbett in this case, as in many others, was rather apt to jump toconclusions after too short experience; for if the tree had continued to grow asit did at first, there must by this time have been many good-sized ones in England.TIMBERThe timber of the white ash is as highly valued in America as ours is inEngland for the same purposes, and is largely imported to England where it is usedas a substitute for English ash. Laslett in writing of American white ash says:" It is tough, elastic, clean, and straight in the grain, and stands well after seasoning,hence we get from this tree the best material for oars for boats that can be produced.They are much and eagerly sought after by foreign governments as well as our own,and also by the great private steamship companies and mercantile marine. The bestquality wood has a clean, bright uniform whitish colour ; the second is slightly stainedwith red and yellow shades alternating; the third and least valuable quality isthat in which the red and yellow colours predominate. It is much slower in growththan the English, and is probably not so durable." On visiting the principalimporters of this wood in Liverpool, I found large quantities of American oarsimported ready-made; and was told that the timber had now become so scarce in theeast, that it came from the west side of the Mississippi. 3 I have not been able toprocure a sample grown in this country for comparison; but I am indebted to1 Catalogue of Hardy Trees, Arley Castle, No. 43 (1907). " Woodlands, art. 135.3 A lot of oars which I saw in the Portsmouth Dockyard were stamped on the blade "De Vails Bluff. Ark. U.S.A."I am informed by experienced naval carpenters, and officers, that they believe that English ash, if it can be procured ofsufficient straightness and length, would be at least as good, if not better.Fraxinus 90SMr. A. Howard for a plank believed to be of this species, which was importedfrom Canada, and was cut from a log measuring 56 feet long by 24 inches quartergirthat the top. It cost £66 and was cut into twenty-four boards containing 2763feet board measure. (H. J. E.)FRAXINUS TEXENSIS, TEXAN ASHFraxinus texensis, Sargent, Silva N. Amer. vi. 47, t. 270 (1894), and Trees N. Amer. 7 68 (1905).Fraxinus albicans, Buckley, Proc. Philad. Acad. 1 862, p. 4 (in part).Fraxinus americana, Linnteus, var. texensis, Gray, Syn. FI. N. Amer. ii. part i. 75 (1878).Fraxinus americana, Linnteus, var. albicans, Lingelsheim, in Engler, Bot. Jahrb. xl. 219 (1907).A small tree, with a short stem, rarely attaining 50 feet in height and 9 feetin girth. Bark dark grey, deeply divided by narrow fissures into broad scaly ridges.Branchlets glabrous. Leaflets, five to seven, often nine in young plants, 2 to z\inches long, distinctly stalked, with glabrous petiolules about \ to \ inch long, ovateor oval (the terminal leaflet often obovate), base rounded and unequal, apex acute orrounded (acuminate in young plants), crenate-serrate and non-ciliate in margin ;upper surface shining bluish green, glabrous; lower surface whitish, usuallypubescent on the sides of the midrib and lateral nerves, the latter being forked nearthe margin ; rachis slender, glabrous, terete, slightly grooved on the upper side.Flowers (section Leptalix) dioecious in glabrous panicles in the axils of theleaf-scars of the preceding year; corolla absent. Fruit, spatulate, with persistentcalyx at the base; body short, terete; wing terminal, rounded or emarginateat the apex.This species, while very close to F. americana in technical characters, isdistinct in appearance, and differs in the smaller leaflets, which are shining bluishgreen, and not dull on the upper surface.It was discovered in 1852 by Dr. J. M. Bigelow, and grows on high dry limestone bluffs and ridges in northern, central, and western Texas, from near DallasCity to the valley of the Devil's River.Young plants are growing in the nursery at Kew, which are thriving, andabout 6 to 8 feet in height, after six years' growth. They were raised from seedsent by Mr. Bush in 1901. These plants have usually 7 to 9 leaflets, whilespecimens in the Kew Herbarium from Texas have only five leaflets ; but this isprobably a juvenile character. It is also growing well at Aldenham.(A. H.)IV 2 C

906 The Trees of Great Britain and IrelandFRAXINUS BILTMOREANA, BILTMORE ASHFraxinus Biltmoreana, Beadle, Bot. Gazette, xxv. 358 (1898); Sargent, Trees N. Amer. 7 73, fig. 618Fraxinus catawbiensis, Ashe, Bot. Gazette, xxxiii. 230 (1902).A tree attaining in America, according to Ashe, over 100 feet in height, with agirth of about 7 feet. Young shoots covered with a dense white pubescence, retainedin the second year; lenticels few, conspicuous, narrow, long, white. Leaflets (Plate266, Fig. 30), seven to nine, about 4 inches long, oval or oblong (the terminalone on a long stalk, broadly oval or obovate), abruptly tapering and unequalat the base, acuminate at the apex, remotely serrate (the serrations oftenobsolete, so that the margin is nearly entire), with occasional scattered cilia ; distinctlystalked with pubescent petiolules, f to \ inch long ; upper surface dark green,glabrous except for a little pubescence towards the base of the leaflet ; lower surfacewhite in colour, with a thin fine short pubescence, densest on the sides of themidrib and nerves. Rachis of the leaf slender, terete, finely pubescent, not groovedor only slightly grooved towards the apex.Flowers (section Leptalix) dioecious in pubescent panicles in the axils of theleaf-scars of the previous year ; corolla absent. Fruit girt at the base by thepersistent calyx ; body, short, elliptical, many- nerved ; wing not decurrent, onlyslightly narrowed at the ends, emarginate at the apex.Buds shortly ovoid, with four outer visible scales, equal in length, the externalpair overlapping the inner pair ; scales carinate, obtuse at the apex, orange-coloured,and covered with a scaly pubescence.This species, as regards leaf- characters, looks like a pubescent Fraxinusamericana ; 1 and is readily distinguished from F. pennsylvanica by the leaflet beingwhite in colour beneath, with a finer pubescence, and being more abruptly taperingand unequal at the base, with shallower and remoter serrations, which often becomeobsolete. The rachis of the leaf is like that of F. americana.According to Sargent, it occurs on the banks of streams from northern WestVirginia through the foothills of the Appalachian Mountains to northern Georgia andAlabama, and to middle Tennessee.This species in the United States has apparently been considered to be a formof F. pennsylvanica, and has been in cultivation probably as long. At Fawley Court,Oxfordshire, the residence of W. D. Mackenzie, Esq., there are two fine trees. Thelargest, in a shrubbery, rather crowded by other trees, is about 80 feet in height by7 feet 3 inches in girth. The other, standing in the open, is a very well-shapedvigorous tree (Plate 247), measuring 68 feet by 6 feet 6 inches. Both are bearingmistletoe and grow in good alluvial soil. The bark is grey in colour, and fissuredlike that of the white ash. (A. H.)1 Beadle says that this species bears the same relation to F. americana as F. pennsylvanica bears to F. lanceolate.Lingelsheim, op. cit. 191, 222, considers this species to be a hybrid, between F. americana and F. pennsylvanica ; but itswide distribution and abundance in the forest are not favourable to his view.Fraxinus 907FRAXINUS LANCEOLATA, GREEN ASHFraxinus lanceolata, Borkhausen, Handb. Forst. Bot. i. 826 (1800); Sudworth, Check List ForestIrees U.S. 1 07 (1898).Fraxinus vt'ridis, Michaux, Hist. Arb. Atner. iii. 115, t. 10 (excl. fruit), (1813).Fraxinus juglandifolia, Willdenow, Sp. PL iv. 1104 ( I 8os) (not Lamarck); Loudon, Arb. et Frut.Brit. ii. 1236 (1838).Fraxinus pennsylvanica, Marshall, var. lanceolata, Sargent, Silva N. America, vi. 51, t. 272 (1894), andTrees N. America, 7 71 (1905).A tree rarely attaining more than 60 feet in height, with a girth of stem of6 feet. Shoots dark green, glabrous, with conspicuous white lenticels. Leaflets(Plate 264, Fig. 16), seven to nine, usually subsessile, 3 to 6 inches long, palegreen beneath, ovate-lanceolate, tapering and unequal at the base, long-acuminateat the apex, under surface glabrous except for slight pubescence along the midrib,variable in serration. Rachis of the leaf glabrous, distinctly grooved on its upper side.Flowers and fruit similar to those of F. pennsylvanica.This species' is considered by Sargent to be a variety of F. pennsylvanicabecause west of the Mississippi trees occur, which are intermediate in character andcan be as readily referred to one species as to the other. As seen in cultivation inEngland, it is very distinct, and on account of its glabrous shoots, it is very oftenmistaken for F. americana. The green ash is, however, readily distinguished fromthat species by the usually subsessile leaflets, which are pale green and not whitebeneath. 2 The rachis is, moreover, more deeply grooved than is ever the case inF. americana, and the buds in the two species are different. (A. H.)This tree was described and figured by the younger Michaux, who says, " Thegreen ash is easily recognised by the brilliant colour of its young shoots and leaves,of which the two surfaces are nearly alike." He found it more common in thewestern districts of Pennsylvania, Maryland, and Virginia, than anywhere else, andspeaks of it as a tree of moderate dimensions, laden with seed when only 25 to 30feet high.Sargent, who treats it as a variety off. pennsylvanica, and distinguishes it by theleaves being rather narrower, shorter, and usually with more sharply serrate leaflets,bright green on both surfaces, says that it rarely exceeds 60 feet high, and occursfrom the shores of Lake Champlain through the Alleghany mountains to westernFlorida and west to the valley of the Saskatchewan, the valley of the Coloradoriver, Texas, the eastern ranges of the Rocky Mountains, the Wasatch range, Utah,and Arizona. It is comparatively rare east of the Alleghany mountains, and mostabundant in the Mississippi basin. East of the Mississippi it seems distinct, butwestward is connected with the red ash by intermediate forms.1 A small tree growing in Kew Gardens and labelled F. coriacea, resembles in many respects the green ash. It isapparently the Texan form of the Mexican F. Berlandieriana, DC., which was formerly considered to be a variety ofF. lanceolata. The Kew tree has glabrous shoots like those of the green ash; but the leaflets (Plate 263, Fig. 14) aresmaller, more coriaceous and very reticulate, usually three to five in number, the terminal one largest and obovate, the lateralones oval and acuminate at the apex, all sparingly and irregularly serrate in the upper two-thirds, with the under surface palegreen and glabrous except for some pubescence along the midrib. Rachis of the leaf glabrous and slightly grooved.F. coriacea, Watson, a tree occurring in desert regions from Utah to California is probably a variety of F. Berlandieriana,differing in the pubescent branchlets and leaf-rachis. It is in cultivation at Aldenham, where a young tree, about 10 feet higb,obtained from Barbier's nursery at Orleans, is thriving.2 Mr. G. B. Sudworth, adds, that the leaflets are usually sharply serrate, while those of F. americana are commonlyundulate, entire, or with only a few teeth.

908 The Trees of Great Britain and IrelandRidgway, in his Additional Notes on the Trees of the Wabash? says that Dr.J. Schneck of Mount Carmel measured a tree 92 feet high and 5 feet in girth; butwhen I visited the remains of this wonderful forest, in September 1904, I saw noash trees of considerable size.Pinchot and Ashe 2 say that the wood is inferior in quality to that of thewhite ash, but in North Carolina is not distinguished from it commercially.Having raised a large quantity of plants from seed sent me as that ofFraxinus americana, which I did not identify as the green ash until Mr. F. V. Covilleof Washington saw them growing in my nursery in 1904, I have distributedthem to many friends as F. americana, and it is probable that the tree will thusbecome common in England under a wrong name, as has happened in so many casesbefore. For this mistake, which was unavoidable, I now apologise; but as the treegrows faster than any other American ash in a young state, and is likely to makeuseful poles, if not large trees, I have planted out some thousands of them atColesborne.Like all the American ashes which I have raised, the seed germinates quicklyafter sowing, and though liable to be injured by late frosts is at least as hardy as thecommon ash. When young the shoots continue to grow late in autumn and do notripen their young wood, which for three to four years at least is liable to be killedback by winter frosts. Some of these seedlings are now, at four years old, 6 to 7feet high and growing very vigorously. Michaux says that this species wasintroduced by his father to France in 1785, but I cannot hear of any survivingunder this name.Loudon says that at Stackpole Court, Pembrokeshire, it had in forty yearsattained a height of 60 feet, and had ripened seeds from which many plants had beenraised and distributed in the plantations, but the Earl of Cawdor tells me that hisgardener can find no trees in the woods which resemble the American ash, and thatnone of the men on the place can remember any peculiar ash trees there. In 1906 Ialso searched the woods at Stackpole without finding any trace of these trees.Loudon also mentions a tree in the garden of Pope's villa at Twickenham, whichno longer exists. There are several young trees in Kew Gardens ; and the tallest,about 40 feet high, is widely branching in habit, differing remarkably from a whiteash of the same height beside it, which has narrow branches and a straighter stem.Their foliage is also very different in colour. (H. J. E.)FRAXINUS PENNSYLVANICA, RED ASHFraxinus pennsylvanica, Marshall, Arb. Amer. 5 1 (1785); Sargent, Silva N. Amer. vi. 49, t. 271(1894), and Trees N. Amer. 7 70 (1905).Fraxinus pubescens, Lamarck, Diet. ii. 548 (1786); Loudon, Arb. et Frut. Brit. ii. 1233 (1838).A tree, attaining 60 feet in height and 5 feet in girth of stem. Bark brownishred and slightly furrowed, with scaly ridges. Young shoots stout, covered with1 Proc. U.S. Nat. Mils. xvii. 4 11 (1894). 2 Timber Trees of North Carolina, 7 3.Fraxinus 909dense white pubescence, which is retained in the second year; lenticels white,inconspicuous. Leaflets (Plate 263, Fig. 13), seven to nine, occasionally five, 3 to5 inches long, three times as long as broad, ovate-lanceolate to oblong-lanceolate,tapering at the base, acuminate at the apex, finely serrate and ciliate in margin ;upper surface with scattered fine pubescence; lower surface densely pubescentand green in colour. The leaflets are usually distinctly stalked, with pubescentpetiolules; but forms occur in which they are subsessile, the substance of theleaflet being prolonged to its insertion. Rachis of the leaf densely white pubescent,with a distinct shallow groove on its upper side.Flowers (section Leptalix) dioecious, in tomentose panicles in the axils of theleaf-scars of the preceding year's shoot; corolla absent. Fruit linear - spatulate,surrounded at the base by the persistent calyx ; body slender, terete, many-rayed ;wing slightly decurrent, narrow, and rounded or acute at the apex.For the distinctions between this species and the Oregon ash, see under thelatter. Fraxinus Biltmoreana differs conspicuously in having the leaflets whitebeneath. Fraxinus prof unda, Bush, 1 which is remarkably distinct in its fruit, differsalso in having the leaflets entire or undulate in margin, their base being usually veryasymmetrical.There are several forms of F. pennsylvanica in cultivation, some having theleaflets very firm in texture and set close on the rachis, others having thin leafletswider apart on the rachis. The leaflets also vary in the length of their stalklets, inthe size of the serrations, and in the shape of the base, which may be graduallytapering or abruptly tapering and almost rounded. F. Richardi, F. Boscii, andF. glabra, names given to certain horticultural varieties, are all probably referable tothis species.Var. aucubczfolia (F. aucubcefolia, Kirchner, A rb. Muse. 507 (1864)), in whichthe leaves are variegated with yellow, is considered by Lingelsheim 2 to be a hybridbetween F. pennsylvanica and F. lanceolata. At Aldenham 8 this forms a handsometree about 30 feet high. (A. H.)It is neither so large nor so common a tree as the white ash in the United Stateswhere, according to Sargent, it has nearly the same distribution as the latter; beingmost common and largest in the north Atlantic States, smaller and less abundantwest of the Alleghanies. Macoun says 4 that in Canada it ranges farther west thanthe white and black ashes, growing along the Assiniboine river and the tributaries ofLake Manitoba. It is usually 40 to 60 feet high, with a diameter rarely exceeding18 inches to 20 inches; and is here of no value for timber, but makes good firewood,even when green. Emerson measured a tree at Springfield in September 1840which was 9 feet in girth at 3 feet from the ground; and Ridgway says 5 thatDr. Schneck measured a tree in the Wabash forests 138 feet high by 16 feet ingirth.1 See Sargent, Trees N. America, 772. This ash, which is probably not yet introduced, grows to a great size inriver swamps in Missouri, Arkansas, and Florida. It is considered by Lingelsheim, in Engler, Bot. Jahrb. xl. 220 (1907), tobe a variety of F. pennsylvanica. 2 Op. cit. 222.3 It is cultivated here under the erroneous name, F. americana, var. aucubtffolia, which is given in Kew Handlist of Trees,533 (1902). < Forest Wealth of Canada, p. 23. 6 Proc. US. Nat. Mas. xvii. 411 (1894).

910 The Trees of Great Britain and IrelandFraxinuspennsylvanica was introduced into England in 1783*; and it is oftenmet with in cultivation in public parks and botanic gardens, where it grows well asa small tree. One in the Botanic Garden at Oxford measures about 50 feet by3 feet, and I have seen smaller trees at Stowe and elsewhere. I have raised it fromAmerican seed, and it seems to grow as fast as the white ash, but not so fast as thegreen ash. It ripens its wood better, and when young loses its leaves earlier thaneither of these.Macoun says 2 that the red ash and the green ash are not separated commercially from the other species, the wood of the latter resembling that of the whiteash, while that of the former is more like the black ash. Therefore there is somedoubt whether Laslett, who writes of the Canadian ash, whose timber is oftenconfounded with that of the white ash, is speaking of this tree or of the black ashwhich he does not mention. He says that it was, until recently, imported in considerable quantity in the form of oars, and that it is reddish brown in colour, considerablydarker than the wood of the English ash. (H. } . E.)FRAXINUS OREGON A, OREGON ASHFraxinus oregona, Nuttall, Sylva, iii. 59, t. 99 (1849); Sargent, Silva N. Amer. vi. 57, t. 276 (1894),and Trees IV. Amer. 7 76 (1905).A tree attaining 80 feet in height and a girth of stem of 12 feet. Bark deeplydivided by interrupted fissures into broad flat scaly ridges. Young shoots stout,covered with dense white tomentum, which persists in the second year; lenticelswhite, inconspicuous. Leaflets (Plate 263, Fig. 15), 3 to 4 inches long, subsessile,usually seven, sometimes five or nine, oval, about twice as long as broad, baserounded or abruptly tapering, apex acute or shortly acuminate; margin entire orminutely and remotely crenate, ciliate; upper surface with scattered fine pubescence;lower surface covered with dense white tomentum. Rachis white tomentose, with adistinct shallow groove on its upper side, basal part wide and flattened.Flowers (section Leptalix) dioecious in glabrous panicles rising out of theaxils of the preceding year's shoot; calyx present, persisting under the fruit, corollaabsent. Fruit obovate-oblong; body slightly compressed; wing long, decurrent,many-nerved, and rounded, apiculate or emarginate at the apex.Fraxinus oregona can only be confused with sessile forms of F. pennsylvanica,which has longer serrate leaflets, with more nerves, tapering gradually to the base.In F. oregona the leaflets are shorter in proportion to their breadth, and are usuallyentire in margin; but this last character is not absolutely distinctive, as the leafletsof the two species vary in regard to the presence or absence and size of theserrations.1 Aiton, Hort. Kew. v. 476 (1813). 2 Forest Wealth of Canada, p. 23.Fraxinus 911This species was discovered by David Douglas in 1825 on the banks of theLower Columbia river. It is mentioned by Koch 1 as having been in cultivationin the Botanic Garden of Berlin prior to 1872. (A. H.)This tree is the common ash of north-west America, extending from PugetSound through western Washington and Oregon and the coast region of Californiato San Francisco, and along the western foothills of the Sierra Nevada to SanBernardino and San Diego counties. It is an important timber tree in Oregon andWashington, but if it occurs in British Columbia, 2 is too scarce to be noticed byMacoun. I saw no trees of great size during either of my visits to the Pacificcoast, but Sheldon 3 says that it attains as much as 100 feet in height and i to 4 feetin diameter, and describes the wood as hard, tough, firm, straight-grained, and takinga high polish.In California, according to Jepson,4 F. oregona only attains a length of 15 to 30feet, growing along the Sacramento river, and on the banks of the streams of thecoast ranges. The leaflets become glabrous with age ; and on this account Lingelsheim6 has distinguished the Californian form as var. glabra.In England the tree grows well in a young state; but I know of none of anysize, except one at Nuneham Court, Oxfordshire, which in 1907 was 63 feet highand 3 feet 8 inches in girth. A tree of F. americana growing near it, and believedto have been planted at the same time, was only 38 feet by i foot 8 inches. SirCharles Strickland has a plantation at Hildenley, Yorkshire, of this species mixed withlarch; and as these trees, when I saw them in 1901, were 20 to 30 feet high, and hadripened seed from which I have raised plants, I think their hardiness in this countryis abundantly proved; and that the tree is more likely to succeed in the westand north of England than the eastern American ashes. At Tortworth CourtLord Ducie has made a mixed plantation of F. americana and this species, whichare now 20 to 30 feet high at about fifteen years old; but the soil and situation arenot very favourable. In 1908 the Oregon ash were bearing seed freely.(H. J. E.)1 Dendrokgie, ii. I, p. 260 (1872).2 Piper, Flora of the State of Washington, 449 (1906), says its range is " British Columbia to California, in the coastregion" ; but gives no localities for British Columbia. It is not mentioned as occurring in Vancouver Island by the authorsof Postelsia, the year-book of the Minnesota Seaside Station for 1906.3 Forest Wealth of Oregon, 3 2 (1904). 4 Flora W. Mid. California, 3 85 (1901).6 In Engler, Bot.Jahrb. xl. 220 (1907). .

912, The Trees of Great Britain and IrelandFRAXINUS CAROLINIANA, SWAMP ASHFraxinus caroliniana, Miller, Did. No. 6 (1768); Loudon, Arb. et Frut. Brit. ii. 1237 (1838);Sargent, Silva N. Amer. vi. 55, tt. 274, 275 (1894), and Trees N. Amer. 7 62 (1905).Fraxinusplatycarpa, Michaux, Fl. Bar. Amer. ii. 256 (1803).Fraxinus triptera, Nuttall, Gen. ii. 232 (1818).Fraximts atbensis, Grisebach, Cat. El. Cub. 1 70 (1866).A tree attaining 40 feet in height, with a stem 3 feet in girth ; bark marked byirregularly - shaped brown patches, separating on the surface into thin scales.Branchlets glabrous or pubescent, with white minute scattered lenticels. Leaflets(Plate 263, Fig. 12) seven, occasionally five, stalked (petiolule \ to f inch), about 3inches long, oval; unequal, rounded, or broadly cuneate at the base; apex shortlyacuminate; finely and irregularly serrate; green and glabrous on the under surfaceexcept for some white pubescence along the sides of the midrib and nerves, or insome forms pubescent throughout. Leaf rachis, glabrous or pubescent, with twoslight wings on the upper side, forming a groove.Flowers (section Leptalix) dicecious in panicles arising in the axils of leaf-scars ofthe preceding year's shoot; calyx present, persisting under the fruit; corolla absent.Fruit broad, elliptic or spatulate; body short and compressed, surrounded by apinnately-veined broad thin wing.This species grows in river swamps in the coast regions of the Atlantic andGulf States from southern Virginia to the valley of the Sabine river in Texas,extending through western Louisiana northwards to south-western Arkansas.It also occurs in Cuba.This species was introduced into England in 1783, according to Loudon, who,however, mentions no trees of any size as growing in England in 1838. We haveseen no specimens, except small trees at Kew, which are thriving. (A. H.)Fraxinus 9*3apex; coarsely serrate in the upper half or two-thirds, but sometimes entire, ciliate ;upper surface pubescent; lower surface densely white pubescent. Rachis of theleaf white pubescent and deeply grooved on its upper side.Flowers (section Leptalix] dicecious in short panicles in the axils of the leafscarsof the preceding year's shoot; calyx present and persisting under the fruit;corolla absent. Fruit spatulate, with terete body and terminal wing.This species is readily distinguished by the dense white pubescence over theshoot, leaf-rachis, and leaflets, the latter being variable in number and smallerthan those of the other pubescent ashes, except F. xanthoxyloides, which has stillsmaller leaflets with a broadly winged rachis and is much less strongly pubescent.Fraxinus velutina occurs usually in elevated canons beside streams in Texas,New Mexico, Arizona, southern Nevada, and south-east California.Young trees are doing well at Kew, where they are of considerable interest fromtheir peculiar foliage, which gives them a neat and elegant appearance. The oldest,planted in 1891, are now about 15 feet high. (A. H.)FRAXINUS VELUTINAFraxinus velutina, Torrey, Emory's Report 1 49 (1848); Sargent, Silva N. Amer. vi. 41, t.(1894), and Trees N. Amer. 7 74 (1905).Fraxinus pistacicefolia, Torrey, Pacific ft. Report, iv. 128 (1856).267A tree, 40 feet high, with a girth of stem of 2 feet. Bark deeply divided intobroad flat broken scaly ridges. Shoots purple, covered with dense white pubescence ;lenticels white. Leaflets (Plate 265, Fig. 20), small, about i£ inch long, three orfive, occasionally seven or nine, or rarely only one, and variable in shape,margin, and insertion ; usually sessile, occasionally stalked, the terminal leaflet oftenobovate, the lateral leaflets commonly lanceolate with cuneate base and acuminateIV 2 D

Zelkova 915Japan, and Z. crenata, Spach, inhabiting the Caucasus and North Persia. The thirdspecies, Z. cretica, Spach, is a shrub growing in Crete and Cyprus, which has not yetbeen introduced, and does not come within the scope of our work. The followingspecies, also a shrub, is only known in cultivation :ZELKOVAZelkova?- Spach, in Ann. Sc. Nat. se"r. 2, xv. 356 (1841); Bentham et Hooker, Gen. PI. iii. 353(1880); Nicholson, in Woods and Forests, 1 884, p. 176.Abelicea, Reichenbach, Consp. Veg. 84 (1828); Schneider, Laubholzkunde, i. 224 (1904).Planera, Gmelin, subgenus Abelicea, Planchon, in Ann. Sc. Nat. se"r. 3, x. 261 (1848).DECIDUOUS trees or shrubs, belonging to the order Ulmacese. Branchlets slender,distichous. Leaves alternate, distichous, simple, shortly stalked, penni-nerved,crenately serrate. Stipules in pairs, membranous, lanceolate, caducous.Flowers monoecious; corolla absent; calyx, four- or five-lobed. Staminateflowers, clustered, two to five together, on the branchlets below the leaves or in theaxils of the lowermost leaves; disc absent; stamens, four or five, with short, erectfilaments and exserted anthers; ovary rudimentary or absent. Pistillate flowers,solitary in the axils of the uppermost leaves; disc cupular, fleshy; staminodespresent or absent; ovary sessile; styles, two, stigmatiferous on the inner side;ovule solitary, pendulous. Fruit, a small drupe, sessile, subtended by the persistentcalyx, subglobose, oblique, veined or rugose on the surface, crowned by the remainsof the styles, persisting as two minute beaks; with a membranous or slightlyfleshy outer covering, and a thin, hard endocarp or stone, containing a compressed,concave, horizontal seed, without albumen. The fruit ripens late in autumn, andpersists on the branchlets till the following spring.In Zelkova no true terminal bud is formed, and the tip of the branchlet falls offin early summer, leaving a small circular scar at the apex of the twig. The base ofthe shoot is ringed with the scars of the inner scales of the previous season's bud,and shows, as a rule, a few of the outer scales persisting dry and membranous.The buds, all axillary, and composed of numerous imbricated scales, are oftenmultiple, two being then developed, side by side, in a single axil. The leaf-scarsare narrow, crescentic, and three-dotted ; with a linear stipule-scar on each side. InUlmus the buds are single in the axils, and none of the scales persist at the base ofthe shoot.Three species of Zelkova 2 are known to exist in the wild state, two of which arelarge trees well known in cultivation, Z. acuminata, Planchon, a native of China and1 The name Zelkova is sanctioned, and Abelicea rejected in Actes Congrts Internal. Bet. Viennt, ^^ (1906).2 Hemiptelea Davidii, Planchon, Compt. Rend. Acad. Paris, Ixxiv. 1496 (1872), a thorny tree, occurring in northernand central China and Korea, is united with Zelkova by Bentham and Hooker, in Gen. PI. iii. p. 353. It diners from thatgenus in having winged fruit. Cf. Schneider, Laiibhohkunde, i . 224. This species does not appear to be in cultivation inEurope.914Zelkova Verschaffeltii, N icholson, Kew Handlist of Trees, 1 45 (1896).Zelkova japonica, Dippel, var. Verschaffeltii, Dippel, Laubholzkunde, ii. 39, fig. 14 (1892).A shrub or small tree. Branchlets slender, pubescent, with white hairs.Leaves (Plate 267, Fig. 8), coriaceous, variable in size, from \\ inch long byf inch wide to 2^ inches long by i^ inch wide, oval, acuminate at the apex,cuneate and unequal at the base, divided by the midrib into unequal halves,the larger half with six to eight nerves, the smaller half with four to sevennerves, each nerve ending in a long triangular tooth, tipped with a shortcartilaginous point; margin ciliate; upper surface dark green, with scattered whitepubescence; lower surface light green, with downy white pubescence, densest on themidrib and nerves; petiole, \ inch to \ inch, pubescent. Buds, often two togetherin an axil, small, globose, pubescent. Fruit similar to that of Z. crenata, butslightly smaller in size.This species, which resembles an elm in having asymmetrical oblique leaves, wasconsidered by Schneider' to be a peculiar variety of Ulmus glabra, and is occasionally met with in cultivation, as Ulmus Verschaffeltii, and Ulmus pendula laciniataPittcursii.A tree, 15 feet high, in the nursery of the Paris Municipality, specimens of whichhave been sent us by M. Vacherot, produced flowers and fruit this year; and thefruit, hitherto unknown, proves to be that of a Zelkova. Z. Verschaffeltii is notknown in the wild state, though Koehne 2 states that O. Kuntze collected specimensof Z. crenata in the Caucasus, which strongly resembled it. It is possibly a hybridbetween Z. crenata and Z. cretica, and was first noticed by Dippel in 1892.(A. H.)ZELKOVA CRENATAZelkova crenata, Spach, Ann. Sc. Nat. xv. 358 (1841); Boissier, Fl. Orientalis, iv. 1159 (1879).Zelkova carpinifolia, Dippel, Laubholzkunde, ii. 38 (1892).Zelkova ulmoides, Schneider, Laubholzkunde, i . 806 (1906).Rhamnus ulmoides, Giildenstadt, Itin. i. 313 (1787).Rhamnus carpinifolius, Pallas, Fl. Rossica, i. 2, 24 (1788).Planera Richardi, Michaux, Fl. Bar. Am. ii. 248 (1803); Loudon, Arb. et Frut. Brit. iii. 1409Planera carpinifolia, Watson, Dendrol. Brit. 1 06, t. 106 (1825); Koch, Dendrol. ii. i. 425 (1872).Planera crenata, Desfontaines, Cat. Hort. Paris (1829).Abelicea ulmoides, Kuntze, Rev. Gen. ii. 621 (1892); Schneider, Laubholzkunde, \. 2 24 (1904).A tree attaining about 100 feet in height, and 1 5 feet in girth. Bark thin, smooth,greyish-brown, marked with persistent lenticels ; on older trees, scaling off in small1 Laubhohkunde, i. 226 (1904). * Deutsche Dendrologie, 1 37 (1893).

916 The Trees of Great Britain and Irelandirregular plates. Branchlets covered with dense, white pubescence. Leaves (Plate267, Fig. 6), slightly coriaceous, about 3 inches long and i J inch wide, oval-lanceolate,acute at the apex, cordate and unequal at the base, with nine to eleven pairs ofnerves, each ending in a crenate serration, the apex of which is minutely pointed;margin ciliate; upper surface dark green, with scattered minute pubescence; lowersurface more or less covered with white pubescence, densest on the midrib andnerves; petiole, | to ^ inch, pubescent.Fruit, about J inch long, pubescent, with a very slight depression on eitherside of the prominent ridge on the upper surface.In winter the twigs are pubescent, and bear elongated conical buds, which arebrownish, tinged with white in colour, owing to the scales, which are glabrous on thesurface, being fringed with long, white cilia.This species is readily distinguished by the pubescent oval leaves, acute andnot acuminate at the apex, the serrations only showing minute points. Z. acuminatahas glabrous ovate leaves, with a long acuminate apex, the serrations ending in long,sharp, often recurved points.DISTRIBUTIONThis species occurs in the Russian provinces, lying south of the main range ofthe Caucasus, and in northern Persia, in the territory bordering on the Caspian Sea,extending as far eastward as Asterabad.The best account of its distribution in Transcaucasia is given by Scharrer, 1 whostates that it grows wild in two distinct areas, one in the government of Kutais andthe other in Talysch, while there are a few scattered trees at Araxes, in theKarabagh district. In Kutais it grows in the Mingrelian plain, east of Sennakh,and ascends in the lower mountains of Imeritia to about 1000 feet, occurring at lowlevels in small groups in the oak forests, and at higher elevations mixed with ash,maple, and beech, and never forming pure woods. In Talysch it is not found onthe marshy plains, but is common in river valleys, ascending on the mountains to5000 feet, and often forming pure and dense woods. Scharrer measured a tree, 100feet in height, with a stem 8 feet in diameter and free of branches to sixty feet.The climate in which it thrives is humid, with a rainfall of 50 or 60 inches; andit requires a moist, permeable, rich soil to come to perfection. At Tiflis, however,where the rainfall is only 19^ inches, the tree is met with growing on good loamysoil and on rocky mountain slopes, but is slower in growth than in Mingrelia. Ithas borne without injury a temperature of - 24 C.The elder Michaux, 2 who travelled in Persia in 1782, and saw the tree growingin the forests of Ghilan, states that it commonly attains a height of 80 feet, with agirth of 9 to 12 feet, with a straight trunk, branching at about thirty feet up, andresembling the hornbeam in its bark, fluted trunk, and mode of branching. Scharrer,however, states that in the forest it produces clean stems, very uniform in thickness1 Gartetiflora, xxxvi. 187 (1887).2 Cf. Andrc Michaux, Mfyioire stir le Zelkoua (Paris, 1831),Zelkova 917and free from branches, except at the summit, excelling in this respect most broadleavedtrees. It grows fast in youth, continuing its growth in height to sixty oreighty years old, afterwards mainly increasing in girth. It is moderate in its demandsfor light, and gives good coppice shoots.CULTIVATIONThis species was introduced into cultivation in 1760, the oldest known tree inEurope being one x in the garden of M. Lemmonier at Petit Montreuil, near Paris,which was cut down in 1820, when it was 72 feet in height and 6 feet 8 inches in girth.It is probable that the elder Michaux, who saw this tree growing in northern Persia in1782, also introduced seed. 1 Further consignments 1 were sent to France in 1831,by Chevalier Gamba, French Consul at Tiflis. Seeds from this source germinatedafter lying eighteen months in the ground; but Gay, in a note in the KewHerbarium, states that this tardy germination was probably accidental, as seeds fromKarabagh, which he sowed in the last days of March, produced seedlings, whichwere peeping out of the ground at the end of May. (A. H.)This tree is now rarely seen in nurseries, though it is easily propagated bysuckers, and seed could be procured without difficulty from its native country.In consequence it is hardly known to modern gardeners, though both from itsornamental habit and valuable timber it would be much better worth planting thanmany trees of more recent introduction.The principal point to be attended to is to protect it from frost, and prune itcarefully until the main stem has attained the desired height; and to plant it in adeep, rich alluvial soil, and warm, sheltered situation. So far as I have been able tolearn, no tree has produced fertile seed in this country.REMARKABLE TREESA remarkable tree at Wardour Castle, Tisbury, Wiltshire, is reputed by traditionto have been sent, when quite young, from North America by the second LordBaltimore, about 1632, and has been supposed 2 to be Planera aquatica, Gmelin, 3 anative of swamps in the south-eastern United States. There must be some errorin the tradition, as the tree is undoubtedly Zelkova crenata. It is known as the Irontree, and the late Lord Arundell of Wardour assured me that this name, used inAmerica for the Hornbeam and Hop Hornbeam, was a proof of its American origin.He also believed that the tree had been cut down during the siege of WardourCastle in Cromwell's time, and had afterwards produced from the stool the seventeen1 Andre Michaux, MAiioire stir le Zelkoua (Paris, 1831).2 See a lengthened correspondence concerning this tree in Garden, xxiv. 370 (1883), xxvi. 38 (1884), and xxxii. 92(1887).3 This species was introduced into England in 1816, according to Loudon, Art. et Frut. Brit. m. 1413 (1838); but itappears to be unsuitable for our climate, and no specimens are known to us to exist, except two plants in the Elm collectionat Kew, about 8 feet high, which were introduced in 1897, and are thriving so far.

918 The Trees of Great Britain and Irelandtall stems which it now shows (Plate 248). These measure from 5 to 8 feet in girth,the whole forming a group about 12 feet wide, and some of them reaching nearly 100feet in height. There are many small suckers, as usual, and some of these havebeen transplanted to the front of the present mansion, where they were, when I sawthem in 1903, about 25 feet high. Lord Arundell was good enough to give me tworooted suckers from the old tree, one of which is now planted at Tortworth Court,and the other at Colesborne.At Holme Lacy, Hertfordshire, the seat of the Earl of Chesterfield, there is avery fine tree standing on a walled depression near the house. Its bole was untilrecently surrounded by laurels, which have now been cleared away, so that I wasable in October 1908 to measure it carefully. I found it to be about 95 feet high by19 feet in girth at the smallest point, about 3 feet from the ground. A photographof this tree is at Kew, and it was reported in 1884 to be 70 feet high.There are several trees in Kew Gardens, the largest growing in front of theHerbarium, and measuring 60 feet in height and 9 feet 3 inches in girth. Alarger specimen was cut down a good many years ago, and a section of thetrunk is exhibited in the Timbers Museum.There are several fine trees at Syon, 1 the largest near the lake being, in 1905,98 feet in height and 12 feet 7 inches in girth, while another is 89 feet by 13 feet.Both these trees are remarkable for their buttressed stems. Near the bridge thereis a slender specimen, crowded by other trees, which is 92 feet high, with a stem7 feet 2 inches in girth, and free of branches to about 50 feet.At Albury there is a remarkable specimen, with a bole of only 4 feet inheight, but 16 feet in girth, dividing into numerous stems.In the Wilderness, at Croome Court, Lady Coventry showed me a tree of thisspecies, which was supposed to be a species of hornbeam. It measured, in 1906, 65feet by 14^ feet, and grew in an angle between two hedges, into which its suckershad spread profusely, and being clipped with the hedge, may eventually formpart of it when the original tree dies. Another tree grows in the TempleShrubberies at the same place, about 70 feet by 7 feet, with a clean trunk about 15feet long, and was of a better shape, but is now partially decayed.At Pitt House, near Chudleigh, Devonshire, the seat of Captain Morrison Bell,there are several Zelkovas which seem to thrive well in this climate. The largestis a very well-shaped and healthy tree measuring 80 to 85 feet by 13 feet. Itsleaves were beginning to unfold on I5th April 1908. Some trees have thrown upsuckers in the hedge by the high road here, but the gardener has not observed anyflowers.At Oxford, in the University Park, on the banks of the Cherwell, there are two,the larger of which, not a well-shaped tree, is about 80 feet by 12^ feet.At Kyre Park there are two trees on the banks of a pond near the house, both ofwhich have begun to decay, and large pieces of the smooth bark were dropping offthem when I visited Kyre in 1904. The largest measured about 75 feet by 16 feetZelkova 9197 inches. Mrs. Childe tells me that in the hot summer of 1905 flowers were produced by this tree.At Belshill, near Belford, Northumberland, the property of Sir W. Church,Bart, there is a fine tree in a sheltered situation which measures 70 feet by 9 feet 10inches, and looks healthy, though it is believed to be over 100 years old.In Scotland we have not seen or heard of any trees, though there is no doubtit would grow well in the south and west, where the climate is much better than atBelshill.There are two trees at Glasnevin, one 50 feet and the other, a remarkably fineone, 61 feet in girth (Plate 249). Both are 9 feet in girth, and divide at 10 feet upinto numerous branches.A tree* at Verrieres, near Paris, is 70 feet in height and 8 feet in girth.There is a large tree in the grounds of the Petit Trianon at Versailles, about90 feet by 10 to 12 feet, which appears to be grafted on the roots of an elm, andMr. Hickel informed me that most of the older trees in France were so grown.This species 2 is represented in the United States at Woodlands, Philadelphia,where there are growing in a cemetery a few low bushy trees, with short trunks,4 feet in diameter, and numerous erect branches.TIMBERAccording to Scharrer the wood is homogeneous, prettily veined, very tough andflexible, does not crack and warp, takes a fine polish, and is very durable even whenplaced in wet situations. It is very suitable for cabinet-work and carriage-building.The native name of the tree, dzelk-wa, signifies " stone-wood," so-called on account ofthe hardness of the timber, into which nails are driven with difficulty.The younger Michaux, who examined a tree cut down at Paris in 1820, statesthat the sapwood is white, and the heartwood reddish in colour, the latter beingheavier and stronger than that of elm, while even the sapwood equalled the ash instrength and elasticity.A plank of this wood cut from a tree which grew at Boynton, in Yorkshire,was given me by Sir Charles Strickland, and resembles the wood of the Japanesespecies in texture and colour. Mrs. Baldwyn Childe has also sent me a specimen ofit from a branch of her tree. Though unknown in the trade, and, as far as I canlearn, never cut for export, I believe that this wood would prove valuable formaking furniture if it could be obtained at a reasonable price. (H. J. E.)1 Hortus Vilmorinianus, $2 ( 1906). Garden and Forest, x. 488 (1897).1 According to Loudon the largest tree at Syon was in 1834, 54 feet high and 2 feet 3 inches in diameter.

The Trees of Great Britain and IrelandZELKOVA ACUMINATAZelkova acuminata, Planchon, in Compt. Rend. Acad. Paris, Ixxiv. 1496 (1872), and DC. Prod. xvii.166 (1873)-Zelkova Keaki, Maximowicz, Mil. Biol. ix. 21 (1872); Sargent, Garden and Forest, vi. 323, fig. 49(1893); Shirasawa, Icon. Ess. Forest. Japan, text 65, t. 36, figs. 1-17 (1900); Mayr,Fremdland. Wald- u. Parkbaume, 5 25, figs. 247-249 (1906).Zelkova serrata, Makino, in Tokyo Bot. Mag. xvii. 13 (1903).Zelkova hirta, Schneider, Laubholzkttnde, i. 806 (1906).Corchorus hirtus, Thunberg, Fl. Jap. 2 28 (1784).Ulmus Keaki, Siebold, Verh. Bat. Gen. xii. 28 (1830).Planera acuminata, Lindley, Gard. Chron. 1 862, p. 428Planera japonica, Miquel, Ann. Mus. Bot. Lugd. iii. 66 (1867).Planera Keaki, Koch, Dendrol. ii. i. 427 (1872).Abelicea Keaki, Schneider, Dendrol. Winterstudlen, 2 38 (1903).Abelicea hirta, Schneider, Laubholzkunde, i. 226 (1904).A tree attaining in Japan 120 feet in height and 15 feet or more in girth.Bark smooth, greyish, resembling that of a beech, but on old trees dividing on thesurface into irregular rounded scaly plates. Branchlets slender, at first with ascattered slight pubescence, but becoming glabrous in summer. Leaves (Plate 267,Fig. 7), membranous, thinner than those of the Caucasian species, about 3 inches longand \\ inch broad, ovate, with a long acuminate apex, slightly cordate and unequalat the base; nerves, nine to twelve pairs, each ending in the long, sharp, oftenrecurved tip of a serration; margin ciliate at first, the cilia falling off in summer;upper surface dark green, with scattered, short, stiff hairs; lower surface lightgreen, at first slightly pubescent, becoming glabrous in summer; petiole r3& to \inch, glabrescent.Fruit about \ inch broad, sub-globose, oblique, with a concave depression oneither side of a transverse ridge on the upper surface, which bears the remains ofthe styles ; prominently veined, glabrescent.In winter the twigs are glabrous, and the buds are minute, ovoid, uniformlybrown in colour, with glabrous slightly ciliate scales.DISTRIBUTIONZelkova acuminata occurs in China and Korea, as well as in Japan, but appearsto be only common in China in the Tsin-Ling mountain in Shensi, where it wascollected by Pere Giraldi. 1 It has also been found a in the province of Chekiang,on the hills near the Taihoo Lake, and on the mountains inland from Ningpo.Carles 2 collected it near Seoul in Korea.It is much more widely distributed in Japan, occurring throughout Kiushiu,Shikoku, and Honshu, forming forests in mixture with maple, beech, Qttercusgrosseserrata, and other broad-leaved trees. According to Shirasawa, it is usually ofbranching habit in the forest, the tallest and stoutest trees being those cultivated1 Diels, in Engler,/a/jri. xxxvi. No. 5, p. 33 (1905). 2 Hemsley, Journ. Linn. Soc. (Bot.), xxvi. 449 (1894).Zelkova 921near houses and temples, which often have clean stems due to pruning. Mayr,however, records a forest tree, 123 feet in height, with a stem 33 inches in diameter,and clean of branches to 57 feet. Dupont, 1 who gives many interesting particularsconcerning this species, states that it ascends in Kiushiu to 3000 feet on northernslopes, and in Honshu to 4000 feet on southern slopes; and that it requires for itsbest development a deep, permeable, and rich soil, such as is found on alluvial tracts.It thrives well also on volcanic soils and sandy loams, but does not succeed on poorsands or on stiff clays. Dupont states that on suitable soils and situations it growsremarkably straight, whether isolated or crowded in the forest, attaining on soil ofmiddling quality at 1600 feet elevation in the latitude of Fuji-yama, 5 feet in girth at60 years old, 9 feet at 120 years, and 12^ feet at 180 years. The growth of isolatedtrees on alluvial soil is still more rapid, the annual rings averaging \ inch in width.He advocates the planting of this tree on account of its rapid growth and the valueof its timber, which he considers to be superior to that of oak. (A. H.)It seems strange that this tree, whose wood is more highly valued by theJapanese than any other hard wood, should be planted on so small a scale in Japan.Probably it requires too many years to come to maturity to induce private persons toplant it when bamboo, Cryptomeria, and pine offer so quick and certain a return.But though I saw no plantations of Keaki, 2 I believe the Government are makingefforts to preserve and increase the area under this species. It is said to be foundwild in the south up to about 5000 feet, and in the north up to about 2000 feet. Italso grows wild in Hokkaido, but not to so great a size as in the north-eastern districtsof the main island which are famous for their large trees. I heard of, but was notable to see, one said to be the largest in Japan at Sendai. Rein 8 speaks of onewhich was felled at Meguro, near Tokio, in 1874, and measured 11.7 metres in girthat one metre high.The largest I measured myself was growing by the side of the Nakasendo road,at a place called Hideshiwa, near the village of Sooga in Shinano, in a grove of treesjust below the road, and may have been wild or planted. It was about 115 feethigh and 20 feet in girth, dividing at about 20 feet, into two tall upright limbs, each10 feet or more in girth, and seemed to contain about 800 to 1000 cubic feet of timber.Close to it was a very large Êsculus, and on the other side of the road another Keaki,113 feet high, by 13 feet 6 inches in girth, clean and straight to about 70 feet high.Such trees as these are found only where they have been crowded when young,the tendency of the species being to assume a branching and spreading habit, so thatmost of these which are seen planted singly are thick and spreading rather than tall.In the forest the Keaki grows scattered among other trees, and is said by theauthor of the handbook on Japanese forestry to love calcareous soils. I never sawany such soil in Japan, but it seemed to grow equally well on all kinds of soil providedit is deep and moist. As to the age to which the tree attains I cannot speakpositively, but it looks like a very long-lived tree. Its bark is smooth and greyishin colour, somewhat like that of the beech. It seeds freely and reproduces itself1 Essences Fortst. du Japan, 45 (1879).The Japanese name of this species is Keyaki, occasionally spelled Keaki.IV3 Industries of Japan, 225 (1889).2 E

922 The Trees of Great Britain and Irelandeasily, the seedlings being hardy and fast growing and bearing shade well. Inappearance the Zelkova is not an ungraceful tree, resembling a beech perhapsmore than an elm, but its small leaves make it a poor shade tree, and its habit ofgrowth varies very much according to the situation in which it grows.CULTIVATIONThis species was introduced into cultivation in England by J. Gould Veitch, 1who sent seeds from Japan in [862. It was apparently introduced on the Continenta few years earlier by Siebold; and Koch, writing in 1872, mentions that it hadbeen cultivated previously for several years in the Botanic Garden at Berlin, whereit had sustained severe frost without injury.I raised a quantity of seedlings in 1901 from Japanese seed, which grewrapidly at first, and seemed quite hardy; but those which I have planted out growslowly, and, where not protected from spring frost, have been killed back everywinter, so that they produce bushy shrubs. I should, therefore, suppose that itrequires more summer heat and moisture than most parts of England afford, and thatit should be planted only in rich, deep soil, where it can be shaded and drawn up byother trees. Careful pruning is also evidently necessary to check its tendency toproduce lateral branches when young, and I do not anticipate that it will ever attainlarge dimensions in Great Britain or be worth planting for its timber.So far as we have ascertained there are no large trees existing in England,the best specimens we have seen at Kew Gardens, at Tortworth, and in LordKesteven's woods at Casewick, Lincolnshire, not exceeding 20 to 30 feet in height.Mr. C. Palmer tells me that in November 1864, he planted a specimen receivedfrom Veitch, then 3 feet high, in an exposed situation, at about 500 feet above thesea, near Stukeley Grange, Leighton Buzzard. This tree in 1874 was 18^ feethigh by 6 inches only in girth. In 1892 it had increased to 2 feet in girth.A good-sized and healthy-looking tree, of whose age no record can be found,grows near the pond below the entrance from Woburn village to Woburn Park onthe right-hand side of the drive.At Kilmacurragh, Co. Wicklow, a healthy tree measured, in 1904, 41 feet inheight and 3 feet i inch in girth.The oldest and largest tree of this species that we know of in Europe is growingin the Botanic Gardens at Carlsruhe. I am indebted to Herr Max Leichtlin, ofBaden-Baden, for a photograph of this tree (Plate 250), and Herr Grabener has alsosent us one. He informs us that it is one of three seedlings which was raised atSt. Petersburg (or brought from Leyden). It was planted sometime between 1859and 1861, and has never suffered from frost, having sustained the severe winter of1879-1880 without injury. 2 In 1904 it measured at ij metre from the ground 3.10metres in girth.1 Hortus Veitchii, 386 (1906). Mr. H. J. Veitch informs me that the plants raised at Coombe Wood only attained 3 to4 feet in height, and gradually died out, the whole being finally lost during a severe winter.2 Mayr says that at Grafrath it has endured 25 Cent, without injury.ZelkovaThis species x has been planted pure in small experimental plots at five differentforestry stations in Germany; but the results have not been encouraging. BothSchwappach and Mayr consider that it probably would succeed if planted in mixturewith other hardwood trees.In the Botanic Garden at Copenhagen, Henry saw, in 1908, a fine tree, about40 feet high, but dividing into three stems at 2 feet from the ground. This tree wasplanted about 1870, and exceeds considerably in size a Z. crenata planted beside itthe same year. At Christiania, Z. acuminata remains a bush about a foot high,being repeatedly cut by frost.The largest specimens in the United States are two trees, growing in Dr. Hall'sgardens at Warren, Rhode Island, which were raised from seed sent home in 1862.According to Sargent, 2 who says that this species is probably the only Japanesetree worth introducing into North America, on a large scale for timber, thesehave received no special care, the soil is not particularly good, and theirgrowth has been checked by overcrowding. They were about 50 feet high in 1893,with trunks about a foot in diameter, and had produced flowers and fruit. Thereappear 8 to be several other trees of this species in the same garden, and hundredsof self-sown seedlings were observed near them in 1893. The only drawback 4 tothe cultivation of this species in America is that it is subject to the attack of theelm-leaf beetle.TIMBERThe wood 5 resembles in structure that of the elm, the vessels being disposed insimilar broken lines. The sapwood, according to Mayr, about i^ inch in thickness,and white in colour, is separated from the brown heartwood by a narrow pink zone.According to Dupont, 6 the wood, while like that of the elm in appearance, is morelike the ash in working, as it bends readily, and is of great strength, surpassing eventhe oak in this quality. It resists exposure to moisture, and is very durable forbuilding purposes. It is much used in Japan for making furniture, an especial kind,cut from burrs, and called jo-rin-moku, being especially esteemed. The mostbeautiful trays and cabinets 7 which come from Japan are made of dark, irregularlygrained, and wavy-lined wood of the Zelkova. Many of these trays are ornamentedwith the bark of Pterocarya. Dupont says that the wood contains an oil, with adisagreeable odour, which prevents it from being made into articles used for containing liquids.The wood is the most valuable in Japan for building, for furniture, and for allpurposes where a strong, tough, durable wood, not liable to warp, crack, or decay,is required ; and it is also valuable for carving and lacquering. It is the highestpriced wood in Japan, worth in ordinary sizes up to 45. per cubic foot, and for finelygrained or very wide planks much more. I was told that a single large plank of1 Schwappach, AnbaTcversuihe mil Fremdlandischen Hoharten, 79 (1901).* Garden and Forest, vi. 324 (1893). 3 Ibid. 468. 4 Ibid. 369.6 Figured in Mayr, op. cit. t. xx. fig. 44. 6 Ess. Forest, dujapon, 45 (1879).7 Sargent, Garden and Forest, x. 40 (1879).

92,4 The Trees of Great Britain and Irelandfinely grained wood of this tree shown at the Osaka Exhibition in 1903 was pricedat 800 yen (over /8o); and I saw beautifully figured pieces myself of a peculiarreddish tint which were held for fancy prices, such pieces being much valued forthe construction of the dais which is a marked feature in Japanese rooms, andon which is the seat of honour. Most of the pillars, beams, gateways, gates, andcarved roofs, which are so striking a feature in Japanese temples, old and new,are made of Keaki wood, which seems indestructible by time or damp when coveredin, and I was told that some of these which looked sound, though much weatherworn, were 1000 years old. It seemed to me, however, that 1000 years is simply aconvenient expression in Japan for anything very old, though no doubt historicalevidence could be found if wanted as to the durability of this fine wood. The fancygrained varieties are known as Jorin, Uzura (partridge), Tama (gem), or Botan(peony), and these are used for cabinetmaking and fancy work. The colour,according to Rein, is deepened by long submersion in water. Rein gives the specificgravity of the wood at 0.682. I bought some of the wood in Japan, and haveused it in making a large wardrobe; it takes polish well, makes good joints, andseems equal to mahogany for furniture making, but so far as I can learn has not asyet been imported, and is unknown in the trade. (H. J. E.)CELTISCeltis, Linnaeus, Gen. PI. 3 37 (1837); Bentham et Hooker, Gen. PI. iii. 354 (1880).Mertensia, Humboldt, Bonpland et Kunth, Nov. Gen. et Spec. ii. 30 (1817).Momisia, F. G. Dietrich, Lexic. Garten. v. 128 (1819).Solenostigma, Endlicher, Prod. Fl. Norf. 4 1 (1833).THE genus Celtis, belonging to the order Ulmaceae, comprises about sixtyspecies, spread over the temperate and tropical regions of the northern hemisphere;and was divided into four sub-genera by Planchon. 1 In the following account, thecharacters of one of the sub-genera, Euceltis, are given, as all the species incultivation belong to it.Deciduous trees, without spines. Leaves stipulate, alternate, distichous, simple,stalked, serrate or entire, usually oblique and three-nerved (rarely four- or fivenerved)at the base, the midrib and basal nerves giving off pinnately secondarynerves.Flowers minute, pedicellate on the branchlets of the year, polygamo-moncecious.Staminate flowers in few-flowered fascicles from the axils of caducous bud-scales.Perfect flowers solitary or in two- to three-flowered fascicles in the axils of the lowerleaves. Calyx four- or five-lobed, imbricate in aestivation, deciduous. Corolla absent.Stamens, four or five, inserted under the margin of a pubescent disc; filaments,subulate, erect and exserted in the staminate flowers, shorter and included, occasionally absent, in the perfect flowers ; anthers two-celled, extrorse, opening longitudinally.Ovary sessile, one-celled, crowned with a short style, divided into two divergent,elongated, reflexed lobes, papillo - stigmatic on the inner surface; ovule solitary,suspended. Fruit, a fleshy drupe, with a firm epicarp, a succulent thin mesocarp,and a thick-walled bony stone, containing one seed. Cotyledons emarginate at theapex, raised above ground in germination.Seven species of Celtis are in cultivation in this country, which may bedistinguished as follows :I. Leaves ovate.* Leaves quite glabrous.i. Celtis Davidiana, Carriere. China. See p. 929.Leaves shining on both surfaces, toothed in the upper third, shortly acuminate,minutely punctate when viewed with a lens.1 De Candolle, Prod. xvii. 169 (1873).925

926 The Trees of Great Britain and Ireland2. Celtis glabrata, Steven. Asia Minor, Caucasus. See p. 929.Leaves bluish green, serrate except near the base, acute or very shortlyacuminate, conspicuously punctate when viewed with a lens.** Leaves pubescent.3. Celtis occidentalis, Linnseus. North America. See p. 930.Leaves caudate-acuminate, serrate in the upper half or two-thirds, smooth tothe touch above, pubescent on the nerves beneath.4. Celtis crassifolia, Lamarck. North America. See p. 932.Leaves shortly acuminate, serrate in the upper half or two-thirds, scabrous tothe touch above, pubescent on the nerves beneath.II. Leaves lanceolate.* Leaves usually entire.5. Celtis mississippiensis, Bosc. North America. See p. 933.Leaves, entire in margin, rarely dentate at the apex, glabrous except foraxil tufts at the base beneath.** Leaves serrate.6. Celtis australis, Linnseus. Southern Europe, North Africa, Caucasus. See below.Leaves caudate - acuminate, covered beneath throughout with a softpubescence.7. Celtis caucasica, Willdenow. Caucasus, Persia, Afghanistan, Baluchistan,N. India. See p. 928.Leaves shortly acuminate, pubescent beneath only on the midrib and nerves.(A. H.)CELTIS AUSTRALIS, NETTLE TREECeltis australis, Linnseus, Sp. PI. 1 043 (1753); Loudon, Arb. et Frut. Brit. iii. 1414 (1838);Planchon in DC. Prodr. xvii. 169 (1873); Boissier, Fl. Orientalis, iv. 1156 (1879) ; Willkomm,Forstliche Flora, 5 45 (1887); Mathieu, Flore Forestiere, 2 93 (1897).A tree, usually attaining 50 to 70 feet in height, and 10 feet in girth, but in rarecases becoming as much as 20 feet in girth. Bark thin, greyish, smooth, somewhatresembling that of the beech, but on old trunks sometimes covered with wartyexcrescences. Young branchlets pubescent. Leaves (Plate 267, Fig. 5), about 4inches long by i|- inch broad, oval-lanceolate, unequal and cuneate at the base, contracted above into a very long caudate-acuminate apex, serrate except near the base ;upper surface dark green, scabrous, shortly-pubescent; lower surface greyish, coveredwith a soft tomentum ; petiole greyish-tomentose, about ^ inch long. Fruit globose,up to £ inch in diameter, at first whitish, then red, and finally dark-brown ; with ascanty sweetish flesh. Fruit pedicel, very slender, an inch or more in length. Theseedling 1 is similar to that of C. occidentalis ; but the cotyledons are wider, rhomboidalin shape, and with a shallower emargination; and the primary leaves are longer,narrower, and more acuminate. The seedling attains about 8 inches in height in thefirst year.1 Lubbock, Seedlings, ii. 495 (1892).I 1Celtis 927In winter, the twigs are slender, tomentose. Leaf-scars crescentic, 3-dotted, onprominent pulvini. Stipule - scars linear, one on each side of each leaf-scar.Terminal bud not formed and scar present at the apex of the twig, as inC. occidentalis. Lateral buds, appressed to the twig, compressed, ovoid, acute, with2 to 3 pairs of loose, tomentose, ciliate, imbricated scales.No varieties of C. australis have been described ; but in the eastern part of itsarea there are forms connecting it with C. caucasica, which is a very closely alliedspecies.Celtis australis is widely distributed throughout the Mediterranean region,and extends into Asia Minor and probably farther eastward, in the Caucasusand north Persia. In France, it is common in Provence and Languedoc,where it is often cultivated as coppice, and is met with as a rare tree as farnorth as Poitiers and Lyons. It is usually an inhabitant of the plains and low hills,but occasionally ascends to 3000 feet in the mountains. In the north of France,it is scarcely hardy, at least when young. It occurs in Switzerland in the cantonof Tessin. Farther east, its northern limit is the southern parts of Tyrol, Styria,and Hungary, whence it extends southward through the Balkan States toGreece and Crete. In Banat, Istria, and Dalmatia it often forms small woods, andascends to 1600 feet. It is in all these regions extensively planted, and has becomenaturalised in many districts. It also occurs in Spain and Portugal, Sardinia, Italy,Sicily, Morocco, Algeria, and Tunis; and is said to grow in the Madeira Islands.Celtis australis attains a great age, and trees of extraordinary size are recorded,there being one ] in the public square at Aix in Provence, estimated to be 500 yearsold, which is 19 feet in girth, and higher than any of the adjoining houses. Thereare fine specimens also in the Botanic Gardens at Montpellier, which are 10 to 12feet in girth. Willkomm saw very large trees in the Balearic Islands, and saysthat enormous trees are to be seen in Istria and Dalmatia, one at Pisino beingsupposed to be 1000 years old.Celtis australis produces suckers from the roots, and when cut gives goodcoppice shoots. In the south of France, coppice woods of this species are veryvaluable, as the shoots 2 attain about 4 inches in diameter in ten or twelve years, andare worth one to two francs each. The wood resembles much that of the ash, ofwhich it has all the good qualities, and is used in carriage-building, and for makingnumerous kinds of small articles, as tool-handles, hay-forks, trenails, tent-pegs, etc.Whip handles in France are almost universally made of this wood, and are called" perpignans," because the chief place of manufacture is at Perpignan. The foliage isgiven to cattle as fodder, the seed contains a sweet oil, and the bark and rootyield a yellow colouring matter.Celtis australis was introduced 8 in 1796, according to Loudon, who mentions a1 Mouillefert, TraiU des Arbres, ii. 1207 (1898). 2 Jolyet, Les Fortts, 226 (1901).3 In the London Catalogue of Trees, 1 8 (1730), three species of Celtis are mentioned as being in cultivation in England :_" I. Virginian Nettle Tree. Red fruit In several gardens near London there are large trees.II. Nettle Tree, with black fruit. European sort, is most rare in England.III. Celtis with large yellow fruit. Has been grown in Devonshire many years, where there are some large trees,which produce ripe fruit, from which many plants have been raised ; but we know not where it came fromoriginally."

92,8 The Trees of Great Britain and Irelandtree at Mitcham, which was 6 feet 8 inches in girth, and had a spread of 60 feet;and a tree at Kew, 40 feet in height, which no longer exists, there not being atpresent a single specimen there of this species. It appears to be very rare incultivation at the present time, the only trees which we have seen being oneat Liphook, and another at Hursley Park. The latter is a small unhealthy-lookingtree about 20 feet high, though of considerable age. There is also a small tree atTortworth.According to Bureau, 1 this species supports at Paris the severest winters withoutinjury; but according to Parde, 2 it bears with difficulty severe frosts in the northof France. The seedlings which Elwes raised at Colesborne were killed by 20 offrost, and though the tree may succeed in the warmest and driest parts of thesouth-east of England, it seems hardly worth planting elsewhere. (A. H.)CELTIS CAUCASICACeltis ca-ucasica, Willdenow, Sp. PI. iv. 994 (1805); Loudon, Art. et Frut. Brit. iii. 1415(1838);Boissier, Flora Orientalis,\v. 1 156 (1879).Celtis australis, Brandis, Forest Flora of N.W. India, 4 28 (1874), and Indian Trees, 5 95 (1906) (notLinnaeus); Hooker, Flora Brit. India, v. 482 (1888); Gamble, Indian Timbers, 6 29 (1902).A tree of moderate size, very similar to C. australis, of which it is possibly onlya geographical form. It differs in the following characters : Leaves ovate-lanceolate,broader in proportion to their length, and more rhomboidal, with a shorter and noncaudateacuminate apex ; upper surface glabrescent, scarcely scabrous ; lower surfacewith slight pubescence, confined to the nerves and midrib. Drupes yellow. 3This species, which is connected with the European species by var. cuspidata?with long - acuminate leaves, is widely spread through the Caucasus, Persia,Afghanistan, Baluchistan, and northern India. In the Caucasus, it is associatedwith C. australis; but farther east the latter species is scarcely met with. InAfghanistan, according to Aitchison,6 it is usually a planted tree near shrinesand in graveyards; but it is quite wild along the Darban and Shendtoi rivers;and in Baluchistan, its leaves, according to Lace, 6 are often used, as they are inIndia, for feeding sheep and goats, the trees being pollarded for this purpose. Itoccurs in India in the north-west Himalaya, as far east as Nepaul ascending to8000 feet, where it is a common tree, wild in the forests, and around villages.According to Webber, 7 in Gorakhpur, it reverses the season of casting its leaves,which wither and fall off in the hot weather, and it flowers in the early months of thecold season. The wood is tough and strong, and is used for oars, tool-handles,sticks, and other purposes requiring toughness and elasticity.This species, though mentioned by Loudon, was not in cultivation in Englandin his day; and Schneider 8 doubts if it has yet been introduced on the Continent.1 Nowv. Arch. Mus. Hist. Nat. vi. 181 (1894).3 Brandis mentions a variety with purplish-black fruit.6 Journ. Linn. Soc. (JBot.) xviii. 93 (1880).7 Forests of Upper India, 232(1902).* Art. Nat. Des Barres, 242 (1906).1 Planchon, in DC. Prod, xvii., 170 (1873).6 Ibid, xxviii. 305 (1891).8 Laubhokliunde, 2 31 (1904).Celtis 929There is, however, a tree of this species in the Kew Collection, which is marked" Aitchison " on the label, and was probably raised from seeds sent by Aitchison fromAfghanistan, about the year 1881. (A. H.)CELTIS GLABRATACeltis glabrata, Steven, ex Planchon, in Ann. Sc. Nat. ser. 3, x. 285 (1848).Celtis Tournefortii, Lamarck, var. glabrata, Boissier, Fl. Orient, iv. 1157 (1879).A shrub or small tree. Young branchlets with a minute scattered pubescence.Leaves (Plate 267, Fig. 10), about 2 inches long, i^ inch broad, quite glabrous,ovate, unequal and rounded or broadly cuneate at the base, acute or very shortlyacuminate at the apex, coarsely serrate except near the base ; upper surfacebluish green, roughened with minute papillae; lower surface lighter green in colour ;punctate with numerous translucent minute dots, when viewed with a lens ; petioleglabrous, 3- inch. Fruit pedicels about an inch. Drupes globose, reddish brown.This species occurs in Asia Minor, in Lycia and Cilicia, and in the Caucasus.Schneider doubts if it has been introduced into cultivation ; but there is at Kew asmall tree, undoubtedly of this species. C. Tournefortii, Lamarck, a closely alliedspecies, occurring in Sicily, Greece, and Asia Minor, is mentioned by Loudon, ashaving been introduced into England in 1739, and cultivated in 1838 in the LondonHorticultural Society's Garden; but I have seen no specimens in this country.(A. H.)CELTIS DAVIDIANACeltis Davidiana, Carriere, Rev. Hort. 1 868, p. 300.Celtis Bmigeana, Blume (in part); Hemsley, Journ. Linn. Soc. (Hot.) xxvi. 449 (1894).Celtis sinensis, Persoon (in part); Maximowicz, Mil. Bid. ix. 27 (1872); Bretschneider, BotaniconSinicum,'\. 1 17 (1882).A small tree. Young branchlets slightly pubescent. Leaves (Plate 267, Fig.u), about 2|- inches long, \\ inch broad, ovate or ovate-lanceolate, base rounded,contracted above into a short acuminate apex, rarely entire, usually slightlytoothed in the upper third; glabrous and shining on both surfaces, dark greenabove, light green beneath, punctate when viewed with a lens; petiole -J inch,pubescent. Fruit-pedicels, slender, f inch long. Drupes small, ovoid, black incolour.This species occurs in north China, in the hills around Peking, and in themountains of the province of Shingking; and was found by me growing as a smalltree, about 20 feet in height, in the mountains of Hupeh. It has been confusedwith two other Chinese species, C. Bungeana? and C. sinensis? which do not appear tobe in cultivation in Europe. It is readily distinguishable from all the other cultivatedspecies by the very shining glabrous leaves; and is a very distinct and handsometree.1 Blame, Mus. Bot. Ltidg. Bat. ii. 71 (1852). 2 Persoon, Syn. PI. \. 292 (1805).IV o F

93 The Trees of Great Britain and IrelandIt was introduced into Kew Gardens, where there is a small tree about 15 feethigh, by seeds sent from Peking by Bretschneider in 1882. It had, however, beenpreviously introduced into France by Pere David, who sent seeds to Carriere in1868, from which a tree was raised in the Jardin des Plantes. This tree, accordingto Franchet l had become with age identical in character with C. Bungeana ; butthis is incorrect. It fruited for the first time 2 at Paris in 1894. Schneider 3mentions trees of this species in the Botanic Gardens at Strassburg and Darmstadt.(A. H.)CELTIS OCCIDENTALIS, HACKBERRYCettis occidentalis, Linnaeus, Sf. PI. 1 044 (1753); Michaux, Hist. Arb. Am. iii. 225, t. 8(1813);Loudon, Arb. et Frut. Brit. iii. 1417 (1838) ; Sargent, Silva N. Amcr. vii. 67 (in part), t. 317(1895), and Trees N. Amer. 2 99 (1905) (in part).A tree, attaining in America, 100 feet in height and 9 feet in girth. Bark grey,broken on the surface into appressed scales, and often roughened on old trees withthick discontinuous corky ridges. Young branchlets glabrous or pubescent. Leaves(Plate 267, Fig. 4), uniform in size, about 2|- inches long and \\ inch wide, ovate,unequal and rounded or shortly cuneate at the base, with a long caudate-acuminateusually non-serrated apex; serrate in the upper half or two-thirds; upper surfacesmooth to the touch; lower surface pubescent on the nerves ; petiole \ inch or more,glabrous or pubescent. Fruiting pedicels short, about f inch. Drupe, purplish-blackor orange when ripe, globose or ovoid, about f inch in diameter.In winter the twigs show the following characters : Branchlets slender, zigzag,reddish-brown, shining, glabrous. Leaf-scars oblique on prominent pulvini, threedotted.Stipule scars minute, linear, one on each side of each leaf-scar. Terminalbud not formed, the end of the branchlet falling off in summer, and leaving aminute orbicular scar at the apex of the twig. Buds 4 all axillary, uniform in size,about fs inch long, alternate, distichous, appressed to the twig, ovoid, acute, compressed, covered by three pairs of pubescent, ciliate, imbricated scales.Seedling? Primary root long, tapering, flexuose, with numerous lateral fibres.Caulicle erect, pubescent, about i^ inch long. Cotyledons oblong, cuneate andthree-nerved at the base, emarginate at the apex, green above, pale beneath, aboutJ inch long. Stem hispid. First pair of leaves opposite, ovate, acuminate, serrate,three-nerved, covered in the young stage with clear dot-like glands. Succeedingleaves similar, but alternate.Scarcely any varieties are known, unless C. crassifolia be considered ageographical form of this species. C. pumila, Pursh,6 a low shrub, of xerophytic1 Planta: Davidiawz, \. 269 (1884). 2 Rev. Hort. 1894, p. 97. 3 Laubhohkunde, 228 (1904).4 Sometimes in this species, the axil of the leaf produces three buds side by side. The middle bud sends out a shoot inthe following year, whilst the lateral ones are left as a reserve. If the shoot happens to die in the year after, one of the twoaccessory buds develops. Cf. Kerner, Nat. Hist. Plants, Eng. transl. ii. 32 (1898).6 Cf. Lubbock, Seedlings, ii. 493, fig. 646 (1892).6 Ft. Amer. Sept. i. 200 (1814); Hill, Bull. Torr. Bot. Clut, xxvii. 496 (1900).Celtis 93 1habit, usually growing on sand dunes, rocky places, and dry hills and mountains, andwidely distributed in the United States from Delaware and Pennsylvania to Utahand Colorado, is considered by some to be a variety of this species, but is probablydistinct.This species is widely distributed, occurring in the north from Quebec toManitoba, and extending southward to Louisiana, Missouri, Kansas, and NorthCarolina. Its exact distribution is not clearly known, as it has been confused withother species ; but it appears to be commoner in the north and east, while C. crassifolia is most prevalent in the Mississippi valley and west of the Alleghanies. It isreplaced by C. mississippiensis in the extreme south, and by C. reticulata in the dryregions of the far west.In New England it is a low round-headed tree, and is well depicted by Sargentin Garden and Forest, iii. 39, fig. 43 (1890), which represents a tree growing closeto the seashore in Massachusetts. It is not common east of the Hudson, 1 but on thebanks of this river grows with a slender trunk and long graceful pendulous branches.Amongst North American species none, perhaps, retains its foliage green and freshso late in the season. (A. H.)The American Nettle Tree or Hackberry was introduced into England in 1656by Tradescant, and the first description of the species, made from a tree cultivatedin England, was published in Ray's Historia Plantarum, ii. 1917. It seems to bethe only species of Celtis which bears our climate well enough to be worth planting,but is so rare in cultivation that very few people know it, and it is rarely found innurseries.I have found it easy to raise from seed, and though the seedlings grow slowlyand are somewhat susceptible to frost when young, it will, as it grows older, endurea greater degree of cold in winter than C. australis. The largest tree mentioned byLoudon was one at Syon, which, in 1838, was 54 feet high and 7 feet in girth.This cannot be the same as one which now grows there, which, according to thegarden catalogue, was 50 feet by 3 feet 3 inches in 1849, and when I measured it in1903, was 60 feet by 4 feet 4 inches. There is a fine tree in Kew Gardens, which,owing to its being crowded on one side by an evergreen oak, leans considerably, andis about 45 to 50 feet by 6 feet 4 inches. This is probably one of the original treesof the Kew Arboretum, though not mentioned by Loudon, unless, as is possible, hemistook it for C. australis.The finest and best-shaped tree, however, that we know of, is one at West DeanPark (Plate 251), which is probably the one mentioned by Loudon as then fourteenyears planted and 19 feet high. Now it is 50 feet by 5^ feet, and when I saw it inJuly 1906 was bearing fruit abundantly. This tree evidently lives to a considerable age, as I saw one in the Botanic Gardens at Padua which was planted in 1760,and measures no less than 32 metres high by 2 metres in girth, with a fine clean bole.Henry saw a good specimen in the Botanic Garden at Copenhagen.(H. J. E.)1 Garden and Forest, i. 465 (1888).

932- The Trees of Great Britain and IrelandCELTIS CRASSIFOLIA, HACKBERRYCeltis crassifolia, Lamarck, Encycl. iv. 138 (1797); Michaux, Hist. Arb. Amer. iii. 228, t. 9 (1813);Louden, Arb. et Frut. Brit. iii. 1418 (1838); Britton, Man. Fl. North. States and Canada,339 (i9°0-Celtis cordata, Persoon, Sp. PI. i. 292 (1805).Celtis Audibertiana, Spach, Ann. Sc. Nat. ser. 2, xvi. 41 (1841).Celtis occidentalis, Sargent, Silva N. Amer. vii. 67 (1895), and Trees N. Amer. 299 (1905). (Inpart.)A tree, attaining rarely as much as 130 feet in height, and 10 feet in girth.Bark as in C. occidentalis. Young branchlets pubescent. Leaves usually aboutz\ inches long by \\ inch broad, but on isolated branches, commonly found in theinner part of the crown, often 6 inches long and 3 inches broad; ovate, unequal,and shortly cuneate at the base, shortly acuminate at the apex, serrate, as a rule, onlyin the upper half; upper surface scabrous to the touch ; lower surface pubescent onthe nerves ; petiole, \ inch or more, pubescent. Fruiting pedicels long, f inchor more. Drupe purple, red or black when ripe, globose or ovoid, about § inch indiameter.This species, which is not distinguished by Sargent, even as a variety, fromC. occidentalis, is remarkably distinct in foliage, and appears to be a more uprightand faster-growing tree in cultivation than that species. Michaux states that it isone of the finest species of the genus, remarkable for its great height and straighttrunk, and that it is common in the states west of the Alleghanies, especially inOhio and Kentucky, where, however, its timber was little esteemed on account of itsweakness and liability to speedy decay on exposure to the weather. Its distributionhas been confused with that of C. occidentalis; but, according to Britton, it occursfrom New York to South Carolina, Ohio, Illinois, Missouri, and Tennessee.(A. H.)Ridgway x speaks of this species of Hackberry under the name ofC. occidentalis,as "a very tall and beautiful tree in rich bottoms, growing frequently 120 to 130 feethigh and 3 feet in diameter, with a tall straight trunk of 60 to 70, or even 80 feet tothe first limb. When growing to its full perfection in a dense forest, there is anindividuality about the aspect of this tree which it is difficult to describe, owing tothe extreme slenderness and great length of the trunk, which not unfrequentlycomprises three-fourths of the total height of the tree; and the smooth grey barkconspicuously clouded on the north side with blackish moss or lichen for its entirelength. This striking appearance is sometimes increased by vines of the Virginiacreeper ascending to the topmost branches, which are wreathed and matted with itsfoliage. One tree was seen whose silvery shaft gleamed among the surroundingtree tops, in a wood where the summit level was considerably more than 100 feetaloft, and though only 10 feet in circumference must have been upwards of 90 feet1 Proc. U.S. Nat. Museum, 1 882, p. 72.Celtis 933to the first limb, which grew not more than 25 feet from the extreme summit ofthe tree."Though I was not fortunate enough to find any such trees standing, when Ivisited the remains of this forest in 1904, yet I saw enough to make me wish that anarea of this unique forest could be preserved to show what the virgin forests of theWabash valley were once like; for there is no other part of the temperate worldwhere so many species of hardwood trees grow to such a size as they formerly didhere.This species appears to have been introduced about the beginning of the nineteenth century; several trees, 10 to 15 feet in height, being mentioned by Loudon.At Kew it appears to be straighter and more vigorous in growth thanC. occidentalis ; and all the specimens have a few branches mainly in the upper andinner parts of the tree, which bear very large leaves. One of the trees, growing onthe walk behind the Aroid House, is 38 feet high by 3^ feet in girth.(H. J. E.)CELTIS MISSISSIPPIENSISCeltis mississippiensis, Bosc, Diet. Agric. x. 41 (1810); Sargent, Silva N. Amer. vii 71, t. 318(1895), and Trees N. Amer. 300 (1905).Celtis Icevigata, Willdenow, Berlin Baumz. 8 1 (1811); Loudon, Arb. et Frut. Brit. iii. 1420 (1838).Celtis occidentalis, Linnaeus, var. integrifolia, Nuttall, Gen. i. 202 (1818).Celtis occidentalis, Sargent, Forest Trees N. Amer., i o//z Census U.S. ix. 125 (1884) (in part); andGarden and Forest, iii. 39 (in part), ff. 9, 10, u (1890).A tree, attaining in America, 80 feet high and 9 feet in girth. Bark bluishgreen,and covered with prominent excrescences. Young branchlets glabrous.Leaves (Plate 267, Fig. 9), up to 3 inches long and \\ inch wide, ovate-lanceolate orlanceolate, unequal and rounded or broadly cuneate at the base, long-acuminate atthe apex ; margin usually entire, occasionally irregularly serrate towards the apex ;light green and glabrous, except for slight axil tufts at the base beneath ; petiole,about \ inch, glabrous. Fruiting-pedicels, about f inch. Drupes, ovoid, ^ to ^ inch,bright orange red, with thin dry flesh and a smooth light brown stone.This species is distributed from southern Indiana and Illinois, throughKentucky, Tennessee, and Alabama to Florida, and through Missouri, Arkansas, andTexas to Nuevo Leon. It is also a native of the Bermudas. It is very abundantand of its largest size in the basin of the lower Ohio River, a tree measured bySchneck in Richland County, Illinois, being 95 feet high and 5^ feet in girth. Hereit is often associated with C. crassifolia, from which it may be distinguished* by itsusually smaller size, shorter trunk, entire leaves, and bright orange-red fruit. It isthe most common species in Kentucky and Tennessee; but is rare in the GulfStates. Though apparently found in Texas and Nuevo Leon, it is replaced to the1 Elwes noticed that the wrinkled bark of this species easily distinguished it in the forest from C. crassifolia.

934 The Trees of Great Britain and Irelandwestward by Celtis reticulata, Torrey, 1 which occurs in Texas, New Mexico,Arizona, Utah, Nevada, and Lower California, and has not yet apparently beenintroduced.C. mississippiensis was probably introduced, by the elder Michaux, into France,where the species was first described in 1810 from a tree cultivated at Paris. It wasnot introduced into England in Loudon's time. The only cultivated specimen knownto us is a small tree in Kew Gardens. (A. H.)> Ann. Lye. N. York, ii. 247 (1828). C. occidentalis, var. reticulata, Sargent, Forest TreesN. Amer., loth Census, U.S.ix. 1 26 (1884); Garden and Forest, iii. 40 f. 12 (1890). C. mississippiensis, var. reticulata, Sargent, Sifva N. Amer. vii.72, t. 319 (1895), and Trees N. Amer. 301 (1905).ALNUSAlnus, Linnaeus, Gen. PI. 2 85 (1737); Bentham et Hooker, Gen. PI. iii. 404 (1880); Winkler, inEngler, Pflanzenreich, iv. 61, Betulaceee, 1 01 (1904).Betula, Linnreus, Gen. PL 485 (1764) (in part).Alnaster and Clethropsis, Spach, Ann. Sc. Nat. sdr. 2, xv. 200, 201 (1841).Semidopsis, Zumaglini, Fl. Pedem. i. 249 (1849).Alnobetula, Schur, Verhand. Siebenb. Ver. Naturw. iv. 68 (1858).DECIDUOUS trees or shrubs, belonging to the order Betulaceae. Leaves alternate,simple, stalked, usually serrate or dentate, rarely entire, penninerved. Stipules 1enclosing the leaf in the bud, caducous or deciduous. Flowers opening eitherin early spring before or with the unfolding of the leaves, or in two species inautumn, monoecious, unisexual, without petals, in few-flowered cymes in theaxils of short - stalked peltate scales of pedunculate catkins. Staminate catkins,at first naked and erect, afterwards pendulous, in the axils of the last leaves or ofleafy bracts; scales three-flowered; bracteoles, three to five, adnate to the baseof the scale; calyx, four-partite; stamens four, filaments short, undivided; anthersdorsi-fixed, not pilose at the apex. Pistillate catkins erect, solitary or racemose, inthe axils of the leaves and produced in autumn, or terminal on a short leafy branchand produced in spring; scales two-flowered; bracteoles, two to four, adnate to thebase of the scales; calyx absent; ovary two-celled; styles two, stigmatose at theapex; ovule, solitary in each cell, suspended. Fruit, a strobile or cone, formed bythe scales of the pistillate flower becoming, when ripe, thick, woody, obovate, threetofive-lobed or truncate at the thickened apex. Cones persistent on the branchafter the opening of the closely imbricated scales and the escape of the nutlets.Nutlet, compressed, minute, bearing at the apex the remains of the styles, marked atthe base by a scar, with or without lateral wings. Seed, solitary by abortion, fillingup the cavity of the nutlet.About twenty-five species are distinguished, inhabiting Europe, Algeria, extratropicalAsia, North America, Central America, and the Andes of South Americafrom Colombia to Peru.The following key includes all the species in cultivation, with the exception ofAlnus serrulata, Willdenow, a North American shrub.I. Buds sessile, -with several (two to six) outer scales, which are unequal in length.i. Alnusfirma, Siebold et Zuccarini. Japan. See p. 952.1 Cf. Lubbock, "On Stipules of the Alder," in/earn. Linn. Soc. (Bot.), xxx. 527 (1895).935

936 The Trees of Great Britain and IrelandLeaves ovate-lanceolate, green beneath, with fifteen or more pairs of lateralnerves.2. Ahms viridis, De Candolle.1 Mountains of Central Europe, plains of NorthernRussia, Siberia, Labrador, and Greenland.Leaves broadly oval, green beneath, with eight to ten pairs of lateral nerves.This species is a shrub, rarely attaining more than 10 feet in height, andwill not be further dealt with by us.II. Buds stalked, with two outer scales, almost equal in length.A. Leaves -white or grey beneath.3. Abms incana, Mcench. 2 Europe, Caucasus, North America. See p. 945.Branchlets pubescent, buds rounded at the apex. Leaves grey beneath, withthe lateral nerves running to the tips of serrated lobules; margin notrevolute.4. Alnus rubra, Bongard. Western North America. See p. 956.Branchlets glabrous, buds beaked at the apex. Leaves whitish beneath, withnerves as in A. incana ; margin revolute.B. Leaves green beneath.* Leaves simply serrate, not lobulate, rounded or acute at the apex.5. Alnus elliptica, Requien. A hybrid between A . cordata and A . glutinosa,occurring wild in Corsica. See full description under A . cordata, p. 949.Branchlets glabrous and covered with wax glands. Leaves elliptic, roundedat both base and apex, with prominent brown axil-tufts beneath.6. Almis rkombifolia, N uttall. Western North America. See p. 958.Branchlets pubescent. Leaves ovate or oval, rounded at the base, acute orrounded at the apex, pubescent throughout beneath.** Leaves simply serrate, not lobulate, conspicuously acuminate at the apex.7. Alnus cordata, Desfontaines. Corsica, Southern Italy. See p. 949.Branchlets glabrous, covered with wax glands. Leaves broadly ovate, cordateat the base, cuspidate-acuminate at the apex, with conspicuous axil-tuftsbeneath.8. A lnus japonica, Siebold et Zuccarini. Japan, Manchuria, Korea, Formosa.See p. 953.Branchlets glabrous. Leaves lanceolate, cuneate at the base, long-acuminateat the apex, with minute axil-tufts beneath.9. Alnus maritima, N uttall. Delaware, Maryland, Indian Territory. See p. 955.Branchlets glabrous. Leaves ovate or obovate, cuneate at the base, shortlyacuminate at the apex, glabrous beneath.10. Alnus nitida, Endlicher. North-west Himalaya. See p. 954.Branchlets densely and minutely pubescent. Leaves ovate-elliptical, cuneate1 Fl. Franc, iii. 304 (1805). Almis Aliwbettila, Hartig, Naturges. Forst. Kulturpfl. 372 (1851). Betula Alnobelula,Ehrhart, Beit. ii. 72 (1788).2 Alnus tinetoria, Sargent, a closely allied species, occurring in Japan and Manchuria, is distinguished by its large,broadly ovate leaves, acuminate at the apex. A . incana has small, narrowly ovate leaves, acute at the apex. Cf. p. 946.Alnus 937or rounded at the base, short- or long-acuminate, with slight axil-tuftsbeneath.*** Leaves with serrate lobules.11. Alnus glutinosa, Gsertner. Europe, Siberia, Western Asia, North Africa,See below.Branchlets usually glabrous. Leaves obovate, cuneate at the base ; obtuse,truncate or rounded at the apex ; with prominent axil-tufts beneath.12. Alnus glutinosa, Gsertner, var. barbata, Ledebour (Alnusbarbata, C . A. Meyer).Caucasus. See p. 938.Branchlets pubescent. Leaves elliptical, rounded at the base and apex,covered on both surfaces with pubescence, densest on the midrib andnerves beneath.13. Alnus tenuifolia, N uttall. Western North America. Seep. 957.Branchlets glabrous. Leaves ovate, broad and rounded at the base, acuteor shortly acuminate at the apex, pubescent on the midrib beneath withinconspicuous axil-tufts.14. Alnus subcordata, C. A. Meyer. Caucasus, North Persia. Seep. 951.Branchlets pubescent. Leaves ovate-oblong, unequal and rounded or subcordateat the base, cuspidate-acuminate at the apex, pubescent on themidrib and nerves beneath.ALNUS GLUTINOSA, COMMON ALDERAlnus glutinosa, Gaertner, De Fruct. ii. 54 (1791); Loudon, Arb. et Frut. Brit. iii. 1678 (1838);Willkomm, Forstliche Flora, 3 39 (1887); Mathieu, Flore\ Forestttre, 4 21 (1897); Winkler,Betulacece, 1 15 (1904).Alnus nigra, Gilibert, Exerc. ii. 401 (1792).Alnus communis, Desfontaines, Tabl. Hort. Paris, 2 13 (1804).Alnus vulgaris, Persoon, Syn. ii. 550 (1807).Betula Alnus glutinosa, Linnaeus, Sp. PI. 9 83 (1753).Betula Alnus, Scopoli, FL Cam. ii. 233 (1772).Betula glutinosa, Lamarck, Diet. i. 454 (1783).Betula palustris, Salisbury, Prod. 3 95 (1796).A tree, occasionally attaining 100 feet in height and 12 feet or more in girth.Bark of young trees smooth and greenish ; after twenty years old becoming brownishblackand divided on the surface into broad flattened plates. Young branchlets,three-angled at the tip, usually glabrous, occasionally pubescent, covered with glands,which secrete a waxy resin, often seen on the dried twigs as a bluish bloom. Leaves(Plate 268, Fig. 6) averaging 3^ inches long and 3 inches broad, variable in shape,but nearly always broadest above the middle, obovate, sub-orbicular or elliptical;cuneate at the base; obtuse, truncate, or retuse at the apex; margin entire in thebasal third, elsewhere lobulate, each lobule serrate or dentate; upper surface darkgreen, shining, glabrous; lower surface light-green, pubescent along the midrib andIV 2 G

938 The Trees of Great Britain and Irelandnerves, with conspicuous tufts of rusty-brown hairs in the axils ; petioles, glabrousor pubescent, i^ to i inch long; stipules conspicuous, deciduous, ovate to lanceolate,obtuse, fringed with glandular hairs. The leaves turn blackish in autumn.Flowers appearing very early, before the leaves, in February or March.Catkins, three to six in a raceme, at the tip of a branchlet. Staminate catkins, 2 to 4inches long, at first erect and rigid, afterwards lax and pendent; stamens, 1 four,yellow, opposite the segments of the four-lobed calyx. Pistillate catkins, alwayserect, at first about £ inch long, smooth, with reddish-brown stigmas ; afterwards^ inch long, ovoid, cone-like, the scales ending in purple shield-like expansions, eachwith a central brown point. Cones at first green, ultimately black, persistent on thetree after the escape of the nutlets. Nutlets obovate, blunt-angled, wingless orwith a very narrow coriaceous wing.The nutlets 2 are gradually shaken out of the cones by the wind during autumnand winter. Their walls are provided with small air-tight cavities, which enable themto float in water, and secrete an oil, which protects them from being wetted.Usually falling into streams and ditches, they float undamaged and unchanged duringwinter, and germinate in the water in early spring. The young seedlings, drifting tothe bank, establish themselves where they happen to be stranded in a suitableplace.In winter, the twigs are glabrous and usually covered with a thin waxy secretion.Leaf-scars pentagonal or rhomboid, parallel to the twig on a projecting cushion, fivedotted,the lowermost three dots coalesced together. Stipule-scars linear, one oneach side of a leaf-scar. Terminal bud present, similar to the lateral buds; allconspicuously stalked, ovoid, obtuse, with two external scales, viscid-glandular, andoften covered with a purplish bloom. Pith triangular in section.The common alder coppices freely from the stool; but rarely if ever producesroot-suckers.VARIETIESThe common alder, distributed over a wide area, shows considerable variationin the wild state, and several varieties have been described.i. Var. barbata, Ledebour, Fl. Rossica, Hi. 657 (1851); Winkler, Betulacea, 1 18(1904). Alnus barbata, C. A. Meyer, Verz. Pfl. Kauk. 43 (1831).This variety is remarkably distinct in the foliage, but has the flowers and fruitof typical A . glutinosa. Young branchlets pubescent. Leaves, about 3 inches longand 2 inches wide, rounded at the base, rounded or occasionally acute at the apex,margin with serrate lobules, ciliate; nerves eight or nine pairs, running parallel andcurved to the margin; upper surface dark green, shining, minutely pubescent; lowersurface pubescent, the pubescence densest along the midrib and nerves ; petiole £inch, pubescent. Buds pubescent.1 Kerner, in Nat. Hist. Plants, Eng. trans. ii. pp. 119, 133, 135, gives an elaborate account of the way in whichpollination is effected by the wind, and of the devices for the protection of the pollen in rainy weather.* Cf. Miall's account of the fruit of the alder in Routid The Year, p. 279.Alnus 939This variety occurs in the Caucasus, and is very similar to var. denticulata?occurring in the same region and in north Persia, which is less pubescent. Var.barbata is in cultivation at Kew.2. Var. quercifolia, Willdenow, Berlin Baumz. 44 (1796). Oak-leaved alder.Leaves obovate, lobed like the common oak. This variety has been found wild inSweden.3. Var. sorbifolia, Dippel, Laubholzkunde, i i. 161 (1892). Service-leaved alder.Leaves oval, lobed like those of Pyrus intermedia. This variety has been foundwild in Finland.4. Var. laciniata, Willdenow, he. cit. Cut-leaved alder. Leaves (Plate 268,Fig. 7), divided half-way to the midrib into three to six pairs of non-serrated triangularsegments ; petiole slender, about an inch long.The cut-leaved alder, according to Duhamel, occurs wild in the north ofFrance, particularly in Normandy, and in the woods of Montmorency near Paris.Thouin states, according to Loudon, that it was first found by Trochereau dela Berliere, and planted by him in his garden near St. Germain, where thestool remained in 1838, from which all the nurseries of Paris were supplied withplants. The largest trees we have seen of this variety are described on p. 942.5. Var. imperialis, Petzold and Kirchner, A rb. Muse. 5 99 (1864). Alnusimperialis, Desfosse-Thuillier, Illust. Hort. v i. 97, fig. (1859).Leaves (Plate 268, Fig. 8) oval, divided more than half-way to the midrib, intosix or seven pairs of long narrow lanceolate non-serrated curved segments. Thisvariety, so far as we know, does not attain to as large a size as the ordinary form ofthe cut-leaved alder. A specimen at Ponfield, Hertford, is 25 feet high by i foot 8inches in girth.6. Var. incisa, Willdenow, Sp. PI. iv. 335 (1805) (var. oxyacanthczfolia,Loddiges, Catalogue, 1 836). Thorn-leaved alder. Leaves (Plate 268, Fig. 10)small, deeply incised, resembling those of the common hawthorn. A fine specimen,44 feet high by 2 feet 8 inches in girth, is growing in the arboretum at Barton, nearBury St. Edmunds.7. Var. rubrinervia, Dippel, loc. cit. A tree, pyramidal in habit, with largeand shining leaves, furnished with red petioles and nerves, vigorous in growth andhandsome in appearance. In cultivation at Aldenham.8. Var. pyramidalis, Dippel, loc. cit. Branches erect, leaves as in the type.9. Var. aurea, Verschaffelt, ex Dippel, loc. cit. Lemaire, Illust. Hortic. 1 866,t. 490. Leaves yellow. Found as a seedling in Vervaene's nursery at Ledebergles-Gand.In cultivation at Aldenham.10. Var. maculata, Winkler, loc. cit. Leaves variegated with yellow. There isa small specimen at Aldenham, which is slow in growth.Hybrids 2 between Alnus glutinosa and Alnus incana are common in the wildstate, where the two species are growing together, and have been observed in1 Ledebour, loc. cit. Alnus dcnticulata, C. A. Meyer, loc. cit.2 A. glutinosa X incana; A. spuria, Callier. Schneider, Laubholzkunde, 1 30 (1904), distinguishes three forms ofthis hybrid.

940 The Trees of Great Britain and IrelandSweden, Norway, Russia, Denmark, Bosnia, and Montenegro. These hybrids areintermediate in the characters of the leaves and fruit; and so far as we know, are notin cultivation in England.A hybrid l between A . glutinosa and A . serrulata has also arisen in cultivationin Silesia and Brandenburg.DISTRIBUTIONThe common alder is distributed through nearly the whole of Europe, Siberia,Western Asia, and North Africa. In Europe, its northerly limit extends from lat.63 52' in southern Norway at Anderoen, to lat. 63 20' on the west side of the Gulfof Bothnia in Sweden, reaching Uleaborg in Finland in lat. 65 , where, however, itonly exists as a shrub, and is continued through the interior of Finland and Russiaalong the parallel of 62 . In Siberia, its distribution is not accurately known ; but itoccurs in the Ural and Altai mountains, and in the district around Lake Baikal. Itssouthern limit, commencing in the province of Talysch in the Caucasus, between 39and 40 lat., extends through Asia Minor and Greece to Sicily, where it reaches atCatania, lat. 37 25', its most southerly point in Europe. It occurs in Spain andPortugal, as far south as the Sierra Morena, about lat. 38 . It is also distributedthrough the mountains of Algeria and Morocco. A variety, 2 recorded for Japan, isprobably a distinct species.According to Sir Herbert Maxwell a the Anglo-Saxon name for alder was air,in Norse olr (now, according to Schiibeler, aar, older, and or); and the Gaelic name/earn, the names surviving in place names such as Allerton, Allerbeck, Ellerslie,Balfern, Farnie, Glenfarne.It is generally distributed throughout the British Isles, growing usually onriver banks, along the sides of lakes, and in wet or marshy places; and ascends inthe Grampians to 1600 feet.It is common throughout France, in similar situations, and ascends to 5700feet in the Pyrenees. Although most usual on siliceous soils, it grows on chalk inChampagne; and in Germany, has been shown to be indifferent to the mineralconstituents of the soil, provided a sufficient quantity of moisture is present. It ismet with as pure woods, on peat-bogs and marshy places, in north Germany, in theBaltic provinces and Lithuania in Russia, and also in Hungary ; but more usually ismixed with birch and aspen, and more rarely grows in company with other hardwoods. It ascends in the mountains of Norway to 1100 feet, in the Central Alpsin Switzerland to 3200 feet, and in the Carpathians to 3800 feet. Thealder is susceptible to late frosts and is injured by cold dry winds, and on thisaccount thrives best in the colder parts of Europe on slopes with a westerlyaspect.It is naturalised in various localities in the eastern United States, particularly insouthern New York and in New Jersey. It succeeds well in North America, when1 A. glutinosa X serrulata ; A. silesiaca, Fiek. Cf. Schneider, loc. cit.2 Var. japonica, Matsumura,y-«. Coll. Set. Tokyo, xvi. 2, p. 9 (1902).3 Green's Encyclopedia of Agriculture, i . 62.Alnus 94planted in wet situations; but is not, as a rule, a long-lived tree, and never growsto be so large as it does in England. 1 (A. H.)PROPAGATION AND CULTUREThough old writers tell us that the alder was often planted by inserting longcuttings, or by burying pieces of the root in the soil; and though layering is the modeusually adopted by nurserymen for propagating the varieties ; yet, as a general rule,it is best raised from seed. The cones ripen in autumn and are freely producedalmost every year so far as I have observed. As soon as they begin to open, theyshould be gathered when quite dry; and though it is said that the seeds will keep fortwo years or more in the cone, yet, as a rule, they should be sown in autumn or inearly spring and lightly covered with earth. Though I have not tried it myself, Ibelieve that the germination is more rapid and regular if the seed is soaked in waterbefore sowing, but seedlings can be procured so cheaply from nurserymen that Ihave always bought them at one or two years old. They are easy to transplant andgrow fast if the soil is moist, being fit to plant out at three or at most four years old ;and I have had as good or better success by planting them in spring as in autumn.The alder bears coppicing well, if not cut too near the ground, but the stools havea tendency to decay in the centre and to spread outwardly. True suckers are notproduced, though the roots when exposed by running water will throw up shoots.The usual age for coppicing is fifteen to twenty years, and I am informed by SirHugh Beevor that he obtained a yield of 1700 poles per acre, which at seventeenyears' growth from the stool averaged 20 feet long with a girth of 7 to 11 inches,giving a yield of about 1100 cubic feet per acre. If the trees are allowed to standfor timber they should be cut at fifty to seventy years, when they may average 50 to70 feet high by 4 to 5 feet in girth. The only lot of alders I ever sold standing, 300in number, realised ^"100, being at the rate of 4d. or 5d. per foot. Sir HerbertMaxwell states, that as long as clogs remain in common use, there will be littledifficulty in realising ^40 per acre for mature alder coppice, and this on land so wet asto be worthless for any other purpose. Except in localities where a good and regularmarket is assured, I should not recommend the planting of alder except in places toocold, wet, and marshy for willow or poplar to thrive; but Selby, 2 whose opinion ofthe tree as an ornamental one was better than my own, states as the result of hisown experience, that the nature of the roots of the alder causes the tree to attractand retain the moisture in the soil, to such an extent that it will convert into amorass, land which, if drained and planted with other trees, might be rendered dryand productive. He adds that from experiments he has made he is " fully convincedthat a plantation of alders would soon render the ground (even if previously oftolerably sound and dry quality) soft and spongy, and in time convert it into adecided bog." I cannot learn that this observation has been confirmed by others,and am inclined to doubt its being of general application.1 Hough, Trees N. States and Canada, 1 31 (1907). * British Forest Trees, p. 218 (1842).

P42- The Trees of Great Britain and IrelandThe common alder is rarely planted as an ornamental tree, and only on wetsituations or on the banks of ponds and streams is it able to attain its full dimensions.But if desired for ornament in such a situation, I should recommend the cut-leavedalder in preference to the common one; and for drier ground either the grey orthe Italian alder.REMARKABLE TREESAmong the finest trees that we have seen or heard of, those at Pain's Hill take ahigh place. One of these on an island is probably over 90 feet, but we could notmeasure it. Another growing by the lake has seventeen stems about 75 feet inheight, growing from a stool which is 19 feet in girth. I estimated the contentsof these poles at about 200 cubic feet. At Whitton there is a tall slender treeabout 90 feet by 6 feet without a branch for 46 feet, and with a clean stem to 70feet up. At Betchford Park, Surrey, Henry measured a tree 90 feet high by 11 feet4 inches at 3 feet from the ground, dividing into two stems at 4 feet. At EnvilleHall, Stourbridge, he saw one which was 87 feet by 8 feet 2 inches. Sir HughBeevor tells us of a large one at Shottisham, Norfolk, 70 feet high and 18 feet ingirth near the ground, out of the base of which a mountain ash of large size isgrowing.At Holme Lacy there is a large tree near the home farm, which has a shortbole 6 feet high by 18 feet in girth, dividing into four main stems about 60 feethigh ; I estimated the contents of this tree to be not far short of 300 cubic feet.On the banks of the Nene near Lilford Hall, Northamptonshire, there is a finerow of large and picturesque alders, of which Plate 252 gives a good representation,but I was unable to measure them on account of the water.In Boughton Park, near Kettering, the property of the Duke of Buccleuch, aremarkable alder is growing near the Broad Avenue, in comparatively dry ground.It measures 70 feet by 9|- feet, and has the lower part of the trunk covered withbark so like that of an elm that it was difficult to recognise it as an alder, by thetrunk alone.At Aldermaston Park there is a very large old tree which looks as if it had beenpollarded, and which in 1906 was 17 feet 4 inches in girth.At Elvaston Castle, Derby, Mr. A. B. Jackson has seen a remarkably finetree, which he estimates at 90 feet by 7^ feet, with a clean bole 60 to 70 feetlong.In Wales the finest alder I have seen is in a wood at Penrhyn Castle whichis about 75 feet by 6 feet 9 inches, with a clean bole 40 feet long. The contentsof this tree were estimated by me at over 100 feet of timber.Of the cut-leaved alders the finest I have seen is about 68 feet bv ioA- feetJ &on the banks of the lake at Syon. At Melbury there is a tree very similar insize and appearance to that at Syon ; and Henry measured, in 1905, one atCassiobury Park, 85 feet by nj feet; and another at Belton, 85 feet by 10 feet2 inches. Colonel Birch Reynardson sends me a photograph of a tree atAlnus 943Holy well Hall, Stamford, which is 65 feet high and no less than 17 feet in girthat two and a half feet, dividing into two trunks of equal height. At Audley End,Henry measured, in October 1908, a fine specimen, 65 feet high, and 15 feet ingirth at a foot from the ground, dividing above into three great stems, with aspread of branches 80 feet in diameter.In Scotland the alder attains as large a girth and perhaps a greater age than inEngland. One of the oldest and largest on record is in the flower gardens at GordonCastle, and was figured in iSSi. 1 It was only 35 feet high but 14 feet 9 inches ingirth and with a spread 63 yards in circumference. The late Mr. J. Webstersupposed it to be nearly three centuries old, but when I saw it in 1907, nothingremained but a hollow stump 16^ feet in girth. At Fasnakyle, in Strathglass,there is a very large old alder, of which Mr. Stevenson Clarke sends me a photograph, and which in 1904 measured 18 feet in girth at two feet from the ground.It has a rowan tree 2 feet in diameter growing on it.At Shanbally, near Dumfries, Henry measured an alder in 1904 which was 65feet by 10 feet 5 inches, dividing at eight feet into two stems ; and at Scone he sawa cut-leaved alder, which was 66 feet by 6 feet 3 inches in the same year.Some of the most remarkable alders in Ireland grow in the old Deer Park atKilmacurragh on strong wet land covered with tall bracken and rushes. Henrythinks that in former times this park may have been part of the virgin forest ofWicklow. The trees are scattered in groups as if self-sown; and though fullymature, are in most cases sound. They average 55 to 60 feet in height, and in somecases have clean boles 30 to 40 feet high and 8 to 10 feet in girth. The onefigured (Plate 253), was about 60 feet by n feet 4 inches, and stands near thetop of the park.At Powerscourt, an immense alder was felled in 1902, the butt being 20 feet ingirth near the ground. At Churchill, Armagh, a tree growing in peat soil, measured,in 1904, 94 feet in height by 6 feet 4 inches in girth, with a clean stem to 60 or70 feet.TIMBERAlder is now a wood of third-rate importance in the English timber trade, andthough still used for making the clogs 2 which are worn in Lancashire, is so low invalue that it will not bear much cost for carriage. On this account it is usuallyworked up on the ground where it grows into pieces suitable for clog soles bymen who travel about from place to place, and this work is carried on by preferencein summer.In Scotland, however, alder is still used for making herring barrels, in districtswhere it is plentiful, though imported staves as usual are taking its place. Thetimber is said to be most valuable for piles, and to be certainly durable under water,1 Trcms. Scot. Arbor. Soc. ix. pi. I (1881).2 A complete set of specimens, illustrating the manufacture of clog-soles from alder wood, was sent to the Kew Museum,in 1904, from Enniscorthy, Ireland. Cf. Kew. Bull. 1 904, p. 6, where the clog-sole industry is described.

944 The Trees of Great Britain and Irelandso that where milldams, weirs, or similar work is being done on an estate, it may beprofitable to use it in preference to more costly wood.It is useless for posts or fencing, as it decays quickly when exposed to wet anddry conditions. It makes very good panelling, and is strong enough for inside worksuch as window-sills, and may be used for cheap furniture, but is said to be subjectto the attacks of wood-boring beetles, unless previously steeped for some time inlime water. The colour is a pale reddish brown; and the large burrs which arecommonly found on old trees show a very pretty figure when cut in slices, butare usually too small and full of flaws to have any marketable value, though SirThomas Dick Lauder says that handsome tables can be made of them. 1 Loudonquotes Mitchell to the effect that in Dorsetshire the local saying used for willow andpoplar in the midland counties, is applied to alder poles when peeled, viz.Thatch me well and keep me dryHeart of oak I will defy ;but, according to Cobbett, the bark must be taken off with a draw knife as soon aspossible after the poles are cut, and even then they will only last a year or two ashop poles.Alder wood which is dug up from peat bogs is said to become as black asebony, but I have seen none large or sound enough to be used like bog oak. Asfuel the wood is little valued in England, though Mouillefert says that in France it isconsidered specially suitable for heating ovens and glass works, though considerablyinferior in heating power to beechwood. The charcoal made from it was at one timein great demand for making gunpowder, but, so far as I can learn, is now little usedfor that purpose. The wood is said to be used on the continent for making cigarboxes; this may be the case for very cheap cigars, but all the cigar boxes I haveseen appear to be made of the wood of the West Indian "cedar," Cedrela odorata.The bark was used for tanning in the north, but only contains about 16 per centof tannin, and Mouillefert says that in France a black dye used for felt is madefrom the bark and sulphate of iron.Another use for alder wood which seems to be little known is the making ofhat-blocks, an industry carried on in Dunstable, Luton, and other towns near theprincipal hat factories. For this purpose the larger sized trees are preferred, cut intoplank of not less than 10 inches wide and 3 inches in thickness. I am informedthat in consequence of the increasing difficulty in procuring this wood in England,it is now imported from the continent, and as much as is. 3d. a foot is paid at theports on the north-east coast.The following details of the cost of making clog soles from alder were taken onmy own estate in 1908 :About 100 trees, estimated at 56 years old, growing on the bank of a streamat Colesborne, on an area of about one-third of an acre, were sold at yd. and 8d. a1 Alder burrs seem to have been a favourite wood with cabinet-makers in Sweden in former times, as I saw severalhandsome cabinets veneered with this wood in the Northern Museum at Stockholm.Alnus 945foot, and as measured down to 3 inches diameter, produced 817 cubic feet andrealised ....... £24 1 0 oDeduct expenses of cutting and hauling out . . £ 2 16 o„ „ loading and delivering to station at6s. per ton . . . . . 2804Leaving a net return of.£19 6 oFour men were occupied for 42 days in working the timber up on the groundand produced :dozen pair ist size, men's, at a cost of is. 4d. per dozen £13 2 O209 2nd women's, is. 2d.12 3 103rd » boys', „ l od.4 2 i» » 4th » children's, 8d.3 '9 8Total for labour£33 7 7The maker informed me that the cost of carriage to Oldham was ^8, 2s. at therate of £ i per ton, and that the sum realised was £72, in addition to which he hadthe whole of the waste and chips to sell for firewood.ALNUS INCANA, GREY ALDERAlnus incana, Mcench, Meth. 4 24 (1794); Willdenow, Sp. PI. iv. 335 (1805); Loudon, Art. et Frut.Brit. iii. 1687 ( l8 3 8); Willkomm, Forstliche Flora, 349 (1887) ; Mathieu, Flore Forestiere, 4 26(1897); Winkler, Betulacea, 1 20 (1904).Alnus lanuginosa, Gilibert, Exercit. Phyt. ii. 402 (1792).Alnus glauca, Michaux, f., Hist. Art. Amer. iii. 322 (1813).Betula Alnus incana, Linnaeus, Sp. PL 9 83 (1753).Betula incana, Linnaeus, f., Suppl. 4 17 (1781).A tree, attaining about 70 feet in height and 6 feet in girth. Bark smooth andsilvery grey, only fissuring slightly at the base of old trunks. Young branchletsgreyish pubescent. Leaves (Plate 268, Fig. i) about 3 inches long and 2 inches wide,ovate or oval, rounded or cuneate at the base, acute or slightly acuminate at theapex; lateral nerves nine to twelve pairs, running straight to the margin, eachending in a short acute lobe, which is finely serrate and ciliate; upper surface dull,dark green, pubescent; lower surface greyish, covered with soft hairs, densest onthe midrib and nerves, without axil-tufts ; petiole, f inch long, pubescent.Catkins in number and position like those of A . glutinosa ; but male catkinslooser, with distant shining red-brown scales and yellow anthers. Cones smallerthan in A . glutinosa, with more numerous scales, thinner and less distinctly fivelobed.Nutlets depressed, pentagonal, reddish-brown, with wing almost as broad asthe body.In winter the twigs are three-angled at the tip, and densely covered with a fineIV 2 H

946 The Trees of Great Britain and Irelandpubescence. Leaf-scars and stipule-scars as in the common alder. Buds reddishbrown,ovoid, conspicuously stalked, with two external scales, finely pubescent onthe surface and only slightly glandular. Pith triangular or three-lobed.The grey alder exhibits in the wild state considerable variation in the shape andpubescence of the leaves, and the cones may be sessile or shortly stalked. Manyvarieties are mentioned by Winkler and Schneider, most of which are scarcely worthdiscriminating ; but the following are noteworthy :1. Var. argentata, Norrlin. Leaves silvery on both surfaces, and covered witha dense silky pubescence. Observed in Finland, Silesia, Saxony, and Switzerland.2. Var. glauca, Regel. Leaves bluish-green and nearly glabrous beneath.3. Var. orbicularis, Callier. Leaves small, almost orbicular, with five pairs oflateral nerves, wild in Silesia. This is occasionally cultivated under the name of var.parvifolia; but var. parvifolia, Regel, which occurs in Sweden and Finland has stillsmaller leaves, only \ inch in length, and ovate in shape.4. Var. acuminata, Regel. 1 Leaves (Plate 268, Fig. 9) divided more than halfway to the midrib, into three to six pairs of long, narrow, triangular, serrate segments.This form has been observed wild in Sweden, and has been much confused withanother wild variety in the same country, var. pinnatifida, Wahlenberg, 2 whichresembles in the shape of the leaves A . glutinosa, var. incisa ; and has not beenseen by us in cultivation. Var. acuminata is common in gardens, and is usuallyknown as var. incisa or var. pinnatifida.5. Var. aurea, Schelle. Leaves and fruit yellow. This variety is growing wellat Aldenham, and is striking in appearance.6. Var. montrosa, Dippel. A dwarf shrub, with the tips of the branches ribbonlikeand fasciated, which originated in Spath's nursery. In cultivation at Aldenham.DISTRIBUTIONThe grey alder is widely distributed throughout the greater part of Europe andthe Caucasus. It is also met with in North America, where, however, it is only ashrub, commonly growing in swamps and on river banks, and forming dense thicketsrarely more than 10 or 12 feet high, and is spread throughout British territory fromNewfoundland to the eastern base of the Rocky Mountains, descending in theUnited States to New York, Pennsylvania, Wisconsin, and Nebraska.Alnus incana is replaced in northern and eastern Asia by two closelyallied species : A . hirsuta? Turczaninow, not in cultivation, a native of Siberia,Kamtschatka, Manchuria, Saghalien, and Japan ; and A . tinctoria, Sargent, 4 whichis confined to Manchuria and Japan. The latter species differs mainly fromA. incana, in the larger size and different shape of the leaves, which are broadly1 Mem. Sec. Nat. Mosc. xiii. 158, t. 17, f. 8 (1861). * Fl. S-uec. 622 (1824).3 Bull. Sac. Nat. Mosc. 1 838, p. 101. Alnus incana, var. hirsuta, Spach.4 Garden and Forest, x. 472, f. 59 (1897). Alnus incana, var. glaaca, Shirasawa, Icon. Ess. Forest. Japan, t. 19, ff.I-I7 (1900). This species is known in Japan as yama-harinoki, or mountain alder, and is much used for making smallarticles in the Hakone mountains. Cf. Rein, Industries of Japan, 239, 336 (1889).Alnus 947ovate, 4 to 6 inches long, 3^ to 5 inches wide ; base broad and rounded or truncate,occasionally cuneate; apex acuminate or cuspidate; pubescent on both surfaces,glaucous or brownish beneath; with 9 to 12 pairs of nerves, each ending in atriangular serrated lobule; petiole an inch or more in length. The amount ofpubescence on the branchlets, petioles, and leaves is variable; but the buds appearto be always densely pubescent. The cones are much larger than those of A . incana,attaining about f inch in length and \ inch in diameter. A . tinctoria grows in Yezo,according to Sargent, on low slopes in rich moist ground, usually at some distancefrom the banks of streams, which are generally occupied by A . japonica. A.tinctoria attains in Japan 60 feet in height, and 6 feet in girth; and was collectedby Elwes at Asahigawa in Yezo. It was formerly in cultivation at Coombe Wood,where it was probably raised from seed sent by Maries; but no specimens cannow be found there; and the only one which we have seen in England is a treeat Aldenham, 1 about 15 feet in height, which is reported to be growing vigorously.There are trees of A . tinctoria in the Arnold Arboretum, Massachusetts, whichwere raised from seed collected by Sargent in Japan in 1892.The grey alder extends in Europe much farther to the northward than thecommon alder, its northern limit in Scandinavia being about lat. 70 30'. InFinmark it reaches the mouth of the river Tana, and following the shore of theArctic Sea, the northern limit extends throughout Russia along the Arctic Circle.Its distribution is divided into two areas, a northern one extending southward in theplains of Russia to the 55th N. parallel; and a southern area, which comprises themountain ranges of the Carpathians, Alps, Jura, and Apennines, where the treegrows at high elevations in the mountains, and descends along the river valleys tolower altitudes, as along the Rhone, Isere, Drome, Durance, and Var in France, andalong the Rhine and its tributaries in Germany, and along the Danube in Austria.Its southern limit passes westwards from Russia through Transylvania to Banat andServia; but the tree is not found in Croatia, Dalmatia, or Istria. In Italy itdescends along the Apennines as far south as lat. 43 40', and grows as a rulebetween 4000 and 6000 feet, occasionally as low as 3000 feet. It ascends in theErz mountains to 2100 feet, in the Swiss Alps and the Tyrol to 5000 feet, and inFrance thrives at 6000 feet altitude near Barcelonette and Brianqon.In Scandinavia the grey alder is common in the pine and spruce forests, usuallyoccurring as underwood; but in favourable situations near streams attaining a considerable size. There are many fine specimens in the beautiful natural park, close toGefle on the Baltic. These trees, many of which are suckers from the roots of oldtrees that had been felled, are narrowly pyramidal in habit. The largest measured75 feet in height and 5 feet in girth. In Denmark, Mr. Prytz of the forest service,who has measured trees 65 feet high and 7^ feet in girth, informed me that thewood had been tested, and clogs made of A . incana had worn as well as those manufactured from the common alder. I saw several trees in a beech forest nearNykjobing averaging 60 feet in height and 3 or 4 feet in girth. In Denmark itgrows better on dry soil than the common alder.1 At Aldenham, and in gardens on the Continent, this species is cultivated under the erroneous name, A. incana, var.hirstita. Cf. Schneider, LaubhohkunJe, i . 134 (1904).

948 The Trees of Great Britain and IrelandIn the Baltic Provinces of Russia it is common as coppice treated with a shortrevolution ; and often takes possession of forests, when the larger trees have been cutaway, and succeeds in doing so, as it is able to grow very well on dry soil. In Germanyand Austria it grows chiefly on the banks of streams and rivers, but is also met with onhilly ground and on mountain precipices. It is very rarely met with on peat-bogs.In the Alps it is especially common on gravelly soil, and it is the most commonspecies in many places, where the mountain torrents form vast areas of gravel andsand, through which their branches spread in all directions. One of the mostremarkable and beautiful of these woods is situated at 2500 feet elevation onthe river Romanche near Bourg d'Oisans in I sere. The whole area is about 200acres, one-half of which is composed of a dense wood of grey alder, mixed witha small number of aspens and ashes, the other half being more open and consistingof a mixture of grey alder and white willow. The dense wood is treated as coppice,with a revolution of thirty years, forty standards per acre being reserved each timeof felling. When cut, the grey alder produces vigorous shoots, which grow rapidlytill they are thirty-five or forty years old; after which time growth ceases and theshoots begin to die. At Bourg d'Oisans natural seedlings are very numerous.The grey alder, unlike the common alder, suckers freely from the root, often ata great distance from the parent stem. It layers easily, and can also be propagatedby cuttings. This facility of reproduction renders it of great service for there-afforestation of the mountains in France, especially in the difficult work ofplanting trees on the sides of the torrents, where the soil is easily washed away.Alnus incana is not a native of the British Isles, and has not yet been discoveredin the fossil state there. (A. H.)CULTIVATIONThough the tree is hardly known to English foresters, I believe that it maybecome an exceedingly useful one on account of its extreme hardiness, rapidity ofgrowth and ability to thrive in very cold heavy soil, and in places subject to late andearly frosts. I have used it with great success as a nurse to trees like Thuya plicata,in situations which were too wet and cold for that tree when young, and believe thatit might be economically used for quickly suppressing rank herbage which wouldsmother more tender and slower-growing trees in low and damp situations. It canbe procured quite cheaply from French nurseries as one- or two-year seedlings, andgrows with extraordinary rapidity on any soil, providing a dense cover, and renderingthe land fit for planting. It soon overtops other trees, and if left standing requiresthe branches to be lopped so as to allow their heads to get up. It seems to thriveequally well on wet ground, and to grow much better than the common alder on soiltoo dry for that tree. I believe that the wood is at least as good, and according toMouillefert is less brittle, than that of the common alder.Though Loudon says that it was introduced as long ago as 1780, I have neverseen a tree of any size in England ; but Sir Hugh Beevor has sent me a photograph of one at Hargham in Norfolk, which measures about 72 feet high by 3 feetAlnus 949in girth, being drawn up in a wood by other trees. It is growing on good loam overclay, and is the only big pole of the species in the wood; others which havebeen cut throw up many shoots from the stool, of which the majority die. On LordCastletown's property at Doneraile, Co. Cork, there is a wood, 1 partly composed ofgrey alder, which has in places covered the ground with its suckers. (H. J. E.)ALNUS CORDATA, ITALIAN ALDERAlnus cordata, Desfontaines, Tab/. Hort. Paris, 2 44 (1815) ; Winkler, Betulacea, n o (1904).Alnus cordifolia, Tenore, Flor. Neap. i. Prod. p. Ixiv. (1811), and ii. 340 (1820); Loudon, Art. etFrut. Brit. iii. 1689 (1838); Baillon, Nat. Hist. Plants, vi. 223, figs. 158-164 (1880); Masters,Card. Chron. xix. 284 f. 42 (1883); Mathieu, Flore Foresticre, 4 28 (1897).Betula cordata, Loiseleur, Notice, 1 39 (1810), ex Loiseleur, Fl. Gall. ii. 317 (1828).A tree, attaining 80 feet in height. Bark greyish-brown, smooth or slightlywarty. Young branchlets, three-angled at the tip, stout, glabrous. Leaves (Plate 268,Fig. 4) about 4 inches long and 3 inches broad, oval or ovate, cordate at the base,shortly and abruptly acuminate at the apex ; margin not lobulate, regularly serrate ;nerves six to ten pairs, looping before reaching the margin ; upper surface darkgreen, shining, glabrous ; lower surface light green, glabrous, except for axil-tufts ofrusty brown pubescence; petiole i to 2 inches, glabrous. Male catkins, three tofour in a terminal raceme. Cones, solitary or two to three in an erect terminalraceme, i to \\ inches long, ovoid. Nutlet, sub-orbicular, with a thin narrow wing.In winter the twigs are glabrous, with leaf-scars and stipule-scars like those ofthe common alder. Buds long-stalked, arising from the twigs at a wide angle, ovoid,beaked at the apex, glabrous and covered with wax glands ; scales ciliate in margin.Seedling z :—Cotyledons oblong-oval, slightly fleshy, pale green, about % inchlong, with a very short grooved petiole. Caulicle pubescent, about £ inch long,ending in a tapering flexuose tap-root. Young stem brown, pubescent. Leavesalternate; first pair broadly ovate, acute or cuspidate, irregularly and acutely serrate,with pubescent petioles ; ultimate leaves cordate, cuspidate.This species shows no variation in the wild state, except that the leaves areoccasionally rounded and not acuminate at the apex. It differs considerably fromA. siibcordata, which has been supposed to be a variety of it; and is readily distinguished from all other species by the conspicuous cordate base of the leaves.Almis elliptica, Requien, A nn. Sc. Nat. v. 3 81 (1825) is a remarkable naturalhybrid between A . cordata and A.glutinosa, which was originally found growing onthe banks of the river Salenzara in Corsica. It has leaves, similar in size to those ofA. cordata, but thinner in texture, oval or elliptical, rounded at the base and apex;margin not lobulate, finely and equally serrate; glabrous on both surfaces, exceptfor axil-tufts beneath. The fruits are not so large as in A . cordata, and are inter-1 Described by Prof. Fisher in Quarterly Journal of Forestry, ii. 95 (1908).* Cf. Lubbock, Seedlings, ii. 531, f. 666 (1892).

95 The Trees of Great Britain and Irelandmediate between that species and A. glutinosa. The branchlets are glabrous andcovered with wax glands. This hybrid, which in general aspect strongly resemblesA. cordata, but is readily distinguished by the thinner leaves, not cordate at thebase, appears to be very vigorous in growth at Kew, where there is a tree growingbeside the lake, which is 72 feet in height and 5 feet in girth. The bark is likethat of A . cordata, being greyish in colour and slightly warty on the surface.Alnus cordata has a very restricted distribution, being confined to Corsica andsouthern Italy. In Corsica it ascends to 3000 feet, as at Vizzavona, where I sawit in a beech forest, growing not only beside a stream, but also on the side of thehill at some little distance off. Here the trees were about 70 feet high and 5 feet ingirth, with clean timber to 50 feet, and were narrowly pyramidal in habit, withascending branches. It grows in southern Italy from the Bay of Naples southwards ; and according to Tenore occurs both on marshy ground and in themountains. It forms woods on Mt. Serino.This elegant species, with foliage somewhat resembling at a distance that of theCaucasian lime, which is retained late in the autumn, was introduced, according toLoudon,1 in 1820. It flowers in March, before the leaves appear; and seems togrow as fast and to be as hardy as the common alder.It supports well the climate of the north of France ; and at Nancy, wherethe winters are severe, flowers and fruits regularly, and has attained 7 inches indiameter after twelve years' growth. According to Mouillefert,2 it succeeds betteron dry soils than either the common or the grey alder; and has been planted onthe chalky soil of Champagne, where it is treated as coppice with a short revolution.At Grignon it has borne 4 Fahr. without injury, but suffered in 1880, whenthe temperature fell to 13 Fahr. Here on poor chalky soil it has attained, atthirty-five years old, 48 feet in height and 2 feet 8 inches in girth; and on bettersoil, 64 feet by 3 feet i inch.The finest tree that we have seen of this species grows on the lawn at TottenhamHouse, Savernake, Wilts, and is a well-shaped tree, measuring no less than 69 feethigh by 9 feet 3 inches in girth at four feet from the ground (Plate 254). WhenElwes found it on April 3, 1908, it was in full flower, and covered with the cones ofthe previous year. It does not appear to be a very old tree, and is growing in adeep and rather heavy soil overlying chalk, at an elevation of about 400 feet.In the new park at Merton Hall, Thetford, a tree, growing in a wind-sweptsituation, on very dry, light, sandy soil, measured in 1908, 50 feet high and 10 feetin girth, with a spread of branches 56 feet in diameter. Lord Walsingham believesthat this tree was planted about 1843, as the new park was enclosed in the precedingyear. The bark at the base is deeply fissured and scaly.1 Arb. et Frut. Brit. 1689 (1838); but in Loudon, Card. Mag. 1 837, p. 143, and 1839, p. 39, a tree at BritwellHouse, Bucks, growing on gravelly soil, was reported to be 60 feet high ; and this would show that the date of introductionwas earlier than 1820. So far as we can learn, this tree no longer exists.2 Essences Forestiires, 252 (1903). However, two trees at Verrieres near Paris, about 80 years old, have only attained60 feet in height and 5 feet 8 inches in girth ; and M. Philippe L. de Vilmorin states (Hortus Vilmorinianus, 54 (1906)),that their growth seems to have come long ago to a standstill.Alnus 95 1A specimen is growing at Milford House, near Godalming, which was plantedby the famous botanist and traveller, Phillip Barker Webb. The present owner,R. W. Webb, Esq., informed us in 1905 that it was very healthy, measuring 8 feetin girth, and estimated to be about 50 feet in height.There is a fine tree growing near the pond in front of the palm-house in KewGardens, which {371 feet high by 5 feet 8 inches in girth. At Tortworth, a treemeasures 60 feet high by 6 feet in girth; and at Waterer's Nursery, Knaphill,Woking, another is 50 feet by 5 feet 10 inches.At Nuneham Park, Oxford, a tree, growing on hilly dry ground, on the greensandformation, measures 51 feet by 5 feet 5 inches, and is very thriving. Elweshas seen a tree at Bicton, measuring 65 feet by 5 feet 10 inches, and another atMelbury, where it grows vigorously and fruits.In the playing fields at Eton, on the banks of the Thames, there are two trees,the larger of which is 40 feet high by 6 feet 4 inches in girth. These were in fullfoliage on I7th November 1907, having scarcely lost a leaf, and were bearing fruit.They have not developed tall straight stems, as in the other places where the tree isthriving; and this is probably owing to their position being exposed to easterly andnorth-easterly winds. At Ponfield, Herts, a young tree 35 feet by 2 feet 4 inchesin 1906 is doing well on dry soil; and there is a good specimen in the CambridgeBotanic Garden. Another at Yattendon Court, Berks, is 50 feet by 3 feet 8 inches.In Scotland it also grows well at the Botanic Gardens, Edinburgh, where thereis a tree 60 feet by 5 feet 10 inches; and in the west at Castle Kennedy, and atMonreith, where Elwes saw a tree 30 feet high, bearing cones in September 1906.There is a fine specimen in the Glasnevin Botanic Garden, which is 64 feethigh by 5 feet 4 inches in girth. (A. H.)ALNUS SUBCORDATA, CAUCASIAN ALDERAlnus subcordata, C. A. Meyer, Verz. Pfl. Kauk. 43 (1831); Winkler, Betulacea, 1 12 (1904).Alnus cordifolia, Tenore, var. subcordata, Regel, in Mfm. Soc. Nat. Mosc. xiii. 170 (1861), and inDC. Prod. xvi. 2, p. 185, (1868).A tree, attaining about 60 feet in height. Bark grey, warty on the surface,ultimately scaling at the base of old trunks. Young branchlets pubescent. Leaves(Plate 268, Fig. 5) about 4 inches long and z\ inches broad, ovate-oblong, roundedand unequal or subcordate at the base, cuspidate-acuminate at the apex; coarselyserrate or bi-serrate in the upper half, finely serrate in the lower half; nerves, abouteight pairs, running to the margin; upper surface dark green, slightly pubescent;lower surface light green, pubescent throughout, the pubescence densest along thenerves and in the axils; petiole, f inch, pubescent. Staminate catkins, three to five ina raceme. Cones solitary or several, ovoid-elliptic, about an inch long; nutletsbroadly ovoid, with a very narrow wing.This is a moderate-sized tree, occurring in the province of Talysch in

952 The Trees of Great Britain and Irelandthe Caucasus, and in the provinces of Asterabad and Ghilan in Persia, where Dr.Stapf informs me that he has seen large trees south of the Caspian Sea.It is closely allied to, if not identical with Alnus orientalis, Decaisne, a nativeof Asia Minor and Cyprus. The latter has not been introduced, so far as we know,into English or continental gardens.It was known in cultivation a good many years ago in France, as Gay records 1a tree 30 feet high growing at Verrieres in 1861 ; but we are unaware of the exactdate of its introduction into England. It appears to grow as well and to be as hardyin England as A . cordata; and a fine tree, growing near the lake in Kew Gardens,is 52 feet high and 4 feet 10 inches in girth. There is a small specimen at Aldenham.(A. H.)ALNUS FIRMAAlnus firma, Siebold et Zuccarini, Abh. Akad. Miinchen, iv. 3. p. 230 (1845); Sargent, Forest FloraJapan, 6 3 (1894); Winkler, Betulacece, 1 02 (1904).Alnus Sieboldiana, Matsumura, Journ. Coll. Sti. Tokyo, xvi. 5, p. 3 (1902).Alnus yasha, Matsumura, of. fit. p. 4 (1902).Alnus pendula, Matsumura, op. cit. p. 6 (1902).Alnus multinervis, Schneider, Laubholzkunde, 1 23 (1904).A tree, attaining in Japan a height of 30 feet, but usually smaller. Youngbranchlets three-angled at the tip, pubescent. Leaves (Plate 268, Fig. 2) about 4inches long and if inch broad, plicate; nerves deeply immersed above and veryprominent beneath, about fifteen to eighteen pairs, running parallel and straightto the margin; ovate-lanceolate, rounded and unequal at the base, acuminate at theapex; upper surface dark green, shining, glabrous ; lower surface light green,pubescent, the pubescence strongest on the midrib and nerves; margin finely andregularly serrate, ciliate; petiole J inch, pilose; stipules often persistent, ovatelanceolate,£ inch long, membranous, glabrous. Buds sessile, conical, long-pointed,curved, green, glabrous, with two external scales.Flowers appearing in spring. Staminate catkins terminal or lateral, I to 2înches long. Pistillate catkins, one, two, or three to five, arising from one bud.Cones, solitary or racemose, variable in size, \ to i inch long, in the differentvarieties. Nutlets, obovate-oblong or sub-rhomboid with a membranous wing.This alder displays great variation in Japan, no less than three distinct speciesbeing recognised by Matsumura and Schneider. These appear to be geographicalvarieties:i. Var. multinervis, Regel in Bull. Soc. Nat. Mosc. xxxviii. 2. 423 (1865), andin DC. Prod. xvi. 2, p. 183 (1868). Alnuspendula, Matsumura. Alnus multinervisSchneider.Branchlets pubescent. Leaves with numerous lateral nerves, eighteenpairs or more. Cones in pendulous racemes, | inch long. This form is the1 Note with a specimen in Kew Herbarium. This tree is not mentioned in Hortus Vilmorinianus ( 1906).Alnus 9S3only one in cultivation, and grows on the banks of streams in sub-alpine regions inYezo and Hondo.2. Var. Sieboldiana, Winkler, loc. cit. Alnus Sieboldiana, Matsumura.Branchlets glabrous. Lateral nerves twelve to fifteen pairs. Cones solitary,i inch long. A native of the sea-coast in Hondo.3. Var. yasha, Winkler, loc. cit. Alnus yaslia, Matsumura.Branchlets pubescent. Lateral nerves twelve to fifteen pairs. Cones solitaryor racemose, f inch long. Occurs in mountain woods in Kiusiu, Shikoku, andHondo.According to Sargent, Alnus firma is largely planted along the borders of ricefieldsnear Tokyo, to afford support for the poles on which the freshly cut rice ishung to dry. He observed var. multinervis on the mountains of Hondo, where itgrows on dry rocky soil and reaches 5000 feet elevation, and describes it as agraceful tree 20 to 30 feet in height. The species, as mentioned above under thevarieties, is widely distributed throughout the whole of Japan.It was introduced by Sargent into New England in 1892 ; and, according toWinkler, was brought by Zabel into the forest garden of Munden in Germany.There are trees 6 to 10 feet in height in the collection at Kew. It is a remarkablydistinct species, with plicate many-nerved leaves, recalling those of two otherJapanese trees, viz.: Carpimis japonica and Acer carpinifolium ; and is worthy of aplace in collections of shrubs, as it scarcely can be considered to be a tree.(A. H.)ALNUS JAPONICA, JAPANESE ALDERAlnus japonica, Siebold et Zuccarini, Abh. Akad. Miinchen, iv. 3, p. 320 (1845); Sargent, Gardenand Forest, vi. 343, f. 53 (1893), and Forest Flora, Japan, 6 3, t. 20 (1894); Shirasawa, Icon.Ess. Forest. Japan, text 38, t. 19, ff. 18-34 (1900) ; Winkler, Betulacete, 1 14 (1904).Alnus maritima, Nuttall, var. japonica, Regel, in DC. Prod. xvi. 2, p. 186 (1868).Alnus maritima, Nuttall, var. formosana, Burkill, Journ. Linn. Soc. (Hot.) xxvi. 500 (1899).A tree attaining about 80 feet in height. Young branchlets usually glabrous.Leaves (Plate 268, Fig. 12) about 4 inches long and i^ to if inch wide, lanceolateor narrowly elliptical, cuneate at the base, long-acuminate at the apex; margin notlobulate, finely serrate; nerves, about twelve pairs, mostly running to the margin ;upper surface dark green, shining, pubescent on the midrib and nerves ; lower surfacelight green, glabrous except for minute axil-tufts; petiole about ^ inch, slightlypubescent. Buds minute, stalked, glabrous, glandular.Flowers, 1 appearing in spring, the fruit ripening in autumn ; otherwise similarto Alnus maritima.This species occurs in Japan, Manchuria, Korea, and Formosa. In Manchuria 2it grows along the sea-coast from St. Olga Bay southwards, and also inland, eithersolitary or in groups, in sandy soil along the rivers. It has been collected in Korea1 In Formosa, according to Burkill, the flowers are produced later, in summer; and he adduces this as a reason foruniting this species with the American A. maritima. 2 Komarov, Flora Manshuria, ii. 60 (1904).IV 2 I

954 The Trees of Great Britain and Irelandat Port Chusan; and in Formosa, it grows near Tamsui at the north end of theisland. According to Sargent, it is the most beautiful and largest of the alders inJapan, forming a pyramidal tree, often 70 or 80 feet in height, and well furnished tothe ground with branches clothed with large dark green lustrous leaves. It differsfrom A . maritima in the larger, differently shaped and coloured leaves, and in thetime of flowering.Sargent states that it is perfectly hardy in New England, where it grows rapidlyand promises to become a large and handsome tree. It was introduced, according toNicholson, 1 in 1886. There are small healthy trees in the collection at Kew; and atAldenham a specimen is about 14 feet high. (A. H.)ALNUS NITIDA, HIMALAYAN ALDERAlnus nitida, Endlicher, Gen. PI. Suppl. IV. ii. 20 (1847); Brandis, Forest Flora N.-W. India, 460,t. 57 (1874), and Indian Trees, 6 23 (1906); Hooker, Flora Brit. India, v. 600 (1888), andBot. Mag. t. 7654 (1899); Gamble, Indian Timbers, 6 70 (1902); Winkler, Betulacece, 1 08(1904).Clethropsis nitida, Spach, Ann. Sc. Nat. ser. 2, xv. 202 (1841); Cambessedes in Jacquemont,Vcy. dans flnde, Bot. 1 59, t. 159 (1844).A large tree attaining 100 feet in height and 15 feet in girth. Bark blackish,with thin quadrangular scales. Young branchlets densely and minutely pubescent.Leaves (Plate 268, Fig. 3) about 5 inches long, and 2\ inches broad, ovate-ellipticalor elliptical, base rounded or cuneate, apex acuminate; margin entire, obscurelycrenate, or remotely and slightly serrate; thin in texture; nerves, nine to twelvepairs, looping before reaching the margin; upper surface dark green, glabrous,shining; lower surface light green, glabrous except for slight axil-tufts ; petiole, f toi inch, minutely pubescent. Male catkins, 2 inches long, in terminal erect, oftenleafy racemes. Cones, f to i^ inch long, three to five in erect lateral racemes ;nutlet with a narrow thickened margin. The flowers in this species open inSeptember.Alnus nitida occurs in the north-western Himalaya, from Kashmir to Kumaon,usually at low elevations, 2000 to 4000 feet, fringing the banks of rivers, and notuncommonly descending with them into the plains. It occasionally, however, ascendsas high as 9000 feet; and is common on the Sutlej in the dry region of Kunawar, asfar as Spui on the right bank and Namgia on the left bank, according to Brandis,who states that it attains 100 feet high and 15 feet in girth; but Gamble has neverseen it so big, and says it is usually crooked and branching. An attempt wasmade recently to float out the wood from the Tons river forest, but failed,as the timber quickly became waterlogged. The bark is used for tanning anddyeing.Alnus nitida is one of the few Himalayan broad-leaved trees which have1 Diet. Gardening, Suppl. 34 (1900).Alnus 955succeeded in this country, where, however, it is little known, the only trees incultivation' that we know of being three thriving specimens which are growing nearthe lake in Kew Gardens. The largest of these is now 40 feet by 2 feet 3 inchesin girth. They were raised from seed sent by Mr. R. E. Ellis of the Indian ForestDepartment in 1882. (AH)ALNUS MARITIMAAlnus maritima, Nuttall, Sylva, i. 34 (1842); Sargent, Garden and Forest, iv. 268, t. 47 (1891),Silva N. Amer. ix. 81, t. 458 (1896), and Trees N. Atner. 2 15 (1905); Winkler, Betulacea,114 (i904)-Alnus oblongata, Regel, Mem. Soc. Nat. Mosc. xiii. 171 (i861) (in part).Betula-Alnus maritima, Marshall, Art. Am. 2 0 (1785).A tree attaining in America 30 feet in height and i foot in girth. Barksmooth, greyish-brown. Young branchlets slightly pubescent, three-angled at the tip.Leaves (Plate 268, Fig. 11) in cultivated specimens 2| inches long, if inch wide, somewhat larger in wild specimens, ovate or obovate, cuneate at the base ; acute, slightlyacuminate, or rounded at the apex ; nerves, eight to twelve pairs, running to themargin ; margin not lobulate, remotely serrate in the upper two-thirds with minuteincurved glandular teeth ; upper surface dark green, shining, glabrous ; lower surfacelight green, glabrous ; petioles ^ inch, slightly pubescent. Buds minute, stalked,ovoid, glabrous, slightly glandular.Flowers appearing in July on the branches of the year, and opening inSeptember. Staminate catkins in scurfy pubescent racemes in the axils of theupper leaves. Pistillate catkins usually solitary from the axils of the lower leaves.Cones ripening in the following September, so that both flowers and ripe fruit occursimultaneously on the tree, ovoid, |- inch long; nutlet obovate, narrowed andapiculate at the apex, with a thin membranous border.This alder grows on the banks of streams and ponds in Delaware and Maryland,usually near, but not immediately upon the sea-coast, as its name would seem toimply. However, it abounds on the banks of the Nanticoke and Wicomico rivers inMaryland, near the high-water mark. What appears to be the same species wascollected by Hall on the Red River in Indian Territory.It was introduced into cultivation by Thomas Meehan, who sent it in 1878 tothe Arnold Arboretum, where it is tolerably hardy, flowering and fruiting freely,though it was killed to the ground in 1885. There are now two trees, about 6 feethigh, growing in the nursery at Kew, which were sent by Prof. Sargent in 1899.These flower in September, and produce fruit in quantity. (A. H.)1 Mr. A. B. Jackson has lately seen a tree at Grayswood, Haslemere, which is 18 feet high and 9 inches in girth.

956 The Trees of Great Britain and IrelandAlnus 957ALNUS RUBRA, OREGON ALDERAlnus rubra, Bongard, Mem. Acad. St. Petersb. ii. 162 (1833); Winkler, Betulacece, 1 24 (1904);Sargent, Bot. Gazette, xliv. 226 (1907.)Alnus oregona, Nuttall, Sylva, i. 28, t. 9 (1842); Sargent, Silva N. Amer. ix. 73, t. 454 (1896), andTrees N. Amer. 2 10 (1905).Alnus incana, Moench, var. rubra, Regel, Mem. Soc. Nat. Most. xiii. 157 (1861).A tree attaining 80 feet in height and 10 feet in girth. Bark greyish orwhitish, thin, roughened by minute wart-like excrescences. Young branchletsglabrous, three-angled at the tip, scarcely viscid except at the beginning of theseason. Leaves (Plate 268, Fig. 16) about 4 or 5 inches long and 2^ inches wide,ovate or elliptical, rounded or cuneate at the base, acute at the apex; nerves, about15 pairs, each running straight and parallel to the apex of a lobule, which isfurnished with minute gland-tipped serrations; margin slightly revolute and ciliate ;upper surface dark green, slightly pubescent; lower surface whitish or greyish,covered with a minute brown pubescence; petiole, f inch, with a few scatteredhairs. Buds beaked at the apex, glabrous, stalked. Stipules ovate, acute, tomentose,^ to \ inch long.Flowers opening in spring before the leaves. Staminate catkins, three to six ina raceme, 4 to 6 inches long when fully opened. Cones, three to six in a raceme,^ to i inch long, with truncate scales, much thickened at the apex; nutlet orbicularor obovate, surrounded by a membranous wing.This species can only be confused with A . incana, from which it differs in theglabrous branchlets and the usually larger leaves with revolute margins. The budsalso differ, those of A . rubra being elongated, pointed, and glabrous, whilst those ofA. incana are shorter, rounded at the apex, and pubescent.Alnus rubra, according to Sargent, ranges from Sitka, where it oftenclothes mountain-sides to elevations of 3000 feet above the sea, southwardsthrough the islands and coast ranges of British Columbia, and through westernWashington and Oregon, and the canons of the Californian coast ranges, tothe Santa Inez mountains near Santa Barbara. It grows to its largest size in theneighbourhood of Puget Sound, where it commonly fringes the banks of streams andgrows in wet places. 1This species was introduced into cultivation a few years ago, and there are twotrees in Kew Gardens about 15 feet in height. Elwes has raised seedlings fromthese trees, which grow very rapidly in heavy soil at Colesborne, but being plantedin a situation very subject to late frosts, have suffered on several occasions, whenthe grey and common alders standing near were quite untouched.The wood is light, soft, brittle, and not strong, but close-grained and takesa fine polish; and is now largely used in Washington and Oregon for makingfurniture.(A. H.)1 In Vancouver Island, the stem and branches are often covered with Polypodium falcatum, the creeping rhizomes ofwhich find anchorage in its moss-covered bark. Cf. Postelsia, 1906, p. 76. A figure of the tree, growing in a moist part ofthe forest and surrounded by devil's dub (Echinofanax horridum), is given in Piper, Flora of the State of Washington,plate vii. (1906).ALNUS TENUIFOLIAAlnus tenuifolia, Nuttall, Sylva, i. 32 (1842); Sargent, Silva N. Amer. ix. 75, t. 455 (1896), andTrees N. Amer. 2 11 (1905); Winkler, Betulacece, 1 24 (1904); Schneider, Laubholzkunde, 1 33(1904).Alnus incana, Moench, var. glauca, Regel, Mtm. Soc. Nat. Mosc. xiii. 154 (1861) (in part).Alnus incana, Moench, var. virescens, Watson, in Brewer and Watson, Bot. Calif, ii. 81 (1880).Alnus occidental, Dippel, Laubholzkunde, ii. 158 (1892) ; De Wildeman, Icon. Select. Hort. Thenensis,ii. 147, t. 75 (1901).A tree attaining 30 feet in height and 2 feet in girth. Bark bright red-brown,broken on the surface into small scales. Young branchlets glabrous. Leaves(Plate 268, Fig. 15) about 3 inches long by 2 inches wide, ovate, broad and roundedat the base, acute or shortly acuminate at the apex; nerves, nine or ten pairs, runningparallel and straight to the margin, and ending in acute triangular lobes, which arefinely serrate; upper surface dark green, pubescent on the midrib and nerves; lowersurface yellowish green, glandular, pubescent on the midrib with slight axil-tufts;petioles pubescent, f to i inch long. Buds stalked, pubescent at the base. Stipulesdeciduous, lanceolate, acute, pubescent. Staminate catkins, three to four, in slenderstemmedracemes; stamens four. Cones, ovoid-oblong, ^ to \ inch long, three tofour in a raceme ; scales thickened, three-lobed and truncate at the apex; nutletsnearly circular, surrounded by a thin membranous border.This species, distributed over a wide area, shows two well-marked geographicalvarieties :1. Var. virescens, Callier. This is the commonest form, and has been describedabove (Plate 268, Fig. 15).2. Var. occidentalis, Callier. Alnus occidentalis, Dippel. Leaves (Plate 268,Fig. 14) larger, 4 to 5 inches long and 3 to 4 inches wide; nerves, twelve pairs;slightly bluish-green and pubescent throughout beneath ; stipules ovate, broad, obtuse.This variety is rare, and has only been observed in British Columbia and Oregon.This species is widely distributed in western North America. It occurs inBritish Columbia, from Francis Lake in lat. 61 to the valley of the Lower FraserRiver, and extends eastward along the Saskatchewan River to Prince Albert. Itextends southwards along the Rocky Mountains to northern New Mexico, and isthe common species in the northern interior region, east of the divide of the CascadeMountains, in eastern Washington, Oregon, Idaho, and Montana It is veryabundant on the southern California Sierra, forming great thickets at 6000 to 7000feet above the sea, along the head-waters of the rivers of southern California flowingto the Pacific Ocean. It is equally abundant and attains its largest size in Coloradoand northern New Mexico, and is met with in Nevada and Utah.This species is rare in cultivation. There are two or three trees of each varietyin the alder collection at Kew, which are about 15 feet in height, and show nospecial beauty or vigour. Var. virescens is thriving at Aldenham ; and a fine specimen

95 8 The Trees of Great Britain and Irelandat Grayswood, Haslemere, planted in 1888, is now about 30 feet in height and i foot8 inches in girth.According to De Wildeman, var. occidentalis was introduced from BritishColumbia in 1891, by Dieck and Purpus, into the arboretum at Zoschen.ALNUS RHOMBIFOLIAAlnus rhombifolia, Nuttall, Sy/va, i. 33 (1842); Sargent, Silva N. Amer. ix. 77, t. 456 (1896), andTrees N. Amer. 2 12 (1905); Winkler, Betulacece, 1 15 (1904).Alnus oblongifolia, Watson, in Brewer and Watson, Bot. Calif, ii. 80 (1880) (in part).A tree attaining 80 feet in height and 9 feet in girth. Young branchletspubescent. Leaves (Plate 268, Fig. 13) on young trees up to 5 inches long and3 inches broad, on old trees 3 inches long and i^ inch broad, ovate or oval, roundedand unequal at the base, acute or rounded at the apex ; margin slightly thickened andreflexed, finely and irregularly serrate, and ciliate ; nerves, ten or eleven pairs,running parallel and slightly curved to the margin; upper surface dark green,shining, glabrous ; lower surface light green, pubescent, the pubescence strongeston the midrib and nerves; petiole, \ inch, pubescent. Buds stalked, pubescent.Staminate catkins, in pubescent racemes, deciduous before the opening of theleaves; stamens, two or three, rarely one. Cones oblong, \ to \ inch long, fullygrown at midsummer, but remaining closed till the trees flower in the followingyear; nutlet broadly ovate with a thin, acute margin.This species grows on the banks of streams from northern Idaho to the easternslope of the Cascade Mountains in Washington and Oregon, extending southwardover the coast ranges and along the western slopes of the Sierra Nevada to themountains of southern California. It is the common alder of central California, andthe only species at low altitudes in the southern part of this state.It is extremely rare in cultivation, the only specimen which we have seenbeing a small plant in Lord Aldenham's remarkable collection of shrubs and treesat Aldenham, Herts. According to Nicholson, 1 it was introduced into cultivationin 1888.1 Diet. Gardening, Sufpl. 34 (1900).BETULABetula, Linnaeus, Sp. PI. 9 82 (1753); Bentham et Hooker, Gen. PI. iii. 404 (1880); Winkler, inEngler, Pflanzenreich, iv. 61, Betulacece, 5 6 (1904).Betulaster, Spach, in Ann. Sc. Nat. se"r. 2, xv. 198 (1841).Apterocaryon and Chamcebetula, Opiz, in Lotos, v. 258 (1855).DECIDUOUS trees or shrubs belonging to the order Betulacese. Bark smooth withlongitudinal lenticels, often peeling off in papery strips, and becoming on old trunksthick and furrowed near the base. Branchlets of two kinds: long shoots withseveral leaves and axillary buds, no true terminal bud being formed; and shortshoots or dwarf spurs, each with two (rarely one or three) leaves and a terminal bud.Buds viscid, elongated, ovoid, fully grown and green at midsummer, composed ofimbricated scales, but with the two basal ones short and lateral, usually only fourscales being visible externally ; inner scales accrescent, and marking in falling thebase of the shoots with ring-like scars. Leaves alternate, simple, stalked, penninerved;serrate, dentate or incised. Stipules lateral, enclosing the leaf in the bud, fugacious.Flowers monoecious, fertilised by the wind, in cylindrical catkins, composed ofclosely imbricated three-lobed scales, with three flowers on each scale. Malecatkins, 1 formed in the preceding autumn, clustered in the axils of the upper leavesof a long shoot, erect and naked during winter, pendulous in spring. Staminateflowers, with a one- to four- lobed calyx; stamens two, with short bifurcatedfilaments, each of the four branches bearing an erect half-anther, there being thusapparently twelve stamens on each scale. Pistillate catkins, solitary, or two to fourin a raceme, terminal on the short shoots, and appearing with the leaves in spring.Pistillate flowers, without a calyx, two-celled, with one ovule in each cell; stylestwo, stigmatic at the apex. Cones, ripening 2 usually in autumn, composed of woodythree-lobed scales and small fruits, deciduous together; nutlets oval or obovate,compressed, bearing the persistent styles at the apex, and with the outer shellproduced into a marginal transparent wing, interrupted at the apex; seed solitary,pendulous, without albumen.3In winter, species of Betula are readily distinguished by the short shoots onthe older wood, which end in a terminal bud, and are densely clothed with scars, aseach season's growth is very short and marked by two crescentic leaf-scars inaddition to the ring-like scars left by the fall of the scales of the bud of the previousspring. The long shoots show similar ring-like scars at the base, and bear axillary buds1 In some of the shrubby species the male catkins are solitary on the ends of the short shoots, and remain enclosed inthe buds during winter, appearing in spring. * In B. nigra the fruit ripens in May or June.3 The cones, scales, and fruits shown in Plates 269 and 270, were all drawn by Miss F. II. Woolward, except in thecase of Figs. 8 and 16.959

960 The Trees of Great Britain and Irelandarranged alternately. The leaf-scars, semicircular or crescentic, and three-dotted,have on each side a linear stipule-scar. The pith of the twigs is oblong in section.About thirty species of Betula are known, all natives of the northern hemisphere,extending from the Arctic circle to Texas in the New World, and to southernEurope, the Himalayas, China, and Japan in the Old World. A considerablenumber are shrubs, the treatment of which does not come within the scope of ourwork. Of the arborescent species, a few, either not introduced or imperfectly known,are not included in the following account.B. corylifolia, Regel et Maximowicz, though not yet introduced, is includedin the key and fully described below, as it is very distinct and has been muchconfused with other Japanese species.There are young plants in the nursery at Kew, received from the ArnoldArboretum, as B. globispica, Shirai, 1 which appear to be a very distinct species ; but,as there is no authentic material in the Kew Herbarium with which to compare themand ascertain if they are correctly named, it is unadvisable to deal with this speciesat present.Similarly, young plants of B. alnoides, Buchanan-Hamilton, var. pyrifolia,Franchet, growing at Coombe Wood, which were raised from seed sent from centralChina by Wilson in 1901, are left undescribed, as they show considerable variation,and we cannot be certain, until they have borne fruit, of their identification.KEY TO ARBORESCENT SPECIES OF BETULA IN CULTIVATIONI. Branchlets and leaves quite glabrous.1. Betula verrucosa, Ehrhart. Europe, Northern and Eastern Asia. Leaves bi-serrate, shortly acuminate.See p. 966.2. Betula populifolia, Marshall. North America. See p. 987.Leaves lobulate and irregularly serrate, ending in a long caudate acumen.II. Branchlets or leaves or both pubescent.* Leaves cordate at the base.3. Betula Maximowiczii, Regel. Japan. See p. 976.Leaves 5 or 6 inches long, broadly ovate.4. Betula ulmifolia, Siebold et Zuccarini. Japan. See p. 979.Leaves, 3 inches long, narrowly ovate.** Leaves cuneate at the base. *5. Betulapubescens, Ehrhart. Europe, Northern Asia, Greenland. See p. 962.Leaves light green beneath, rhombic-ovate, bi-serrate, with six pairs of nerves.Branchlets not glandular, clothed with minute dense erect pubescence.6. Betula davurica, Pallas. Manchuria, Korea, North China. See p. 974.Leaves light green beneath, narrowly ovate, bi-serrate, with six to eight pairsof nerves. Branchlets glandular, with minute erect pubescence interspersedwith a few long hairs.Betula 9617. Betula nigra, Linnaeus. North America. See p. 988.Leaves greyish beneath, ovate, acute, with large serrated teeth, and six toeight pairs of nerves. Branchlets glandular, tomentose.*** Leaves rounded or truncate at the base.A. Leaves orbicular or oval.8. Betula corylifolia, Regel et Maximowicz. Japan. See p. 975.Leaves acute, pale beneath with conspicuous silky hairs on the midrib andnerves ; nerves twelve to fourteen pairs.B. Leaves ovate-oblong, considerably longer than broad, and -widest near the middle.9. Betula lutea, Michaux. North America. See p. 990.Leaves, with nine to twelve pairs of nerves, pilose on the midrib and nerves ofboth surfaces, and on the petiole. Branchlets pilose.10. Betula lenta, Linnaeus. North America. See p. 991.Leaves with nine to twelve pairs of nerves, pilose on the midrib and nerves ofboth surfaces; petiole glabrescent. Branchlets glabrous, except for a fewhairs above the leaf-insertions.11. Betula utilis, Don. Himalayas, China. See p. 980.Leaves, with nine to twelve pairs of nerves ; lower surface with dense axiltuftsof pubescence; petiole tomentose. Branchlets tomentose.C. Leaves ovate, not much longer than broad, widest near the base.t Branchlets very glandular.12. Betula Ermani, Chamisso. Eastern Siberia, Manchuria, Japan. See p. 977.Leaves truncate at the base, with ten to twelve pairs of nerves, glabrescentand conspicuously glandular beneath. Branchlets glabrous.13. Betula fontinalis, Sargent. Western North America. See p. 992.Leaves thin in texture; rounded, truncate, or subcordate at the base, withsix pairs of nerves ; both surfaces minutely glandular, with scattered longhairs. Branchlets with long pale hairs.\ Branchlets not conspicuously glandular.14. Betula papyrifera, Marshall. North America. See p. 983.Leaves about 3 inches long, with six to eight pairs of nerves; lower surfaceglandular and with conspicuous axil-tufts. Branchlets pubescent orglabrous.15. Betula utilis, Don, vat.Jacquemontii, Regel. Himalayas. See p. 981.Leaves about 3 inches long, with seven or eight pairs of nerves; lowersurface glandular, with long hairs on the midrib and nerves, and withoutconspicuous axil-tufts. Branchlets with short, erect, dense pubescence.16. Betula luminifera, Winkler. Central China. See p. 980.Leaves, 5 or 6 inches long, with ten to fourteen pairs of nerves, pubescenton both surfaces. Branchlets with dense erect pubescence.(A. H.)1 The Japanese name of this species, according to Matsumura and Goto, \sjizo-kamba.IV2 K

962 The Trees of Great Britain and Ireland Betula 9 6 3BETULA PUBESCENS, COMMON BIRCHBetula pitbescens, Ehrhart, Beit. Naturk. vi. 98 (1793); Mathieu, Flare Forestiere, 4 15 (1897);Winkler, Betulacea, 8 1 (1904).Betula tomentosa, Reitter and Abel, Abbild. Holzart, i. 17 (1790).Betula alba, Linnseus, Sp. PI. 9 82 (1753) (in part); Roth, Tent. Fl. Germ. i . 404 (1788); Willkomm,Forstliche Flora, 3 02 (1887); Schneider, Laubholzkunde, i. 116 (1904).Betula alba, Linnaeus, \ax.pubescens, Loudon, Arb. et Frut. Brit. iii. 1691 (1838).Betula odorata, Bechstein, Diana, i. 74 (1797)-A tree, usually attaining 70 or 80 feet in height, and 5 or 6 feet in girth,occasionally larger. Branches ascending or spreading, branchlets usually notpendulous. Bark smooth, white, and papery, often peeling off in transverse shreds,with black triangular markings below the insertion of the branches ; thick and deeplyfurrowed at the base of old stems. Young branchlets covered with short, erectpubescence, often minute and only discernible with a lens, usually retained in thesecond year. Leaves (Plate 269, Fig. i), i^ inch to 2 inches long, f inch to i^ inchwide, rhomboid-ovate or ovate, usually cuneate at the base, acute or acuminateat the apex; margin ciliate and coarsely serrate; nerves, five or six pairs ; uppersurface with scattered pubescence; lower surface pubescent on the midrib andnerves ; petiole, £ inch, pubescent and glandular.Fruiting-catkins (Plate 269, Fig. i), cylindrical, about i inch long, \ inchwide, at first erect, afterwards pendulous, on long pubescent stalks; scales,puberulous, ciliate, with the central lobe more prolonged than is the case in B.verrucosa, and with the lateral lobes angular and usually erect, but occasionallyrecurved. 1In winter the twigs are slender, dark brown, densely covered with short, erectpubescence. Buds, \ inch long, ovoid, rather blunt at the apex, viscid, with glabrous,ciliate scales.Seedling. 2 Cotyledons, about ^ inch long, oblong-ovate, obtuse, glabrous, withpubescent petioles, about £ inch long. Caulicle short, pubescent, raising the cotyledons above the ground. Stem pubescent and non-glandular, bearing primaryleaves, which are alternate, ovate, cordate at the base, simply and coarsely serrate,and pubescent.The birch,3 under ordinary conditions of growth, does not produce root-suckers ;however, when cut down, although coppice shoots are not given off from the stool,the roots give rise to numerous tomentose shoots, which bear leaves larger thanthose of ordinary branches, cordate at the base, dentate in margin, and pubescent onboth surfaces.1 When the lateral lobes are recurved, the scales are similar in shape to those of B. verrucosa; and in such cases wemay suspect a hybrid between the two species or an intermediate form. The pubescent stalks of the catkins and thepuberulous scales are, however, apparently characteristic of B. pubescens.a Cf. Lubbock, Seedlings, ii. 541, fig. 672 (1892).3 Cf. Uubard, in Ann. Sc. Nat. xvii. 169, plate 2, fig. 4 (1903).The witches' brooms, which are so common on birch trees, are generally supposed to be due to a fungus, Exoascus betulinus, the threads of which penetrate theyoung growing twigs, causing them to branch repeatedly and thus form large,irregular, nest-like clusters. Miss Ormerod, 1 however, states that these abnormalgrowths are caused by the development of unhealthy buds, which have beenattacked by a gall-mite. 2 It is possible that in some cases it is the fungus, and inother cases the gall-mite, which is the cause of these witches' brooms.A large number of birches were killed in 1900 in Epping Forest by a fungus,identified, by Paulson,3 with Melanconis stilbostoma, Tulasne, which attacks the younggrowing branches.VARIETIESThis species is very variable in the wild state, both as regards the stature ofthe tree and the shape, size, and pubescence of the foliage. A large number ofvarieties have been distinguished by Continental botanists, of which the nomenclature is very confused; and as most of these are separated by inconstantcharacters and are of no value from the cultivator's point of view, it will be sufficienthere to refer the reader to the works of Willkomm, Winkler, and Schneider, wherethe different forms are fully dealt with. The following varieties are, however,worthy of note :1. In Alpine and northern localities this species is often met with as a smallshrub with twisted branchlets, but with leaves very variable in character. This groupof forms may be distinguished as var. tortuosa, Koehne, Deut. Dend. 109 (1893).2. Var. Murithii, Gremli, Excursions/I, f. d. Schweiz, 365 (1893).B. Murithii, Gaudichaud, Fl. Helv. vi. 178 (1830); Christ, Ber. Schweiz, Bot. Ges. v. 16 (1895).An Alpine shrub, occurring in the Bagnes valley, near Mauvoisin (Valais), inFribourg, and in the Joux valley (Vaud) in Switzerland. This has broadly ovateor ovate-triangular leaves, with large simple serrations, and prominent reticulatevenation beneath.3. Var. denudata, Grenier et Godron, Fl. France, iii. 147 (1855). This namemay be given to a series of forms, characterised by rhombic leaves, cuneate at thebase, and glabrous beneath or with only slight axil-tufts. This is often cultivated asvac.pontica*vz.r. carfiaticafvzx. odorata, etc., and is usually a tree of considerable size.4. Var. urticifolia, Spach, in Ann. Sc. Nat., ser. 2, xv. 187 (1841); Schneider,Laubholzkunde, i. 117 (1904).Betula urticifolia, Regel, in Mhn. Soc. Nat. Mosc. xiii. ITS (1860); Willkomm, ForstlicJie Flora, 3 13(1887); Winkler, Eetulacea, 80 (1904).1 Injurious Insects, 2 12 (1890).2 Gillanders, Forest Entomology, 2 5 (1908), identifies the gall-mite with Eriophes rudis, Canestrini; and gives a figureof swollen buds on the branch of a birch tree. These had been found in close proximity to a witches' broom.3 Essex Naturalist, xi. I, p. 273 (1901). Cf. also Nature, Ixii. 599 (1900). Mr. Massee thinks that root-rot orunsuitable soil conditions, rather than the fungus, were the cause of death of these trees.4 There is a good-sized healthy tree in the Botanic Garden at Glasnevin, under the name of B. alba, var. pontica,which was mentioned by Loudon as being, in 1838, thirty-five years old and 35 feet high.6 The true var. carpatica (B. carpatica, Waldstein and Kitaibel, in Willdenow, Sp. PI. iv. 464 (1805)) is a low treeallied to var. tortuosa, the distinctive characters of which are given in Schneider, of. ct't. 1 19.

964 The Trees of Great Britain and IrelandThis variety, which was long considered to be a distinct species, has beenshown by Beissner and Schneider to be an abnormal form of B. pubescens. Itdiffers from the type in the longer, more acuminate, slightly lobed leaves; and inthe fruiting-catkins, which are very long ( \\ inch or more) and slender, withpubescent ciliate scales, the middle lobe of which is elongated. The leaves andfruit are shown in Plate 269, Fig. 2.This peculiar form has been found wild in the province of Wermland, in Sweden,and is only a tree of small dimensions. It is often planted in botanical gardens.5. Some peculiar forms have arisen in cultivation, as var. aurea, young foliagetinted with yellow, sent out a few years ago by G. Paul, Cheshunt Nurseries; andvar. nana, a dwarf form.HYBRIDS1. Hybrids have often been observed between this species and B. verrucosa,and have received various names, as B. hybrida, Bechstein, in Diana, i . 80 (1797);B. aurata, Borkhausen, Forstbot. i . 498 (1800); B. glutinosa, Wallroth, Sched. Crit.497 (1822); and B. ambigua, Hampe, in Reichenbach, Fl. Sax. 1 20 (1842). Thesehybrids are intermediate in the characters of the branchlets, foliage, and fruit; andmay be suspected in cases where the branchlets are more 'or less glandular andshow slight pubescence.2. B. intermedia, Thomas, in Gaudichaud, Fl. Helvet. vi. 176 (1830), a hybridbetween B. pubescens and B. nana, is a shrub about 10 feet high, with smallleaves, which is found in the Swiss Jura, Greenland, Iceland, northern Europe,and Siberia. This has been collected 1 in a few localities in the Highlands ofScotland.DISTRIBUTIONThis species is widely distributed through Europe and northern Asia, extending farther northward than B. verrucosa, but not descending so far south. It is themost northerly tree in Europe, growing on the shores of the Arctic Sea from NorthCape (lat. 71 ) to the mouth of the White Sea; its northern limit eastwardthrough north-eastern Russia and Siberia to Kamtschatka being near theArctic circle, though in some localities it ascends a degree or two higher. Itssouthern limit in Russia and Siberia appears to be the edge of the steppes, onwhich the tree does not grow; but it occurs in the Caucasus and Armenia. Fartherwestward the southern limit is the Carpathians and the Alps; and the tree is notfound in the Apennines or in the Pyrenees, its most southerly point in France beingnear Grenoble. It is also a native of Iceland and of south-western Greenland.The distribution may then be roughly described as the northern hemisphere,from Greenland in the west to Kamtschatka in the east, between the parallelsof 45 and 71 .As compared with B. verrucosa, this species is found on the continent of Europeon wetter soils and in moister climates, and is the birch which grows on marshyground and on undrained peat-mosses.1 Cf. E. S. Marshall, \njoiirn. P.ot. xxxix. 271 (1901), and Bat. Exchange Club Report for 1904, p. 33.Betula 9 6 5In Norway, Schubeler distinguishes this species as the highland birch, andspeaks of B. verrucosa as the lowland birch, the latter not being found north ofSnaasen in lat. 64 12', or on the fells higher than 1600 to 1800 feet.In Russia it forms large pure forests in the provinces of Olonetz and Vologda ;while in Esthonia, Livonia, and Finland, it is mixed with pine, spruce, and aspen ;farther south it is gradually replaced by B. verrucosa, with which, however, it issometimes associated. Von Sivers* states that in the Baltic provinces this speciesis the characteristic tree of the low-lying moors, and on account of its resistance toMay frosts, holds its own with the spruce and common alder. On better soils itforms immense forests, where Betula humilis and Rhamnus frangula are the underwood, and which are the favourite summer resort of the elk. It never attains suchlarge dimensions as B. verrucosa, scarcely ever surpassing 100 feet in height.In northern Germany, large forests, composed mainly of this species andcommon alder, are common on marshy ground. In France it is usually met with inthe moister parts of the forests or on peat-mosses.It appears to be much more common, as a wild tree, in the British Isles, thanB. verrucosa, the moist climate being favourable to its development; and theextensive birch forests of the Highlands of Scotland are usually B. pubescens. Asthe two species have not as a rule been distinguished by collectors, and no discrimination has been made between planted and wild trees, it is impossible atpresent to give an accurate account of the distribution of the two species in thiscountry. (A. H.)The distinction between the two forms or species of common birch which isalmost universally admitted by Continental botanists and foresters appears to havebeen generally overlooked in Great Britain; and though most local floras admitboth, yet, after much inquiry and investigation, I have found it impossible to definetheir distribution as indigenous trees. In many districts where the birch nowreproduces itself by seed abundantly, the original parents were of both forms, whichare not distinguished by nurserymen, though the name " silver birch " is supposedto be, and should correctly be used for the rough-twigged form. The bark of thisbeing more silvery though this character is variable and disappears with age andthe habit more pendulous and graceful, it should be chosen as an ornamental tree.But, where birch is planted on peat bogs or wet moors in order to produce a timbercrop, and to prepare the ground for planting other trees, as has been recommendedby Mr. G. U. Macdonald,2 or to act as a nurse for other trees, it would be preferable to use the downy-twigged form, which is considered to be more naturally athome and to thrive better on wet than on dry rocky soil. :iBut both forms as well as their hybrids grow together in many parts ofEngland and Scotland, and my experience in planting them does not justify me insaying that there is a marked difference in their relative growth or size. However,1 Forst. Verhalt. Bait. Frtrv. 1 8 (1903).'l " Protection of Young Spruce from Frost," in Traits. Key. Scot. Arb. Soc. xix. 287 (1906).3 M. Bommer, Director of the Botanic Garden at Brussels, pointed out to me in the Museum there, characters in thebark by which he could distinguish the two common birches. It seems to me, however, that bark is, especially in thebirch, so much influenced by climate, soil, and the age and vigour of the trees, that these characters were nofreliable.

966 The Trees of Great Britain and Irelandwhen visiting the nursery at Balmoral in 1904, I noticed self-sown seedlingsof both forms, in which the distinction was very striking, and took some away withme, which I planted on my trial beds at Colesborne. Of these the growth of B.verrucosa\\a& been immensely superior; and Mr. J. Michie, Commissioner to HisMajesty at Balmoral, writing to me on the subject, says, " I have no doubt aboutthe rough and smooth twigged birches remaining constant through life; they aredistinct varieties, and in nature grow side by side frequently. I do not say thatthe rough-twigged variety always develops a pendulous habit with age, but itgrows larger and has a lighter colour of bark. It generally weeps, on the sameground where the other remains rigid, of less size, and with darker bark."Birch is the most Alpine tree in Great Britain, and ascends in the Highlands toabout 3000 feet.The Gaelic name of the birch is beith (pronounced bey), and according to SirHerbert Maxwell, 1 is found in various forms in Scottish place-names, as Drumbae,the birch-ridge, Auchenvey and Largvey in Galloway (achadh-na-beith, birch field,and learg-blieith, birch hill-side). Beoch in Ayrshire, Galloway, and Dumfriesshire isbeitheach (beyagh), birch-land.In Ireland, this word 2 occurs in many names of places, as Ballybay in Monaghan(mouth of the ford of the birch), Kilbeheny, a village in Tipperary (birch-wood),Aghavea in Fermanagh (birch-field). Beagh is also a common place-name in differentparts of Ireland. (H. J. E.)BETULA VERRUCOSA, SILVER BIRCHBetula verrucosa, Ehrhart, Belt. Naturk. vi. 98 (1791); Willkomm, Forstliche Flora, 3 14 (1887);Mathieu, Flare Forestiere, 407 (1897); Winkler, Betulacece, 7 5 (1904).Betula pendula, Roth, Tent. Fl. Germ. i. 405 (1788); Schneider, Laubholzkunde, i . 113 (1904).Betula alba, Linnaeus, Sp. PI. 9 82 (1753) (in part).Betula alba, Linnaeus, var. pettdula, Aiton, Hort. Kew, iii. 336 (1789) ; Loudon, Art. et Frut. Brit.iii. 1691 (1838).Betula rhombifolia, Tausch, in Flora, xxi. 2, p. 752 (1838).Betula lobulata, Kanitz, in Linnaa, xxxii. 351 (1863).Betula odorata, Reichenbach, Icon. Fl. Germ. xii. 2, t. 626, f. 1288 (1850) (not Bechstein).A tree, attaining in Russia 120 feet in height and about 12 feet in girth.Bark like that of B. pubescens, but more silvery white in colour. Main branchesascending, smaller branches and branchlets pendulous. Young branchlets glabrous,with scattered minute glands, which are persistent in the second year.Leaves (Plate 270, Fig. 9) about i^ to 2 inches long, and i to i£ inch broad,deltoid, with a broadly cuneate base and an acuminate apex; margin biserrate ;nerves five or six pairs; both surfaces glabrous and glandular; petiole f inch,glabrous, glandular.Fruiting-catkins (Plate 270, Fig. 9) cylindrical, about i inch long and J inchwide, directed towards the apex of the branchlet, on a slender, glabrous, glandular1 Scottish Land Names, 109 (1894). * Joyce, Irish Names of Places, i. 506 (1883).Betula 967stalk, about % inch long; scales glabrous, ciliate, with rounded, recurved laterallobes larger than the middle lobe.The twigs in winter are slender, shining, glabrous, covered with scatteredglands and waxy patches. Buds ovoid, acute, £ inch long, appressed to thebranchlet, with brown glabrous scales.The shoots, which spring, as in B. pubescens, from the roots, after a tree isfelled, are covered with numerous glandular warts and layers of wax, and bear large,incised, pubescent leaves. The seedlings of this species have simply serrate,pubescent leaves resembling those of seedlings of B. pubescens? but conspicuouslyglandular on both surfaces ; the stem is pubescent, but bears numerous glands.VARIETIES1. Var. dalecarlica, Linnaeus, f., Suppl. 4 16 (1781), Fern-leaved Birch. Leaves(Plate 270, Fig. 10) produced into a long acuminate apex, and with the margin cutinto pinnatifid serrated lobes. This variety has been found growing wild in theprovinces of Dalecarlia and Wermland in Sweden, and is occasionally seen incultivation, there being a good specimen in Kew Gardens. 2 It is sometimes knownin nurseries as var. laciniata.2. Shrubby forms, with leaves smaller than in the type, have been distinguishedas var. oycowiensis, Regel, in DC. Prod. xvi. 2, p. 164 (1868), found growing wildin Silesia and Galicia; and var. arbuscula, Winkler, observed by Fries in the wildstate in Dalecarlia.3. Var. japonica, Rehder, in Bailey, Cycl. Am. Hort. i . 159 (1900); Schneider,Laiibholzkunde, i . 112 (1904).Betula japonica, Siebold, Verh. Batav. Gen. xii. 25 (1830); Winkler, Betulacece, 7 8 (1904).Betula latifolia, Tausch, Fl. Ratisb. 7 51 (1838)5 Komarov, Act. Hort. Petrop. xxii. i. p. 38 (1904).Betula alba, Linnaeus, sub-species latifolia, Regel, in Bull. Soc. Nat. Mosc. xxxviii. 399 (1865), andin DC. Prod. xvi. 2, p. 165 (1868).Betula alba, Linnaeus, var. Tauschii, Shirai, in Tokyo Bot. Mag. viii. 319 (1894).In eastern Asia, in Manchuria, Saghalien, and Japan, the common birch isrepresented by a series of forms which have been grouped together by Winklerunder the name B. japonica, Siebold. In some respects they approach moreclosely B. verrucosa than B. pubescens, and are perhaps best treated as a geographicalvariety of the former species. Trees of Japanese origin cultivated in Kew Gardensshow the following characters : Young branchlets sparingly glandular, glabrous orwith a few scattered hairs. Leaves, 2 inches long, \\ inch wide, ovate, cuneate atthe base, acuminate at the apex; margin ciliate, sharply and simply serrate; nervesseven or eight pairs ; upper surface with scattered pubescence; lower surface lightgreen, glandular, glabrous except for slight pubescence on the midrib and nerves ;petiole, | inch, glabrous, glandular. Fruiting-catkins about i inch long, J inch wide,1 Cf. Watson, Compendium, 560 (1870); and Kerner, Nat. Hist. Plants, Eng. Tians. ii. 514 (1898).8 According to Schiibeler, p. 461, this beautiful variety was first found in 1767 at Lilla Ornas, about seven English milessouth of Falun in Sweden, when it was quite a small tree, 6 feet high, but grew to be in 1878, 64^ feet high, with a trunk6 feet 8 inches in girth. An excellent illustration of it is given by Schiibeler (Fig. 86) with outlines of the leaves (Fig. 87).(H. J. E.)

968 The Trees of Great Britain and Irelandcylindrical, on long, slender, minutely pubescent, glandular stalks; scales longstalked,veined, puberulous, ciliate, with the triangular central lobe slightly longerthan the broad, rounded, recurved, lateral lobes; nutlets with rather broad wings.This variety * is distinct in the larger number of nerves in the narrower, longer,slightly pubescent leaves, which are simply serrate in margin; and in the charactersof the fruit-scales. The bark of cultivated trees is more like that of B. Ermani thanthe common birch, as it is uniformly white in colour, with raised whitish lenticels,and scales off in transverse shreds. According to Sargent,2 it is a slender tree,attaining about 80 feet in height in Yezo.There are three trees of this variety in Kew Gardens, about 20 to 2 5 feet inheight, which were raised from Japanese seed sent by Sargent in 1891 under theerroneous name B. ulmifolia. A similar tree,8 2 5 feet high, cultivated at Kew asB. alba, var. latifolia, was obtained from Madrid in 1887. These trees are narrowlypyramidal in habit, and very ornamental on account of their beautiful white bark;and appear to be fast in growth and very thriving.4. Several varieties have arisen in cultivation, of which the most noteworthy are:Var. Youngi, Schneider, Young's weeping birch; and var. elegans, Schelle,Bonamy's 4 weeping birch. Both these forms have long, slender, pendulous branchlets ;and are usually grafted on stems 6 to 8 feet high, when they assume the habit ofthe weeping sophora. A fine specimen is growing in Smith's nursery at Worcester.Var. fastigiata, Schelle, is characterised by its upright branches, the treeresembling in its appearance a Lombardy poplar. According to a writer inWoods and Forests, this variety retains its foliage later in autumn than any otherform of the silver or common birch.Var. purpurea? Leaves purple, resembling in colour those of the purple beech,valuable for ornamental planting.DISTRIBUTIONThis species is widely distributed in Europe, and in northern and eastern Asia.The northern limit, beginning in Scotland, crosses Norway in lat. 64 , Swedenin lat. 65 , and ascends in Russian Lapland to Lake Ruanjarvi; and thence, crossingLake Onega, passes through the province of Vologda to Siberia, where its exactdistribution has not been made out. In eastern Asia, \wc.japonica is met with inManchuria, Saghalien, and Japan. The type occurs in the mountains of northChina, and was found near Lake Kokonor (lat. 37 50') by Przewalski; and itappears to be the common birch in the Altai and Ural mountains. It is not foundin Persia or Afghanistan, but occurs on the higher mountains of the Caucasus andin Armenia and Asia Minor. In Europe the southern limit extends from the1 The Japanese name for this variety is Shira-Kamba. z Forest Flora of Japan, 6 1 (1894).3 This tree has broadly ovate leaves, subcordate or rounded at the base, and larger than those of the trees raised fromJapanese seed, sent by Sargent; but in other respects is identical, and is probably also of Japanese origin.4 This originated in Bonamy's nursery at Toulouse, and is usually known in gardens as B. alba pendula elegans. Cf.Rev. ffort., 1 869, p. 135, fig. 33, and Card. Chron., 1869, p. 1278.6 Probably identical with var. atrosanguinea, stated by Schiibeler to have originated in France, and to be growing in theBotanic Garden at Christiania. (H. J. E.)Betula 969Rhodope mountains in Bulgaria, through Servia, Bosnia, and Istria to the VenetianAlps; and following the southern limit of the Alps in northern Italy, it extendsfrom the maritime Alps along the Apennines to Aspromonte in Calabria, crossingover to Sicily, where it reaches its extreme southerly point on Mount Etna inlat. 37 40'. It does not occur in Corsica or Sardinia. It is common in the forestsof the plains and lower mountains of France in the north, east, and west; buttowards the south only grows at high elevations, as in the mountains of Auvergneand in the Pyrenees. It grows in Portugal in the Sierra de Gerez; and in Spainthroughout the northern mountains, in Catalonia, Aragon, Navarre; and also in theSierra Guadarrama, the mountains of Toledo, and the Sierra de Gata.The largest forests of the species occur on the plains of the Baltic and centralprovinces of Russia, where it grows either pure or in mixture with aspen and greyalder, or with the common pine and spruce. Von Sivers 1 points out that the twospecies of birch occur in the Baltic provinces in different soils and situations.B. verrucosa grows on the glacial drifts, where it reaches large dimensions, andoften forms pure forests of clean, straight stems, which on the better class of soil,amongst spruce, often reach 100 to 130 feet in height.There are also extensive pure woods in the plains of northern and centralGermany; but farther south the tree is more at home in the mountains, as in theAlps and Carpathians, and only forms small woods, or grows scattered or in groupsamidst other trees.This species is most common in continental Europe on dry soils, thrivingbest in localities where the common pine does well, as in loamy sands with amoderately moist subsoil; but dwindles and ceases to grow on marshy groundor on undrained peat-mosses. It requires more light than the other species, andin woods of B. verrucosa the soil is usually covered with grass; the leaf mouldand moss, so common on the ground in woods of B. pubescens, being usually absent.(A. H.)CULTIVATION OF THE COMMON AND SILVER BIRCHAfter the oak, there is perhaps no tree which has been so generally attractiveto artists and lovers of the picturesque as the birch, which will grow almost anywhere, and is often looked on by English foresters and woodmen as a weed. Thisit may be on land fit to grow fine timber; where, however, it is not often soprevalent as on poor dry soils, or on wet, boggy land ; but when the question ofcovering waste land with timber of some sort at a low cost has to be considered,there is no tree that will do it so cheaply and so surely as the birch.It seeds very profusely, and the seed is so light that it spreads with greatrapidity, and germinates in places where hardly any other tree will live. It isabsolutely the hardiest tree we have, and though its economic value is low atpresent, yet probably it will, when our coal gets scarce and dear, be looked on asthe cheapest and best of firewood.IV1 Font. Verhalt. Bait. Prov. 1 8(1903).2 L

97 The Trees of Great Britain and IrelandIn north Russia it is the usual fuel for railway engines, and all overScandinavia is the principal firewood, but though our climate does not producethe tree as well as that of more northern regions, where the bark is almost theonly material used for covering the roofs of common buildings; yet, havingregard to the great beauty of the tree in landscape, it should be much moregenerally grown than it is. All this was well brought out by Loudon manyyears ago in his great work, and yet the birch remains a neglected tree. Butit has another virtue which must appeal to many in these days. It is of all treesthe one most distasteful to rabbits, and on my property is the only tree whichgrows up from self-sown seed, on land which on account of its poverty has beentreated as a rabbit warren.As a nurse for other trees on poor land, whether of a dry and rocky nature, orwet and peaty, the birch seems to me to have a greater value than most writers onforestry have admitted. For wherever the soil is naturally covered by self-sownbirch, or a fairly thick crop can be obtained by sowing, the land will be made fit forthe planting of more valuable trees, such as larch or Corsican pine, at a lower cost;and after providing shelter for smaller plants than could otherwise be used, it can,when the permanent crop has been established, be cut and sold at an age whenspruce or Scots pine of the same age would be worthless.A very successful instance of its use in this way has been described J byG. U. Macdonald, forester at Raith, Fifeshire, the object here having been to plantmoorland with spruce, in a locality where the late and early frosts were so severethat the spruce would hardly grow at first without some protection. 2On very dry oolite soil, birch is the only tree which reproduces itself naturallyamong long coarse grass, which it will, if thick enough, eventually suppress; and,though a large quantity of small birch wood may not be always saleable at asgood a price as at Raith (205. per ton for crate-wood), yet it is such excellentfirewood, even when quite small, that having regard to the low cost of its seed, I cansuggest no means whereby the desired result could be obtained so cheaply.If desired to establish a birch covert by sowing, I would advise the carefulselection of seed from trees naturally growing on land of similar character, becausethough foresters in this country have not yet realised the preference shown bythe rough twigged birch for dry rocky land, it is universally accepted as a fact inGermany, whilst for wet or boggy land the downy twigged birch is preferable.To raise birch from seed is not always easy; and whether it is better to sowin autumn directly the seed is ripe, or in spring, is a question which, after trying bothplans, I have not yet decided to my own satisfaction. But, as a rule, I wouldfollow nature and sow in autumn, not attempting to cover the seed with earth, butcovering with some fir boughs, fern, or leaves, until it began to germinate. So far1 Trans. A'cy. Scot. Art. Sac. xix. 287 (1906).2 It has been used with great success as a nurse for beech in some of the plantations, which were made by the Danishforester, Ulrich, near Copenhagen, and which were shown to the Royal English Arboricultural Society, in August 1908. Theobject here was to protect young beech trees from spring frosts, and afford shade during their youth. Bircbes were planted inlines about 8 feet apart, and the beech planted between the rows of birch eight or ten years later. When the beech aresufficiently tall, the birch are thinned and finally cut out entirely and used as firewood. This system seemed to me to be onewell worthy of adoption in England for other trees which require shade in youth.Betula 971as I have seen the germination is slow and irregular, and seems to depend a gooddeal on the maturity and age of the seed. For though birch seed is best sown thefirst year, I have had fair results from seed which had been kept a year in a bag,whilst I have sometimes obtained poor results from fresh seed sown in spring. Ialways sow the seed of exotic birches in pots or under glass, and prick out theseedlings at a year old.As a rule the seedlings grow fast, and must not be left long in the nursery, asthey do not transplant well when old, but there is much variation among the differentspecies ; and it seems that some of the American birches do not grow well orlive long in this country, unless grafted on the roots of the common birch.As a rule the birches seem to grow best in nature when unmixed with othertrees; and in some of the best birch woods I have seen in Norway, Japan, andAmerica are almost pure, but if mixed with conifers or other hardwoods, and notsuppressed by their shade, they often attain large dimensions.REMARKABLE TREES, COMMON AND SILVER BIRCHAmong the numerous birches which I have measured in England, I cannotmention any tree which is equal in height to some that I have seen inNorway and Russia, as the tallest do not exceed about 90 feet. Along the roadwhich passes through Savernake Forest from Marlborough to Andover there are agreat number of large and beautiful trees, planted as an avenue on both sides; butof their age I can obtain no record. Near the school is one of the finest (Plate 225),which measured in 1908 about 90 feet high by 8 feet in girth. Another near it wascovered with large witches' brooms; and a third, from which a large branch hadbeen recently torn off by the wind, was pouring out sap in such quantity that a poolof it had formed on the ground. In this park a birch, which was considered bythe woodman to be the largest, and which like the rest appeared to be D. pubescens,measured 77 feet by 10 feet 7 inches.Plate 256 shows the graceful habit of a fine birch in front of Lord Walsingham'shouse at Merton, Norfolk. Plate 257 shows a group of self-sown birches in Sherwoodforest, close to the Queen oak, described on page 322 of this work.At Dropmore there is a birch (B. pubescens) about 55 feet high, with an immensebole 21 feet in girth at a foot from the ground, and dividing a little way upinto three main stems, 9 feet 10 inches, 8 feet 2 inches, and 6 feet 3 inches in girthrespectively. One large limb has rested on the ground for many years, but doesnot seem to have rooted.In Windsor Park there are numerous fine birches, one of the handsomest ofwhich, growing by Prince Albert's chapel, was in 1904 71 feet by 8 feet 8 inches.At Longleat there is a fine tree of which Colonel Thynne has sent me a photograph,and which measured 76 feet by 9 feet 4 inches in 1906. At Barton there is a treefrom 80 to 85 feet high, with a clean stem about 50 feet long and 5 feet 7 inchesin girth, and drooping branches. This tree is still young and vigorous, and is oneof the finest we have seen anywhere.

97 2 The Trees of Great Britain and IrelandAt Beauport, Henry saw one dividing at 5 feet into five stems, and girthingbelow the fork 11 feet. At Arley Castle he measured one 76 feet by 8 feet i inch in1905. At Croome Court there is a tree 77 feet high, of which one large stemhas been broken off short. Mr. J. Smith mentions 1 a birch growing at EmbleyPark, Hants, in 1887, which was 85 feet by 6 feet 7 inches, with a bole of 25 feet.In Wales the birch, so far as I have seen, does not attain so large a size as inEngland and Scotland, the finest I know of being at Ogwenbank, near the entranceto the great slate quarry at Penrhyn. This, though only about 50 feet high, spreadsover an area 25 paces in diameter, and has two main stems which are 13 feet and 12feet 2 inches in girth respectively.In Scotland, Mr. Renwick considers a birch (B. verrucosa) at Auchendrane, inAyrshire, to be the finest in the west of Scotland. He gives an account of thistree, with a photograph, in Trans. Nat. Hist. Soc. Glasgow, vii. 262 (1905). It is 67feet in height, with a bole of 13 feet, girthing 10 feet 8 inches; and was planted,according to Miss Cathcart, by her mother in 1818, having been purchased fromBooth's nursery near Hamburg as a cut-leaved weeping birch.A still larger tree at Newton Don, near Kelso, which was cut down in 1901,measured, in 1893, 80 feet high, with a short bole, 13 feet in girth at i foot 7 inchesabove the ground, and dividing at 3 feet up into two main stems. Captain C. B.Balfour informs us that Jeffrey, in his History of Roxburghshire, describes it in 1859as being then 74 feet high and 14 feet in girth at the base.At Monzie there are some tall birches, drawn up by other trees, one measuring90 feet high by 8 feet in girth. At Blair Drummond there are several old birches,all with large boles, some with remarkable burrs, and one with low spreadingbranches layering. One of these measured 60 feet by 13 feet 10 inches ; and another70 feet high is 10 feet 8 inches in girth.In Darnaway Forest there are many fine birch trees on the banks of theFindhorn, one of which was stated by Sir Thomas Dick Lauder to girth 13 feet;but I am informed by Mr. D. Scott, forester to the Earl of Moray, that many ofthese have died, and those remaining are fast decaying. The largest measure 9 to10 feet in girth, and many of them contain 70 to 75 cubic feet of timber. Therough twigged birch predominates here, but does not as a rule assume a penduloushabit until it is of some age.At Gordon Castle there is a fine tree in the park, which in 1904 was 68 feet by9 feet; and at Murthly, in the drive from Dunkeld, I measured in 1906 a very talland slender birch (B. verrucosa) which was 89 feet high and only 3 feet 9 inches ingirth.In the Pass of Killiecrankie and many other Highland glens the birch growsfreely mixed with oak on the rocky slopes, and in the wide valley of the Spey thereare beautiful open woods of pure birch, covering a large extent of the gravellyflats and terraces which every traveller on the Highland Railway between Kingussieand Grantown must have admired. In the swampy flat at the head of Loch Morlich,in Glenmore, there is an open wood of curiously distorted, twisted, and stunted1 Trans. Scot. Arb. Soc. xi. 5 32 (1887).Betula 973birches, mixed with alders, in the hollow trunks of which the goosander occasionallybreeds, and the goldeneye duck may also sometimes do so (Plate 258).In no part of Great Britain do the birches assume the same tall, cleangrowth, or have the same smooth, silvery bark that they do in the forests of Scandinavia, where pure woods of birch are in some districts very prevalent on the lowerland, whilst on the fjelds and mountains it ascends as a scrubby and stunted tree toa greater elevation than any other. Schubeler 1 describes and figures someinstances of abnormal growths in birch. His figure 88 shows a fallen trunk fromwhich six healthy-looking young trees are growing in a line, and I have a similarthough less striking instance in my own woods. The tallest birch he knew inNorway was at Drobak in Sceterdalen, 30 kilometres south of Christiania, andmeasured 100 feet by 5 feet 9 inches. Another at Sondre Tveten, in Eidanger,was 79 feet by 11 feet. He figures a remarkable tree at Dunserud, in Eker,which divides into six large trunks, and measures 75 feet in height. Perhaps themost shapely and beautiful of those which he figures, are a tree at the farm of Hohls,in south Trondhjem Amt, measuring 80 feet by 11 feet, and another at Gravrok,18 kilometres south of Trondhjem, which measured 80 feet by 16 feet. It seemsfrom these particulars that the birch is the largest deciduous tree in Norway,and attains greater dimensions than in this country ; but though I have spentseveral months in the forests of north and south Trondhjem in the pursuit of myfavourite quarry, the elk, I never saw such trees as those above mentioned, andbelieve that they all grow in the neighbourhood of farms on unusually fertile soil.TIMBER AND BARKThe timber of the birch can hardly be said to have any general recognisedvalue in England, though in some districts it can be sold to coopers, chairmakers,and clogmakers ; 2 in others, especially since charcoal burning has ceased in mostplaces to be a profitable industry, it can only be looked on as firewood. It is soperishable in contact with the soil, that it is of no use for fencing unless creosoted;and though in former times,3 according to Sang, the Highlanders made everythingthat they used of it rafters, ploughs, harrows, carts, and fences yet now it wouldonly be used as a makeshift, when other wood could not be had. I have seen largeold burry birches, which when cut into boards, were fit for small cabinet panels; butthe wood twists a good deal in drying, and is usually inferior, in grain, texture,colour, and figure to the wood of several species of American birch which can beimported in larger size and at a low price. As underwood it has an uncertainvalue for making brooms and tool handles and is also used for bobbins.In Sweden and Russia the burrs found on the trunks of this tree are convertedinto many ornamental articles of great beauty. Beer - pots carved out of theseburrs, and hooped with wood or silver, are often heirlooms in Scandinavian houses;1 Viridarium Norvegicum, 469.2 Birch clog soles are used in the Yorkshire manufacturing towns, whilst in Lancashire alder is preferred.3 Pennant, Tour in Scotland, 1 12 (1772), says that wine was extracted from the live tree.

974 The Trees of Great Britain and Irelandand as the wood takes a high polish, it is highly valued in Siberia for work-boxes,cigarette cases, and other small fancy articles.In northern Sweden and Russia the wood is sometimes found full of undulations, which make it very ornamental for furniture, and some bedroom furnituremade by the Nordiska Kompaniet (Lundberg and Laja) of Stockholm, was almostequal in beauty to satinwood. I have also seen it used as veneer with the besteffect for decorating cabins in steamers built in Denmark and Sweden. This isknown in Sweden as " Flammig bjork." Another curious form of birch wood isthat known in Finland and Sweden as " Masur." I was informed by Mr. Jacobssen,Swedish Vice-Consul at Abo, that this variety in Finland is only found incertain places, Karku, Tyrois, and Kalvola. A number of logs which I saw inthe works of the Finska Colorit Aktiebolag at Abo were covered on the outside withsmall pitted depressions, somewhat similar to those which produce bird's-eye maple,and when cut into veneer, are dyed of various colours, of which French grey seemedto me the most effective; and made up into furniture which commands a high price.At St. Petersburg this form of birchwood is known as " Karelsky," beingsupposed to come from the Karelian peninsula; and is largely used both in thesolid and as veneer for furniture making. Though not so elegant as the wavedform, or as the bird's-eye maple which it somewhat resembles, it is very quaint andstriking in appearance, and can be imported at very reasonable prices.The bark, 1 when taken off in sheets, is used in Scandinavia for covering theroofs of houses, and remains for many years undecayed between the inner boardingand the outer sod of turf. A strong smelling oil, obtained by destructive distillationfrom birch wood, is, when mixed with alcohol and rubbed on the skin, the bestprotection I know of against the swarms of midges and mosquitoes which makelife almost unbearable in the short summer of the far north. This oil is used asa preservative, and gives the fragrant odour to Russia leather. Birch bark has noequal for lighting fires, and in the dripping forests of the north I have often hadgood reason to value it when nothing else would start a fire. (H. J. E.)BETULA DAVURICABetula davurica, Pallas, Fl. Ross, i . 60, t. 39 (1784); Winkler, Betulacece, 86 (1904).Betula Maximowiczii, Ruprecht, in Bull. Phys. Math. Acad. Petersb. xv. 139 (1856) (not Regel).Betula Maackii, Ruprecht, in Bull. Phys. Math. Acad. Petersb. xvi. 380 (1857).A tree, attaining 60 or 70 feet in height. Bark purplish brown, separating insmall, papery scales, which remain attached, curled and ragged, to the trunk, givingthe tree a peculiar appearance. Young branchlets glandular, covered with a minuteerect pubescence, interspersed with a few long hairs. Leaves, about 3 inches longand i £ to 2 inches wide, narrowly ovate or ovate-rhombic, cuneate at the base,Betula 975acute or acuminate at the apex; margin ciliate, coarsely and irregularly serrate ;nerves, six to eight pairs; upper surface at first pubescent on the midrib and nerves,ultimately glabrescent; lower surface glandular, with scattered pubescence on themidrib and nerves ; petiole, £ inch, slightly pilose.Fruiting catkins, f to i inch long, ovoid-cylindrical, acute at the apex; scalesglabrous with scattered glands on the margin and outer surface, middle lobetriangular, lateral lobes broad, rounded and spreading; nutlet obovate, with narrowwings, broadest in their upper part.This species, which is readily distinguished by its peculiar bark, is widelyspread throughout Amurland, Manchuria, Korea, and north China. According toKomarov, 1 it grows throughout the whole of Manchuria, in the drier parts of thevalleys and in open places on the mountains, on rocky or sandy soil, occasionallyforming small woods, but is never seen in the dense virgin forests.It is extremely rare in cultivation, the only specimen which we have seen beinga small tree in Kew Gardens, about 15 feet high, which was raised from seed sentby Bretschneider from Peking in 1882. (A. H.)BETULA CORYLIFOLIABetula corylifolia, Regel and Maximowicz, in Bull. Soc. Nat. Mosc. xxxviii. 417, t. 8 (1865); Regel,in DC. Prod. xvi. 2, p. 178 (1868); Winkler, Bctulacece, 5 9, fig. 17 (1904).A tree, the dimensions of which are not stated. Young branchlets slightlypubescent. Leaves (Plate 270, Fig. 14), 2\ to 3 inches long, about 2 incheswide, oval; rounded or truncate (rarely cuneate) at the base, acute at the apex;coarsely serrate; nerves, twelve to fourteen pairs, impressed above and veryprominent beneath ; upper surface pilose on the midrib, elsewhere glabrous; lowersurface pale in colour, with conspicuous long silky hairs on the midrib and nerves,elsewhere glabrescent; petiole, | inch, at first pilose, later glabrescent. Fruitingcatkins (Plate 270, Fig. 14), i^ to 2 inches long, f inch wide, cylindrical, oftencurved; scales large, slightly pubescent, ciliate in margin, deeply three-lobed,lobes linear-oblong, the middle one about twice as long as the lateral lobes; seedswith very narrow wings.This remarkable species, 2 peculiar in the shape of the leaf and in the stout, longfruiting catkins, was found on the high mountains of the provinces of Senano andNambu in the main island of Japan, by Tschonoski. Very little is known about itin the wild state, and it has never apparently been introduced into cultivation inEurope. (A. H.)1 Flora Manshtiria, ii. 49 (1903).2 The Japanese name of this species, according to Matsumura and Goto, is Urajiro-kamtia.1 Pyrobetulin, obtained by sublimation from the outer bark of birch, is used for depositing films on glass, about to beengraved, and for covering lint with an antiseptic layer. Cf. Wheeler, in Phann. Journ. ix. 494 (1899).

976 The Trees of Great Britain and IrelandBETULA MAXIMOWICZIIBetula Maximmmczii^ Regel, in Butt. Soc. Nat. Mosc. xxxviii. 418 (1865); Winkler, Betulacece, 8 9(1904).Betula Maximowicziana, Regel, in DC. Prod. xvi. 2, p. 180 (1868); Shirasawa, Icon. Ess. Forest.Japan, text 45, t. 23, ff. 1-8 (1900); Mayr, Fremdliind. Wald- u. Parkbiiume, 449 (1906).A tree, attaining in Japan, according to Mayr, 100 feet in height. Bark grey,smooth, peeling off in thin, papery strips. Young branchlets, with scattered glands ;glabrous, except for slight pubescence above the insertions of the leaves. Leaves(Plate 269, Fig. 6) about 6 inches long and 4 inches wide, broadly ovate, deeplyand narrowly cordate at the base, acuminate at the apex; margin non-ciliate, biserrate; nerves, ten to twelve pairs, each ending in a long-pointed serration ; uppersurface at first pubescent with erect hairs, later glabrescent; lower surface withscattered pubescence throughout, or glabrous, except for pubescence on the midriband nerves, gland - dotted; petiole an inch or more in length, pubescent orglabrous.Fruiting catkins (Plate 269, Fig. 6), two to four in a raceme, about z\ incheslong, and nearly \ inch in diameter; scales glabrous, shortly three-lobed, the laterallobes spreading and shorter than the middle lobe; nutlets very small, withbroad wings.A variety of the species is in cultivation in Kew Gardens, distinguished byhaving smaller leaves, with more shortly pointed serrations, and with their undersurface (as well as the young branchlets and petioles) covered with long, softpubescence.This species is readily distinguished by its large leaves, deeply and narrowlycordate at the base. In winter the twigs are stout, shining, yellowish, nearlyglabrous; buds about f inch long, appressed to the branchlet, curved laterally andending in a sharp beak, with glabrous scales.B. Maximowiczii occurs in Japan, in the central chain of Hondo, but is morecommon in Yezo,2 where, according to Sargent, 3 it is a shapely tree, 80 or 90 feet inheight, with a trunk 2 or 3 feet in diameter, covered with pale, smooth, orangecolouredbark. Towards the base of old trees the bark becomes thick and ashygrey,separating into long, narrow scales.The largest that Elwes saw in Japan were growing in a mixed forest of maple,poplar, ash, spruce, and silver fir, on volcanic soil, at about 3000 feet elevation, nearthe Crater Lake, Shikotsu, in Yezo; and one measured 90 feet high by 9 feet9 inches in girth. The Japanese name 4 of this species is Udai-kamba.1 This species, with B. luminifera, Winkler, B. Bcetimkeri, Winkler, both natives ot central China, and B. alnoides,Buchanan-Hamilton, distributed throughout the Himalayas and in central and southern China, constitute the section Seta/aster,distinguished by elongated fruiting catkins and broad-winged nutlets.2 Mayr, op. cit., plate 31, gives a picture of this tree growing in a forest in Yezo.3 Forest Flora of Japan, 62 ( 1894).4 According to Matsumura, in Shoktibutsu Mci-I, 48 (1895). The same name is given in Goto's Forestry of Japan.Betula 977It was introduced into cultivation in England by J. H. Veitch, 1 who sent homeseeds from Yezo in 1888; and a large number of seedlings were raised at theArnold Arboretum, in 1893, from seed received by Sargent 2 from the forestryofficers of Yezo.Sargent has spoken highly of the beauty and value of this tree, which, whereverwe have seen it in this country, is thriving. It is one of the most beautiful of theyoung trees in Messrs. Walpole's lovely garden at Mount Usher, County Wicklow.A tree at Kew, raised from seed obtained in 1893 from the Arnold Arboretum,was 25 feet high and 13 inches in girth in 1907. At Tortworth, a tree, probably ofthe same age, 30 feet high and 17 inches in girth, is growing vigorously. AtGrayswood, Haslemere, a tree obtained as a small plant from Lemoine in 1894, was29 feet by 21 inches in 1906, and is very healthy.This species is common on the Continent in botanical gardens, and is veryhardy, having borne without injury the severe winter climate of Grafrath, nearMunich; and on this account, and because of its rapid growth, it is recommended byMayr as worth cultivating as a forest tree.Shirasawa says that the wood is rather hard, showing no difference in thecolour of the sap and heart wood, and is used in Japan for house-building. Thefishermen of Yezo make torches out of the bark, as it takes fire easily, even whenwet - (A. H.)BETULA ERMANIBetula Ermani, Chamisso, in Linnaa, vi. 537, t. vi. f. 8 (1831); Erman, Reise, t. 17 (1835);Komarov, Flora Manshnrice, ii. 49 (1903) ; Winkler, Betiilacece, 6 6 (1904).A tree,3 attaining about 100 feet in height in Manchuria. Bark creamy-white,with raised whitish lenticels, and peeling off in irregular shreds. Young branchletsglabrous, except for a few hairs above the insertions of the leaves, and covered withnumerous glands which roughen the shoot in the second year.Leaves (Plate 270, Fig. 12), about 3 inches long, 2 inches broad, ovate, with abroad truncate or slightly cordate base, acuminate at the apex; margin slightlyciliate at first, coarsely and irregularly serrate; nerves, ten to twelve pairs; uppersurface with scattered hairs; lower surface glandular, glabrous except for slightpubescence in the axils and on the midrib and nerves; petiole, \ inch, glabrescent,glandular.Fruiting catkins (Plate 270, Fig. 12), ovoid-oblong, about i inch long and| inch in diameter, sessile or shortly stalked ; scales glabrous on the surface, ciliateand glandular in margin, with linear-oblong lobes, the middle lobe longer than thetwo lateral divergent lobes; nutlets with narrow wings, broadest above.1 Hortus Veitchii, 357(1906). 2 Forest Flora of Japan, 62 ( 1894).3 According to Matsumura, Shokutulsu Mei-l, 47 (1895), var- nipponica of this species is known as Take-kamba, whilethe type is called Ezo-no-take-kamba, L e. B. Ermani of Yezo.IV 2 M

978 The Trees of Great Britain and IrelandVar. nipponica? Maximowicz, MM. Biol. xii. 923 (1888). This differs from thetype in the branchlets not being so densely glandular, and in the longer, narrowercones, which are cylindrical, i^ inch long, and \ inch in diameter. This variety wasfound by Maximowicz in the Nikko mountains, and is probably the form of thespecies occurring in the main island of Japan.B. Ermani is widely distributed in eastern Asia, occurring in Kamtschatka,Manchuria, Korea, Saghalien, Kurile Isles, and Japan. According to Komarov,2 itgrows in Manchuria, near the sea-coast, from the river Amur to St. Olga Bay, inthe Sichote-Alin mountains, in the north-eastern part of the province of Mukden,and in the Korean main range; and is a native of mountain forests at elevationsbetween 2000 and 7000 feet. Komarov informs us that it is the largest of thebirches, which he found in Manchuria, attaining a height of 60 to 100 feet.According to Sargent, 3 it is the most common birch in the high mountains ofthe main island of Japan, where it is scattered through the coniferous forests at4000 to 6000 feet, and is very conspicuous from the white bark of the trunk, and thebright, orange-coloured bark of the principal branches. It appears to be the birchfigured and described by Shirasawa, 4 as Betula alba, var. communis (not Regel),which he states to be a tree of vigorous growth, attaining 70 feet in height and3 feet in diameter, occurring in the central chain of Honshu, especially at Shimotsuke,in Nikko, Musachi, in Chichibu, and Kiso, in Shinano. At Kiso the barkis used as material for writing on, and for envelopes; and also for torches,as it contains a considerable amount of resin. Elwes saw a birch, which he believesto be B. Ermani, growing in great abundance in the forest north of Asahigawa,and also on the volcanic cone near Lake Shikotsu, in Hokkaido. Here it seemedto be commoner than, but not so large, a tree as B. Maximowiczii.This species was introduced into cultivation through the agency of theSt. Petersburg Botanic Garden; and is not uncommon in gardens both on theContinent and in England, where it is often wrongly named B. ulmifolia, B. corylifolia,etc. At Kew there are small trees, 20 to 25 feet in height, both of the typicalform and of var. nipponica. The former 5 is the first of the birches to come intoleaf, the foliage often being fully developed at the end of March; and, in consequence, the trees are usually much injured by spring frosts. The largest tree wehave seen of this species is in the arboretum at Westonbirt, and when measured byElwes in 1908, was 51 feet by 2^ feet. There is a good specimen apparently ofthe typical form at Benmore, in Argyleshire, which is grafted near the ground,and has yellowish scaly bark like that of B. lutea. I n 1907 Elwes found it to beabout 40 feet by 3 feet. Var. nipponica is later in leafing, and is not usually injuredby frost. Two trees of this variety at Grayswood, Haslemere, are about 30 feetin height. (A. H.)1 This variety appears to be identical with a specimen in the Kew herbarium collected by Tschonoski on "highmountains not far from Fuji-yama," which is labelled B. Bhojpattra, var. subcordata, Regel, in DC. Prod. xvi. 2, p.177 (1868). 2 Flora Manshnria, ii. 50 (1903).3 Forest Flora of Japan, 62 (1894).4 Icon. Essences Forest. Japan, text 42, t. 21 ff. 1-15 (1900). Betula alba, var. commimis, Regel, is B. pafyrifera,an American species, which certainly does not occur wild in Japan.6 The trees of the typical form were raised from seed sent by the Arnold Arboretum in 1893.Betula 979BETULA ULMIFOLIABetula ulmifolia, Siebold et Zuccarini, in Abh. Bayer. Akad. Wiss. iv. 3, 228 (1846); Winkler,Betulacece, 62 ( 1904) (in part); Schneider, Laubholskunde, i. 101 (1904).A tree, attaining in Japan 70 feet in height and 8 feet in girth. Bark describedas greyish-brown, smooth, shining, not separating into thin layers, and resemblingthat of Prunuspseudocerasus. Young branchlets covered with a white, short, somewhat appressed pubescence. Leaves, about 3 inches long, i£ inch wide, narrowlyovate or ovate-oblong, unequally cordate * at the base, acuminate at the apex ;margin ciliate, bi-serrate, with falcate serrations; nerves twelve to fourteen pairs;upper surface with appressed, long, brownish hairs both on the midrib and nerves,and in bands between the nerves; lower surface similarly pilose on the midrib andnerves, gland-dotted and glabrous between the nerves; petiole % inch, pilose.Fruiting catkins, on short pilose peduncles, about f inch long and £ inch indiameter, ovoid; scales pubescent, ciliate, strongly veined, with the central lobeoblong and obtuse, nearly twice as long as the ovate rounded lateral lobes; nutletswith narrow wings.This species is extremely rare in cultivation, the only specimen which we haveseen being a tree, about 8 feet high, in Kew Gardens, which was raised from seed,received under the name B. grossa? S . et Z., from Tokyo in 1896. It is identicalwith the type specimen of B. ulmifolia, S. et Z., preserved in the Munich herbarium,with which we have compared it.This species, together with B. grossa, S . et Z., B. carpinifolia, S . et Z., bothnatives of Japan, and B. costata, Trautvetter, a native of Manchuria, are closelyallied; and our knowledge of their exact relationship and distribution is veryimperfect. It is possible that B. grossa and B. carpinifolia are varieties orhybrids of B. rtlmifolia, while B. costata is the continental geographical form of thesame species.Shirasawa,3 whose figures of B. grossa and B. ulmifolia do not in either caseexactly agree with the type specimen of the latter species, says that both these speciesare spread throughout the central chain of Hondo, and occur also in Kiushiu andShikoku. B. ttlmifolia, which is the representative in Japan of the AmericanB. lutea, differs much in bark and other characters from B. Ermani, with which ithas been confused. (A. H.)I saw a very fine birch which my guide and companion, Mr. Mochizuki of theJapanese Forest Service, called B. grossa, growing in the forest of central Japan, atOngawa, about 3000 feet above the sea; and measured specimens 80 to 90 feet high.1 The leaves on the lower part of the branchlel and on the thoil shoots are markedly cordate; those on the upper panof the branchlet are usually truncate or rounded at the base.2 Similarly a dried specimen at Kew, collected in the Etchu province on Mt. Tateyama, and labelled B. grossa by theTokyo University Science College, is identical with B. ulmifolia. It bears the Japanese name Yoguso-ininebari.3 Icon. Ess. Forest. Japan, text, 42, 43, t. 22 (1900).

980 The Trees of Great Britain and IrelandThe timber, of which I brought home a specimen now at Kew, is a hard wood of abright pinkish brown colour, and is used for flooring. It seems at least as goodas the best American birch timber. It was, however, very difficult to identify thespecies of birch, of which no less than five 1 are said to be found in this district, andthe foresters of Japan were not themselves sure of their scientific names.(H. J. E.)BETULA LUMINIFERABetula luminifera, Winkler, Betnlacece, 9 1, fig. 23 (1904).A tree, the dimensions of which are not known. Young branchlets coveredwith dense, erect, pale pubescence, non-glandular. Leaves (Plate 270, Fig. 16),about 5 inches long and 3^ inches wide, broadly ovate, sub-cordate or truncate atthe broad base, acuminate at the apex; margin ciliate, irregularly serrate, theserrations ending in cartilaginous points; nerves ten to fourteen pairs; coveredmore or less on both surfaces with white, short pubescence; petiole f inch,pubescent. Fruiting catkins (Plate 270, Fig. 16), solitary, cylindrical, elongated,about 3 inches long and | inch in diameter ; on a peduncle f inch long; scaleslanceolate, auricled on each side a little below the middle ; nutlets pubescent, withbroad wings.This species, which is the representative in central China of B. Maximowicziiof Japan, was discovered by Pere Farges in the north-eastern mountains ofSzechwan, and was introduced into cultivation in 1901 by E. H. Wilson, who sentseeds from the same locality. A young tree in Veitch's nursery at Coombe Woodis now 16 feet high at seven years old, and is very flourishing. This species isremarkable for its fine foliage, and is worthy of cultivation as an ornamental tree.(A. H.)BETULA UTILIS, HIMALAYAN BIRCHBetula utilis, Don, Prod. Fl. Nepal. 5 8 (1825); Hooker, Fl. Brit. India, v. 599 (1888); Gamble,Indian Timbers, 668 (1902); Winkler, Betulacece, 6 1 (1904); Schneider, Laubholzkunde, 1 02(1904); Brandis, Indian Trees, 6 22 (1906).Betula Bhojpattra, Wallich, PL As. Rar. ii. 7 (1832); Brandis, Forest Flora, N.-W. India, 457(1874)-Betula Jacquenwntii, Spach, in Ann. Sf. Nat. ser. 2, xv. 189 (1841).A tree, attaining in the Himalayas about 60 feet in height, but becoming ashrub at high elevations. Bark on young trees thin, smooth, brownish red, with1 Matsumura, in SlioJtubutsu Mei-I. 47 (1895), enumerates eight distinct species of birch as occurring in Japaii; but ofthese, B. utilis, as explained in a note under our account of that species, and B. grossa, mentioned above, are doubtful.There remain six distinct species, undoubted natives of Japan, viz. : B. Maximowic&ii, B. corylifolia, B. Ermani, B.globispica, B. ulmifolia, and B. •verrucosa, var. japonica; all of which are referred to in this account of the genus Betula, andunder each species is given its native name.Betula 981darker coloured horizontal lenticels, peeling off in transverse rolls; on older treesdarkened and thickened at the base. Young branchlets, non-glandular, coveredwith a dense, greyish tomentum; older branchlets smooth, glabrous. Leaves(Plate 269, Fig. 7) coriaceous, about 3^ inches long and i\ inches wide, oval orovate-oblong, rounded at the base, acuminate at the apex; margin slightly ciliate,irregularly serrate ; nerves nine to twelve pairs; upper surface shining, dark green,with scattered pubescence; lower surface yellowish green, glandular, glabrousbetween the nerves, which are slightly pubescent, and with dense axil-tufts ofpubescence ; petiole f inch, tomentose.Fruiting catkins (Plate 269, Fig. 7), 1 cylindrical, i^ inch long, \ inch in diameter,on tomentose peduncles, variable in length; scales with glabrous, ciliate, spatulatelobes, the central lobe dilated above and obtuse at the apex, and often trifid, abouttwice as long as the erect or slightly divergent lateral lobes; nutlets with narrowwings.VARIETIESIn addition to the type, described above, which occurs in the Himalayas andChina, the following varieties can be recognised :1. Var. sinensis, Franchet, Journ. de Bot. xiii. 207 (1899).Betula albo-sinensis, Burkill, in Journ. Linn. Soc. (A'ot.) xxvi. 497 (1899).Leaves glabrescent beneath. Fruit-scales glabrous, not ciliate ; nutlets smallerthan in the type. Discovered in north-eastern Szechwan by Pere Farges. Not yetintroduced.2. Var. Prattii, Burkill, Journ. Linn. Soc. (Bot.) xxvi. 499 (1899). Leavesmore pubescent than in the type, the pubescence extending over the whole undersurface, and very dense in the axils and along the midrib. Fruit-scales stronglyciliate, with spatulate lobes, the lateral lobes spreading and not erect. This varietyoccurs in western Szechwan, at high elevations (13,500 feet), and has not yet beenintroduced.3. Var. Jacquemontii, Regel, in DC. Prod. xvi. 2, p. 177 (1868).Betula Jacquemontii, Spach, in Ann. Sc. Nat. ser. 2, xv. 189 (1841); Cambessedes, in Jacquemont,Voyage dans I'lnde, Botanique, 1 57, t. 158 (1844); Regel, op. tit. 1 78 (1868).A tree, with white bark, marked by brownish horizontal lenticels, and peelingoff in transverse strips. Young branchlets slightly glandular, and covered with adense, erect, short pubescence. Leaves (Plate 270, Fig. 15), about 2| inches long,and if inch broad, ovate, rounded or slightly cuneate at the base, acuminate at theapex, bi - serrate; nerves seven to nine pairs ; upper surface with scatteredpubescence or glabrescent; lower surface gland - dotted and glabrous except forlong hairs on the midrib and nerves; petiole f inch, glabrescent, glandular.Fruiting catkins (Plate 270, Fig. 15), i£ inch long, \ inch wide, cylindrical, onlong pubescent stalks; scales glabrous, ciliate, with an elongated linear central lobe,1 In this figure the middle lobe of the scale is represented short and trifid at the apex, as is occasionally the case ; but asa rule it is more elongated, and broadened and rounded at the apex.

982 The Trees of Great Britain and Irelandacute at the apex, and two or three times as long as the divergent short roundedlateral lobes ; nutlets with narrow wings.This variety is represented in Kew Gardens by two trees obtained from St.Petersburg in 1891 and 1894, and 25 feet and 20 feet high respectively. It is verydistinct in appearance from the typical form of the species, having white bark;smaller, few - nerved, thinner leaves; and different catkins. Moreover, thebranchlets and fruiting peduncles are shortly pubescent in the variety, andtomentose in the type.This variety, 1 judging from the material in the Kew Herbarium, is common inthe Himalayas, and probably constitutes a distinct species, which a careful study inthe field may show to occupy a different area of distribution from that of typicalB. titilis, which is so readily recognisable by its reddish bark and othercharacters.DISTRIBUTIONB. utilis* is widely distributed in the Himalayas and in China. It occurs inWest Tibet, and in the Himalayas from the Kurram valley and Kashmir, to Sikkimand Bhotan, at altitudes usually ranging from 10,000 to 14,000 feet, but descending inthe north-west to 7000 feet. It is often gregarious at the upper limit of arborescentvegetation, where it is commonly associated with Rhododendron campanulatum.According to Brandis, it attains 50 to 60 feet in height, and sometimes 10 or 12 feetin girth. Gamble's account of the bark includes that of the type and of var.Jacquemontii\ as he describes it as smooth, shining, reddish white or white, theouter bark consisting of numerous distinct, thin, papery layers, peeling off in broadhorizontal rolls; the thicker lower part of the bole becoming rough and dark as inthe European birch. He states that the growth is slow, with an average of fifteenrings per inch of radius. The wood is extensively used in the inner arid Himalayanregion for building purposes; it is elastic, does not warp, and seasons well. Thebark is the most valuable part of the tree, and is used for paper, umbrellas, hookahtubes,and roofs of houses.In China it is a moderate-sized tree, growing only at high elevations, between8000 and 13,500 feet, in the provinces of Szechwan, Hupeh, and Kansu. It wasseen in the latter province by Przewalski, 3 who describes the bark as reddish,peeling off and hanging from the tree in long festoons.This species is very rarely seen in cultivation, and the typical form from theHimalayas, like most of the broad-leaved trees from that region, for some unexplained reason, does not appear to have succeeded in this country. At Grays-1 Shirai, in Tokyo Bet. Mag. viii. 320, ff. 23, 24 (1894), states that this variety occurs in Japan; but the plant figuredby him seems to be identical with a specimen gathered by Maximowicz in the province of Shinano, and labelled B.BhojpaUra, var. typica, Regel, which is certainly not that species, and appears to be B. uliiiifolia, S . et Z. Shirai's accountof the bark of this tree, as not heing papery, but greyish brown, smooth, cracking, and falling off in patches, confirms thisidentification.2 There is no evidence that B. ati/is occurs in Japan, where it is represented by the closely allied species B. Ermam.Shirasawa, in Icon. Ess. Forest. Jaf on, text 44, t. 23, ff. 9-22 (1900), figures a tree as B. Bhojpattra which is not thisspecies (B. atilis), as is confirmed by his account of the bark, as being hard, compact, and falling off in scaly facets.B. Ermani, var. nipponica, has also been considered erroneously to be a form of B. utilis.3 Cf. Bretschncidcr, Europ. Bat. Dinoveries, 987 (1898).Betula 983wood, Haslemere, a tree planted in 1882, which was obtained from a nurseryat Newry, was 30 feet high and 2 feet i inch in girth in 1906.There are three trees about 20 feet high in the Botanic Garden of TrinityCollege, Dublin, which, according to Burbidge, 1 were raised from seed sent bySir Joseph Hooker in 1881. At Castlewellan there are some young trees, about8.feet high, which were obtained by grafting branches of the Dublin trees on thecommon birch. fj± H )BETULA PAPYRIFERA, PAPER BIRCH, CANOE BIRCHBetula papyrifera, Marshall, Arbust. Am. 1 9 (1785); Sargent, Silva N. Amer. ix. 57, t. 451 (1896),and Trees N. Amer. 2 02 (1905); Winkler, Betulacece, 8 3 (1904).Betula /en/a, Wangenheim, Nordam. Holz. 45 (1787) (not Linnjeus).Betulapapyracea, Aiton, Hurt. Kew. iii. 337 (1789); Loudon, Arb. et Frut. Brit. iii. 1708 (1838).Betula grandis, Schrader, Ind. Hort. Bot. Goett. 2 (1833).Betula lalifolm, Tausch, in Flora, xxi. 2 p. 751 (1838).Betula alba, Linnseus, var. papyrifera, Spach, in Ann. Sc. Nat. ser. 2, xv. 188 (1841).Betula cordifolia, Regel, Monog. Betulaeece, 86 (1861).Betula macrophylla, Hort., ex. Schneider, Laubholzkunde, i. 115 (1904).A tree, usually attaining in America, in its typical form, 60 or 70 feet in heightand 2 or 3 feet in diameter, the variety found on the lower Fraser River in BritishColumbia being usually much larger in size. Bark thin, smooth, creamy-white,marked with long, narrow, horizontal lenticels, and separating into thin paperylayers; becoming on old trunks near the base £ inch thick, dull brown or blackish,fissured, and scaly. Young branchlets, with scattered long hairs, mostly fallingoff in summer; in the second year dark brown and glabrous.Leaves (Plate 269, Fig. 5), 2 to 3 inches long, i^ to 2 inches wide, ovate ; roundedor slightly cordate at the base, acuminate at the apex ; margin ciliate and irregularlybi-serrate; nerves six to eight pairs ; upper surface dull green, slightly pilose on thenerves; lower surface paler, with numerous minute brown glands, usually glabrousexcept for dense axil-tufts of pubescence and a few long hairs on the midrib andnerves, occasionally minutely pubescent between the nerves; petiole at first pilose,ultimately glabrescent.Fruiting catkins (Plate 269, Fig. 5), cylindrical, about i£ inch long, \ inch thick,hanging on slender stalks; scales pubescent or glabrous, ciliate, with the middlelobe longer than broad ; lateral lobes rounded, erect, spreading or recurved. Nutletwith broad wings.In winter the twigs usually show a few scattered long hairs; buds, | inch long,appressed to the branchlet, ovoid, acute, with glabrous, ciliate scales, glisteningwith resin.VARIETIES AND HYBRIDSThis species, spread over a vast territory in North America, is very variable inthe wild state; and the forms occurring on the Rocky Mountains and in the Pacific1 Proc. Roy. Hort. Sec. 1 901, p. xxxviii.

984 The Trees of Great Britain and IrelandCoast region have not yet been fully studied in the field. The following varietiesare notworthy:1. Var. cordifolia, Regel, in Bull. Soc. Nat. Mosc. xxxviii. 401 (1865); Fernald,in Rhodora, in. 173 (iQO 1 ) ', Sargent, Silva N. Amer. xiv. 55, t. 724 (1902).Leaves distinctly cordate at the base, smaller than in the type. An alpinetree, moderate in size, scarcely exceeding 40 feet in height, which occurs on MountKatahdin in Maine, on the White Mountains in New Hampshire, and in thenorthern Rocky Mountain region.2. Var. kenaica.Betula kenaica, Evans, Bot. Gazette, xxvii. 481 (1899); Sargent, Silva N. Amer. xiv. 53, t. 723(1902), and Trees N. Amer. 205 (1905).A tree in Kew Gardens, about 20 feet high, obtained from Dieck in 1891, andsaid to be a native of Alaska, has been identified by Sargent with B. kenaica.Judging from this specimen and the description given by Sargent, this speciesis only a small-leaved variety of B. papyri/era, from which it cannot be separatedby any characters of importance. The branchlets are minutely pubescent andslightly glandular. The leaves are about 2 inches long, with five to seven pairsof nerves. The fruiting catkins are smaller than in the type, about an inch long,with glabrous ciliate scales, the middle lobe of which is narrow, oblong, not muchlonger than the rounded broad lateral lobes. The bark is like that of ordinaryB. papyri/era, but the white colour is slightly tinged with orange.Betula kenaica was discovered in 1897 by Dr. Evans in Alaska, in the KenaiPeninsula, near Cook Inlet; and was found on Kodiak Island by Coville in 1899.It is described as being a small tree, only reaching 40 feet in height. It is probable Tthat the variety extends south from Alaska through British Columbia; and is aform with small leaves growing on poor soil and in mountainous regions, while var.Lyalliana, with large leaves, occurs at nearly sea-level in rich alluvial soil.3. Var. Lyalliana? Koehne, in Beissner, Schelle, and Zabel, Laubholz-Benennung, 5 5 (1903); Schneider, Laubholzkunde, i. 115 (1904).Betula occidentalis, Lyall, in Journ. Linn. Soc. (Bot.), vii. 134 (1864) (in part) (not Hooker);Sargent, in Bot. Gazette, xxxi. 237 (1901), Silva N. Amer. xiv. 57, t. 725 (1902), and TreesN. Amer. 2 04 (1905) (not Sargent, in Silva N. Amer. ix. 65, t. 453 (1896)).Betula Lyalliana, Koehne, in Mitt. Deut. Dend. Ges. viii. 53 (1899).This variety, which is considered to be a distinct species by Sargent, differsfrom the type in the greater size of the tree, in the orange tint of the bark, and inthe larger leaves, which are thin and membranous in texture, and not so thick andcoriaceous as is usual in this species. The leaves are about 4 inches long and 3inches broad, with seven to nine pairs of nerves, and are coarsely and doublyserrate, broadly ovate, with a broad, truncate base and a slightly acuminate apex.1 Specimens just received from trees cultivated in the Arnold Arboretum, U.S., labelled B. occidentalis, Sargent, andB. kenaica, Evans, are so similar, that they cannot be distinguished even as varieties. This confirms the opinion that var.kenaica and var. Lyalliana (B. occidentalis, Sargent) are merely geographical forms of B. papyrifera, the differences in the size ofthe leaves and fruiting catkins being due to soil and climate. Under cultivation in this country they maintain these differences.2 This name is preferable to B. pafyracea, var. occidentalis, Dippel, Ijaiilholxkunde, ii. 177 (1892), as the plant describedthere is apparently var. grandis, Schneider, a cultivated variety which originated from the eastern form of the species. Cf. ourremarks on B. occidentalis, Hooker, given in p. 993.Betula 985The young branchlets have a minute, dense, erect pubescence, interspersed with longhairs and a few scattered glands. The fruiting-catkins are like those of the type,with slightly thinner scales, the middle lobe of which is triangular and elongated.This splendid tree, which attains a height of 100 or 120 feet, and a diameter of3 or 4 feet, on the alluvial banks of the lower Fraser River, appears to be confinedto the lower basin of that river in south-western British Columbia and north-westernWashington. It was first collected by Lyall in 1859, " in woods by river banks,on the Sumas and Chilukeveyuh prairies and other low grounds to the westward ofthe Cascade Mountains "; and his specimens preserved in the Kew Herbarium areidentical with my own collected near New Westminster.Piper 1 recognises the typical variety as occurring in north-western Washington,where he says that it is a tree with dark grey bark, occasionally 3 feet in diameter.He mentions a similar tree, smaller in size and often white-barked, which grows inStevens County and the Blue Mountains of Washington State. This smaller tree,which also occurs in Idaho, is a connecting link between var. Lyalliana and theform of the species which occurs in the Rocky Mountains.I collected seeds of var. Lyalliana on October 20, 1906, from two trees, about60 feet in height, which I found growing near New Burnaby, on the electric tramline between Vancouver and New Westminster. The virgin forest had been cutdown in this neighbourhood, and the few trees which I saw were young and thriving,and growing in open spaces amidst second-growth Thuya and Douglas. I had notime to descend to the alluvial flats of the Fraser River, where Sargent reports theexistence of trees of large size. At the large lumber mills of Vancouver and NewWestminster, where I made inquiries, the tree is unknown at the present time; butI was informed that some years ago a small quantity of furniture had been madefrom large trees cut down near New Westminster.The seed which I collected has been distributed to various places in GreatBritain, and has germinated well. Seedlings raised in a nursery bed at Casewick,Lincolnshire, by Lord Kesteven, are now (August 1908) 12 to 22inches in height, and for so far have been healthy and vigorous in growth.Some of the seed did not germinate till the following year. At Brocklesby, Mr.Havelock has raised a few plants in a frame, which are 24 to 30 inches high, withfine, large foliage. At Pollokshaws, near Glasgow, Sir John Stirling Maxwell reportsthat they did better when sown in the open than when grown in a frame, andaverage 15 inches in height, the tallest being 29 inches. He adds that this varietyshows every sign of being a thriving tree.Elwes has also raised from seeds sent to him from Kaslo, on Lake Kootenay,British Columbia, in 1904, a few young trees which appear to belong to this variety,and others from seed given him by Professor Sargent, and said to be from thelower Fraser Valley. These are growing vigorously at Colesborne.4. Betula Andrewsii, Nelson, in Bot. Gazette, xliii. 279, with figure of the tree(1907), is a peculiar form with many branching stems from the base, which has beenfound in Colorado.i Contrib. U.S. Nat. Herb. xi. 218 (1906).IV 2 N

986 The Trees of Great Britain and Ireland5. A hybrid between B. papyrifera and B. populifolia, found growing wild inNew Hampshire and Massachusetts, is described by Sargent in Garden and Forest,viii. 356, fig. 50(1895).Several varieties and hybrids have originated in cultivation:6. Var. grandis, Schneider (B. macrophylla, Hort.). Leaves large, cordate,lobulate in margin. Similar leaves appear on coppice shoots and on lower branchesof old trees belonging to the typical form of the species.7. B. Koeknei, Schneider, Laubholzkunde, i. 114 (1907), a hybrid between B.papyrifera and B. verrucosa, is identical with B. cuspidata of Spath's nursery.8. B. excelsa, Aiton, Hort. Kew, iii. 337(1789), long supposed to be either adistinct species or a cultivated variety of B. papyrifera, is considered by Schneider(pp. cit. 1 08) to be a hybrid between this species and B. pumila, and differs from theformer mainly in the smaller size of the leaves.DISTRIBUTIONThe paper birch is the most widely distributed species of Betula in NorthAmerica, the typical form extending northward to Labrador, the southern shores ofHudson's Bay and Great Shore Lake, and southward to Long Island, New York,northern Pennsylvania, central Michigan, central Ohio, northern Nebraska, theBlack Hills of Dakota, and northern Montana. In various forms it also occurswest of the Rocky Mountains to the Pacific Coast, in Alaska, British Columbia,Washington, and Idaho.It usually grows on rich wooded slopes and on the borders of streams, lakes,and swamps; and is common in Canada, New York, and northern New England,becoming rarer to the southward and in the Rocky Mountains. (A. H.)CULTIVATIONNotwithstanding the rarity of this tree in cultivation, it seems to grow freelyat Colesborne, where I have raised it from seed, and planted it out in situationswhere it is exposed to cold and damp. Here it does not suffer from spring frost,and has attained 15 feet in height in seven years. As an ornamental tree, however,it is not in England superior to the common birch, and has no special merit tojustify its being planted except as a curiosity.The paper birch was introduced into England in 1750, according to Loudon ;but is rarely seen except in botanic gardens, as at Kew, where there are severalfair-sized specimens, the largest, a tree with ascending branches, near the Victoriagate, being 45 feet high by 3^ feet in girth. Close to it is another nearly equal insize, with markedly drooping branches. In the Cambridge Botanic Garden, a tree,grafted at i^ foot from the ground, was, in 1906, 47 feet by 4 feet 7 inches.The largest tree we know of in cultivation is in Mr. Kaufman's garden atWhite Knights, near Reading, which Henry measured in 1904 as 82 feet by 4 feet11 inches. Another tall white-barked tree with a clean stem, grafted on commonBetula 987birch near the ground, grows at Bicton, and measured, in 1906, 75 feet by 7 feet2 inches (Plate 259). A large tree is growing at Woburn, near the lake on theright of the main entrance from the village. It is on its own roots, and has barkof a brownish-grey colour, quite unlike the trees at White Knights and Bicton.At Arley, 1 a tree measured 41 feet by 4 feet in 1905. There is also an old andsickly tree at Boynton, in Yorkshire, and a young and healthy one on its own roots,about 40 feet high, at Tortworth.In Scotland and Ireland we have failed to find a single specimen of any size.The handsomest specimen that I have seen in Europe is at the nursery ofSimon-Louis freres at Metz, where on a deep rich loam it has attained 70 feet highby 6 feet 4 inches in girth, and has a fine silvery bark, more beautiful than any thatI know in England. (H. J. E.)BETULA POPULIFOLIA, GREY BIRCHBetula populifolia, Marshall, Arbust. Am. 1 9 (1785); Loudon, Arb. et Frut. Brit. iii. 1707 (1838);Sargent, Situa N. Amer. ix. 55, t. 450 (1896), and Trees N. Amer. 200 (1905); Winkler,Betulacea, 7 9 (1904).Betula excelsa canadensis, Wangenheim, No?-dam. Holz. 8 6 (1787).Betula acuminata, Ehrhart, Belt. Naturk. vi. 98 (1791).Betula cuspidata, Schrader, ex Regel, in DC. Prod. xvi. 2, p. 164 (1868).Betula alba, Linnaeus, var. populifolia, Spach, in Ann. Sc. Nat. ser. 2, xv. 187 (1841).Betula alba, Linnaeus, sub-species populifolia, Regel, in Bull. Sot. Nat. Mosc. xxxviii. 399 (1865).A tree, attaining in America 30 or 40 feet in height and 18 inches in diameter.Bark similar to that of B. verrucosa, but greyish in colour. Young branchletsglabrous, covered with reddish-brown glands, which persist and roughen the shootin the second year. Leaves (Plate 269, Fig. 4), z\ to 3 inches long, i£ to 2 incheswide, deltoid ; broadly cuneate or truncate at the base; prolonged into a longcaudate-acuminate apex; margin lobulate, irregularly serrate; nerves five or sixpairs; both surfaces shining, glabrous, covered with minute brown glands; petiolereddish, long, slender, glandular, glabrous.Fruiting-catkins (Plate 269, Fig. 4), cylindrical, f inch long, \ inch in diameter,pendent or spreading on slender stalks; scales pubescent and ciliate, with shorttriangular middle lobe and recurved broad lateral lobes; wings broader than thenarrow nutlet.This species is closely allied 2 to B. verrucosa; but it is a smaller tree,strikingly different in the colour of the bark, and is readily distinguished by theshape of the leaf, the apex of which is very prolonged, and by the pubescent scalesof the fruiting-catkins.Varieties laciniata and pendula mentioned by Loudon are not known now in1 Hortus Arhyensis, 45 (1907).2 Sargent, in Garden and Forest, ii. 484 (1889), points out the differences between these two species.

The Trees of Great Britain and Irelandcultivation. 1 Var. purpurea, with reddish leaves, was sent out in 1892 by Ellwangerand Barry, of Rochester, New York. B. carulea, Blanchard, is apparently a formof this species, with dull bluish-green leaves, ovate rather than deltoid in outline,which is common on hills in northern New England and eastern Canada.This is the smallest of the arborescent birches of America, and grows usuallyon dry, gravelly or sandy, barren soil, or on the edges of swamps and lakes. Itsarea of distribution extends from Nova Scotia and the valley of the St. Lawrencesouthward to Delaware, and westward through northern New England and NewYork to the southern shores of Lake Ontario. It is very abundant in the coastregion of New England and the middle states, and springs up in abundance afterforest fires or on abandoned farm lands.It was first cultivated in England by Archibald, Duke of Argyll, at Whitton,near Hounslow, in 1750, and is rarely met with now except in botanical gardens.At Kew, the largest specimen, growing near the end of the rhododendron dell, is35 feet high and about 6 inches in diameter. It produces fruit regularly. Loudonmentions a birch supposed to be of this species at Dodington Park in Gloucestershire,60 feet high in 1838, but no such tree now survives there. (A. H.)BETULA NIGRA, RED BIRCHBetitla nigra, Linnaeus, Sf. PI. 982 (1753); Loudon, Arb. et Frut. Brit. iii. 1710 (1838); Sargent,Silva N. Amer. ix. 61, t. 452 (1896), and Trees N. Amtr. 1 98 (1905); Winkler, Betulacece, 5 8(1904).Betula lanulosa, Michaux, Fl. Bor. Am. ii. 181 (1803).Betula rubra, Michaux f., Hist. Arb. Am. ii. 142 (1812).A tree, attaining in America 80 or 90 feet in height, with a trunk occasionally5 feet in diameter, and usually divided into two or three diverging limbs at 15 or 20feet above the ground. Bark at first smooth, reddish brown; with age, separatinginto successive layers, which curl up and persist on the trunk as thin papery scalesof various tints of red and brown ; ultimately turning black and becoming an inchthick and deeply furrowed at the base of old trunks. Young branchlets tomentose,with numerous glands; older branchlets glabrous and roughened with the remainsof the glands. Leaves (Plate 270, Fig. 13), i^ to 3 inches long, i to 2 inches wide,deltoid-ovate, with cuneate base and acute apex; margin non-ciliate, coarselyand irregularly bi-serrate, and often lobulate; nerves seven or eight pairs; uppersurface shining, with fine pubescence mainly on the nerves; lower surface greyish,with pubescence chiefly on the midrib and nerves, and with numerous white glands ;petiole tomentose and glandular.Fruiting-catkins (Plate 270, Fig. 13), cylindrical, i to i| inch long, \ inch indiameter, erect, on stout tomentose peduncles, about \ inch long : scales pubescent1 There is a var. laeiniata, and also a var. purpurea assigned to B . verrucosa in Spath's and Simon-Louis's nurseries,which may be what Loudon referred to.Betula 989and ciliate. The fruit ripens in May or June, and Sargent has called attention tothe fact that the early ripening of the seeds of this and other trees, as the red andsilver maples, growing beside rivers assures their germination, as they fall on thebanks at the season of low water, immediately germinate, and grow speedily.Betula nigra is readily distinguishable by its peculiar bark, the only otherspecies in cultivation which at all resembles it in this respect being B. dahurica. Itis also very distinct in the greyish colour of the leaves beneath, which are cuneateat the base, acute and not acuminate at the apex, and usually lobulate in marginwith sharp double serrations.In winter the twigs are brown, glandular, and almost glabrous; buds minute,| inch long, appressed to the branchlet, with a sharp beak directed inwards; scalesciliate and pubescent.This species, which is known as the red or water birch, 1 grows usually on thebanks of streams and ponds or in swamps, in deep rich soil, liable to inundation. Itoccurs from Massachusetts, southwards to Florida, east of the Alleghany Mountains ;through the Gulf States to Trinity River, Texas; and throughout the Mississippivalley to Indian territory, Kansas, Nebraska, Minnesota, Wisconsin, and Ohio. Itattains its largest size in the damp lowlands of Florida, Louisiana, and Texas, beingthe only birch tree of thes6 warm regions. (A. H.)Sargent says 2 that its distribution is peculiar, as though it grows abundantlyand luxuriantly on the banks of the Merrimac and Spicket rivers in north-eastMassachusetts, it occurs nowhere else in New England, and only becomes commonin the south of New Jersey, extending from thence to Iowa in the west, and toFlorida and Texas in the south, growing in the south on the banks of almost everystream which has a gravelly bed, and of which the banks are not marshy, andattaining a height of 80 or 90 feet with a trunk 3 to 4 feet in diameter ; but in NorthCarolina, according to Ashe, its average size is 40 to 60 feet high by i to 2 feetin diameter.It was introduced into cultivation in England by Peter Collinson in 1736 ; butis rarely met with, though it ought to be more suitable for cultivation 8 in thewarmer parts of England than the more northern species. There are several goodspecimens in Kew Gardens, the largest, near the rhododendron dell, being 57 feethigh and 5 feet in girth. This tree divides at 7 feet from the ground into two mainstems; and one or two smaller trees in the collection of birches branch similarly nearthe ground into two or three stems. A tree near the Victoria gate has a singlestem, 48 feet in height and 4^ feet in girth. Some of these trees bore ripe fruit inJune 1908.Though Sargent says that large specimens may be seen in some of the olderEuropean parks, neither Pard6 nor Correvon mention any trees of this species; butBean 4 saw one 7£ feet in girth at Herrenhausen, Hanover. (H. J. E.)1 It is also known as the river birch, and though known to botanists as B. nigra, it is very seldom called in Americablack birch, the latter name being very commonly applied to B. lenta. " Garden and forest, ii. 591.3 In an article on this tree, in Card. Chron. xxv. 21 (1899), Mr. Bean recommends it for ornamental planting on lowislands in lakes, and beside water-courses.* Kew Bull., 1 908, p. 392.

99 The Trees of Great Britain and Ireland Betula 991BETULA LUTEA, YELLOW BIRCHBetula lutea, Michaux f., Hist. Arb. Am. ii. 152 (1812); Sargent, Silva N. Amer. ix. 53, t. 449(1896), and Trees N. Amer. 1 97 (1905); Winkler, Betulacea, 6 5 (1904).Betula lenta, Linnaeus, van lutea, Regel, in DC. Prod. xvi. 2, p. 179 (1868).Betula lenta, Linnaeus, var. gennina, Regel, in Mhn. Soc. Nat. Mosc. xiii. 126 (1861).Betula excelsa, Pursh, Fl. Am. Sept. ii. 621 (1814) (not Aiton); Loudon, Arb. ef Frut. Brit. iii.1711 (1838).A tree, attaining in America 100 feet in height and 3 or 4 feet in diameter.Bark smooth, shining, silvery or golden grey, breaking into ribbon-like strips andcurls, which long remain attached; in old trees, ^ inch thick, reddish brown, andfissured. Young branchlets covered with long pale hairs; in the second yearsmooth, brown, and usually glabrous.Leaves (Plate 270, Fig. ii), 3 to 4^ inches long, \\ to 2 inches wide, ovateoblong,rounded at the base, acute or slightly acuminate at the apex; margin finelyand sharply serrate, ciliate between the teeth; nerves nine to twelve pairs; bothsurfaces with long silky hairs mainly on the midrib and nerves; pale beneath ;petiole pilose.Fruiting-catkins (Plate 270, Fig. ii) erect, sessile or sub-sessile, ovoid-oblong,i to i^ inch long, f inch in diameter; scale lobes nearly equal, ciliate, pubescent.In winter the slender twigs are more or less pilose; buds fusiform, \ inchlong, rather blunt at the apex, with minutely pubescent, ciliate scales.This species occurs in Newfoundland and along the northern shores of theGulf of St. Lawrence, to the valley of the Rainy River, extending southwards toDelaware and Minnesota, and along the Alleghany Mountains to the high peaksof North Carolina and Tennessee. It is one of the largest broad-leaved treesof the eastern provinces of Canada and New England, where it is abundant,usually growing in rich soil on moist uplands, in company with the beech, sugar andred maples, black and white ash, and white elm. The leaves turn a bright yellowin autumn. (A. H.)The yellow birch was introduced into England, according to Loudon, about1767, but has never become common.The finest we know of in England grows in a shrubbery near the kitchengarden at Tortworth, and measured in 1907 about 50 feet by 4 feet. It hasborne fruit, from which Lord Ducie has raised seedlings. There are smallspecimens in Kew Gardens.At Auchendrane, Ayrshire, Mr. Renwick measured a tree in 1907, 57 feet high,with a bole of 15 feet girthing 5 feet 2 inches.A very fine tree is growing at Oriel Temple, Co. Louth, the seat of LordMasserene, which was mentioned by Loudon under the name B. lenta. WhenI saw it in July 1908, it was in perfect health, and measured 58 feet high by 7 feet4 inches in girth (Plate 260). Loudon states that in his time it was about fiftyyears planted, and 50 feet high with a diameter of i foot 9 inches. (H. J. E.)BETULA LENTA, CHERRY BIRCH, BLACK BIRCHBetula lenta, Linnseus, Sp. PL 9 83 (1753); Loudon, Arb. et Frut. Brit. iii. 1713 (1838); Sargent,Silva N. Amer. ix. 50, t. 448 (1896), and Trees N. Amer. 1 96 (1905); Winkler, Betulacea, 6 4(1904).Betula nigra, Du Roi, Obs. 3 0 (1771) (not Linnseus).Betula carpinifolia, Ehrhart, Beit. Naturk. vi. 99 (1791) (not Siebold and Zuccarini).A tree, attaining in America 80 feet in height and 2 to 5 feet in diameter.Bark smooth, close, dark brown, with pale, elongated, horizontal lenticels, peeling offtransversely in thin strips; on old trunks deeply fissured and broken into large,irregular scaly plates. Young branchlets glabrous, except for a few hairs above theleaf-insertions, slightly glandular.Leaves (Plate 269, Fig. 3), 2^ to 5 inches long, i£ to 3 inches wide, ovateoblong,rounded or slightly cordate at the base, acuminate at the apex; marginnon-ciliate, bi-serrate; nerves nine to thirteen pairs ; upper surface dark green witha few long hairs confined to the midrib and nerves or scattered throughout; lowersurface lighter in colour, with silky hairs on the midrib and nerves, forming axiltufts; petiole pilose at first, ultimately glabrescent.Fruiting-catkins (Plate 269, Fig. 3) erect, sessile, ovoid-oblong, i to i^ inchlong, \ inch in diameter; scales glabrous, with nearly equal lobes, the lateral lobesbeing divergent.This species is characterised by the pleasant aromatic flavour and fragrance ofthe leaves, twigs, and inner bark; and on that account is sometimes named sweetbirch in America. It is, however, more often called, on account of the colour ofthe bark, cherry birch or black birch. In winter the twigs are shining and almostglabrous; buds fusiform, £$ inch long, ending in a sharp beak, brownish, viscid,shining.Var. laciniata, Render, in Rkodora, ix. in (1907). Leaves with six to ninepairs of sharply serrated lobes. A single tree of this variety, which resemblesB. verrucosa, var. dalecarlica, in the form of the leaves, was found in 1901 atNew Boston in New Hampshire.A hybrid between this species and B. pumila, L., has been described bySargent. 1 (A. H.)This species is an inhabitant of Newfoundland and Canada from Nova Scotiato Lake Superior, growing in its greatest perfection in central Ontario, Algoma,and Parry Sound, where Macoun says it is often more than 4 feet in diameter. Inthe United States it extends west to Iowa, and along the Alleghany Mountains toKentucky and Tennessee, attaining a large size in the valleys of North Carolina.Ashe figures on plate 12 of his work a splendid tree 80 feet in height, with a cleanbole 5 feet in diameter. He says the bark is reddish brown and rough on oldtrees, while on young trees and branches it is smooth and dark, resembling that1 Garden and Forest, viii. 243, fig. 36 (1895).

992- The Trees of Great Britain and Irelandof Prunus serotina. It produces seed once in three or four years, and does notbear shade well, though young trees will sprout from the stump. The timber isheavy, hard, and strong, dark brown in colour, and takes a good polish.I saw a very fine tree of this species in a garden at Lancaster, Massachusetts,62 feet by 10 feet, dividing at 3 feet into four stems, which were covered with a verypretty ragged yellow and grey bark.Betula lenta was introduced into England in 1759, according to Loudon. Weare not aware that it has anywhere attained a large size, except at Oakly Park,near Ludlow, the property of the Earl of Plymouth, where on a rich sheltered flaton the banks of the Teme, I found a tree of considerable age, which in August 1908measured about 60 feet by 4 feet 9 inches. The trees in Kew Gardens are about20 feet in height. A specimen at High Canons, Herts, measured 36 feet high by4 feet 2 inches in girth, and bore fruit in 1907. Another at Bicton, 38 feet by3 feet 5 inches, is growing in the Arboretum walk, near the Paper Birch.Beer is sometimes obtained in America by fermenting the sugary sap of thistree. Oil of birch, which is made on a considerable scale in Pennsylvania, is a moreimportant product. This is obtained by distilling the wood, 1 one ton of whichyields about 4 Ibs. of oil. This oil is nearly identical, both in chemical andphysical properties, with oil of winter-green, which is manufactured in the samedistrict ; and commercial oil of winter-green is a mixture of the two oils in varyingproportions. (H. J. E.)BETULA FONTINALISBetula fontinalis, Sargent, in Bot. Gazette, xxxi. 2 39 (1901), and Trees N. Amer. 2 07 (1905).Betula occidentalis, Sargent, SUva N. Amer. ix. 65, t. 453 (1896) (not Hooker); Winkler, Betulaceee,86 (1904); Schneider, Laubholzkunde, i. 114 (1904).A tree, occasionally attaining 40 feet in height and 3 or 4 feet in girth, morecommonly shrubby, with many branching stems. Bark about \ inch thick, darkbrown, shining, not separating into thin layers, marked by pale brown horizontallenticels. Young branchlets viscid, densely covered with resinous glands, interspersed with long, pale hairs ; older branchlets dark in colour and roughened withthe persistent glands. Leaves about i^ inch long, and i to i^ inch broad, thin intexture, broadly or narrowly ovate ; rounded, truncate or subcordate, and oftenunequal at the base ; acute at the apex ; margin ciliate, sharply and doubly serrate ;nerves six to eight pairs ; both surfaces glandular with scattered long hairs, atfirst paler beneath, becoming glabrescent ; petiole, ^ inch, glandular, glabrescent.Fruiting-catkins, about i inch long, \ inch in diameter, cylindrical, on slenderglandular stalks ; scales pubescent, ciliate, with the three lobes triangular and nearlyequal in size, the lateral lobes divergent ; nutlets with broad wings.This species is readily distinguished by its conspicuously glandular branchletsand its small, thin leaves, which are variable in width, and in the form of the base.1 Cf. article by H. Trimble in Garden and Forest, viii. 303 (1895), where the process is described. The oil of birch iscontained in the inner bark only ; and on this account the wood used in distillation is obtained from small trees, usuallycoppice shoots.Betula 993B. fontinalis x B. papyrifera. There are two small trees in Kew Gardens,with wide-spreading pendulous branches, which were obtained from Dieck in 1891,and were said to be B. occidentalis from Alaska. One of these trees has firm, darkbrown,shining bark like that of B. fontinalis; while the other has white bark,peeling off in shreds, indistinguishable from that of B. papyrifera. The branchletsare exactly similar to those of B. fontinalis. The leaves (Plate 269, Fig. 8) arevery variable and not precisely the same on both trees; resembling those ofB. fontinalis in colour, but much larger and much thicker in texture; 2 to 3 inchesor more in length, broadly ovate; rounded, truncate, or cuneate at the base,acute at the apex, coarsely serrate or toothed in margin, pubescent and glanddottedon both surfaces. The fruiting-catkins (Plate 269, Fig. 8) and the scalesare as large as those of B. papyrifera; but the scales are more like those of B.fontinalis in shape, the three lobes being almost triangular, glabrous, and ciliate. In allprobability these two trees, with such variable characters in the bark and foliage,are hybrid between B. fontinalis and B'. papyrifera, which occur in the same region.The small-leaved birch, described above under the name B. fontinalis, wasidentified by Nuttall 1 with B. occidentalis, W . J. Hooker, Fl. Bor. Amer. ii. 155(1839); and most botanists, including Sargent 2 in 1896, Winkler, and Schneider,have followed Nuttall. Sargent, 8 however, in 1901 advanced the opinion that thetall, large-leaved birch of the lower Fraser River, considered by us to be B.papyrifera, var. Lyalliana, is a distinct species, which he identified with Hooker'sB. occidentalis ; and he proposed the name B. fontinalis for the small-leaved birch.The material 4 in the Kew Herbarium, on which Hooker founded his species,includes no less than three distinct birches, none of which, however, is the largeleavedvariety of B. papyrifera; and as his description is confused and not confinedto a single species, the name B. occidentalis, Hooker, must be entirely abandoned;and, in consequence, B. fontinalis is rightly adopted for the small-leaved birch, asbeing the first valid name for this species.B. fontinalis is a small tree or spreading shrub, widely distributed in westernNorth America, where it usually grows on moist soil near the banks of streams inmountain valleys. It extends from the basin of the upper Fraser and Peace riversin British Columbia, Alberta, and the valley of the Saskatchewan, southwards toMount Shasta and the northern Sierra Nevada in California, and through the RockyMountains and the interior ranges to Nevada, Utah, and northern New Mexico ;extending eastwards in the United States to the Black Hills of Dakota and northwestern Nebraska. Mr. M'lnnes, of the Canadian Geological Survey, has recentlydiscovered this species in the district north of Lake Superior. 8i N. Amer. Sylva, i. 22 (1842). " SUva N. Amer. ix. 65 (1896). 3 Bot. Gazette, xxxi. 237 (1901).4 This material includes :I. One specimen collected by Dr. Scouler, labelled "De Fuca Straits"; another, collected by Dr. Tolmie,"N.-W. Coast"; and a third collected by Douglas "west of the Rocky Mountains." These three specimens are a small-leaved variety of B. papyrifera, identical with B. kenaica, Evans.II. Two specimens, collected by Dr. Richardson, labelled "Arctic Sea-Coast," one of which is B. fontinalis, Sargent;and the other, B. alaskana, Sargent.III. A specimen, with young foliage, collected by Drummond, near Edmonton, which is probably B. fontinalis.6 Canadian Forestry Journal, 1 905, 175-IV 2 0

"T994 The Trees of Great Britain and IrelandIt was introduced into cultivation in the Arnold Arboretum in 1874; but israrely met with in Europe, except in Botanic Gardens. Four or five specimens, inthe lower nursery at Kew, which were received from Sargent in 1903, are now 8 to10 feet in height, and are vigorous in growth, promising to become trees of considerable size. Another specimen in the collection at Kew, raised from seedobtained from the Pinehurst Nursery in 1897, is about 10 feet in height and is moreshrubby in appearance. (A. H.)TIMBER OF THE AMERICAN BIRCHES JAll American authors agree in saying that the tree which supplies the bestwood of commerce is that of the Betula lenta, known in New England as blackbirch, in Canada as cherry birch. Macoun says that the yellow birch is scarcelydistinguished in commerce, and judging from the specimens in Hough's AmericanWoods, it would be difficult for any one but an expert to do so. Michaux andMacoun both say that the timber of the white birch is less valuable than either ofthe above; and Sargent says that the wood of the red birch has lately been foundsuitable for furniture of the best quality. Mr. Weale tells me that large quantitiesof red and yellow birch are imported into Liverpool, chiefly in the form ofhewn logs of 15 to 18 inches square, which are valued in proportion to theirsize at from is. 6d. to 2S. a foot and upwards for logs showing figure. The woodoften shows a beautiful undulation called " roll figure," which, when cut into veneers,was a few years ago very fashionable for bedroom furniture, and is sometimes ratherdifficult to distinguish from satin wood, though its colour is rather pinkish thanyellowish and the undulations larger and more open. Mr. C. L. Willey, of Chicago,tells me that this fine " curly birch" is principally found at altitudes of 3000 to 4000feet in North Carolina, and is very light in colour, having a yellowish tint; whilst othertrees produce wood of a reddish colour, resembling that of cherry (Primus serotina).A large quantity of American birch is also imported in the form of planks2 inches to 5 inches thick, and averaging 8 inches wide, which are sold in Liverpoolat about is. per cubic foot, and consumed for chair-making at High Wycombe andelsewhere. At this low price it is the most formidable competitor to the nativebeech, and in some of the factories which I visited, seemed to be the more popularwood of the two.The bark of the Paper Birch, and to a less extent that of other species, is asimportant to the inhabitants of Canada as that of the common birch is to those ofnorthern Europe. Canoes and lodges are covered with large sheets of bark; itis placed on shingled roofs under the shingles to prevent the water from comingthrough; and very ornamental boxes, baskets, and other articles are made from itby the Indians. It also serves as a writing material, and I have a clearly writtenletter from Prof. Elrod, sent me by him, when during an expedition in Montanahe ran out of paper. (H. J. E.)1 Emerson distinguishes the five common birches of New England as follows :1. B. lenta, black birch ; bark dark coloured.2. B. lutea, yellow birch ; bark yellowish, with a silvery lustre.3. B. nigra, red birch ; bark reddish or chocolate coloured, very much broken and ragged.4. B. papyrifera, canoe birch ; bark white with a pearly lustre.5. B. populifolia, grey or white birch ; bark chalky-white, dotted with black.DIOSPYROSDiospyros, Linnaeus, Gen. PL 1 43 (1737)5 Hiern in Trans. Camb. Phil. Soc. xii. i, 144 (1873);Bentham et Hooker, Gen. PI. ii. 665 (1876); Giirke in Engler u. Prantl, Pflanzcnfam. iv. i,161 (1890).Cargillia, R. Brown, Prod. Fl. Nov. Holl. 5 26 (1810).Leucoxylum, Blume, Bijdr. Fl. Ned. hid. 1 169 (1825).Noltia, Schumacher, Dansk. Vidensk. Selsk. Skrift. iii. 189 (1828).Rospidios, A. de Candolle, Prodr. viii. 220 (1844).TREES or shrubs, belonging to the order Ebenaceae. Leaves alternate or rarelysub-opposite, deciduous or persistent, simple, entire, without stipules.Flowers dioecious or rarely polygamous, monoecious or perfect, 1 in cymes or solitaryfrom the axils of the leaves of the current year, or in a few species arising from theold wood. Calyx, three- to seven-, usually four-lobed, pubescent, and accrescentunder the fruit. Corolla urn-shaped, campanulate, tubular, or salver-shaped, three- toseven-, usually four- to five-lobed, pubescent. Male flowers small, usually in cymes;stamens four to sixteen, inserted on the base of the corolla or hypogynous ; filamentsslender and often united by pairs, forming an outer and inner series; anthers openinglongitudinally or by apical pores ; ovary aborted or wanting. Female flowers oftensolitary; staminodes four to eight, sometimes wanting, occasionally with fertileanthers; ovary with four to sixteen cells, which are double the number of the stylesand one-ovuled, or rarely of the same number as the styles and two-ovuled. Fruit aberry, with the enlarged and persistent calyx at its base, containing one to ten ormore oblong seeds, which have a copious albumen.The alternate, simple, stalked, entire leaves, without stipules, and the shootswithout true terminal buds and with two persistent bud-scales at their base, aredistinguishing marks of the genus.About 180 species of Diospyros are known, mostly confined to the subtropicaland tropical regions of both hemispheres. The wood is usually hard and closegrained,the heartwood black, the sapwood soft, thick and yellow.Only three species are in cultivation in this country; and of these DiospyrosKakt, Linnaeus f., the Chinese Persimmon, a shrub or small tree, usually only metwith in England in greenhouses or trained against a wall, does not come within thescope of our work. It ripens its fruit in warm summers in England. (A. H.)1 In many species the sexes are unstable; cf. Wright, Ann. K. But. Card, feradeniya, ii. pt. i. I, 133 (1904).995

996 The Trees of Great Britain and IrelandDIOSPYROS VIRGINIANA, AMERICANPERSIMMONDiospyros mrginiana, Linnaeus, Sj>. PI. 1 057 (1753); Loudon, Arb. et Frut. Brit. ii. 1195 (1838)Sargent, Silva N. Amer. vi. 7, tt. 252, 253 (1894), and Trees N. Amer. 749 (1905).Diospyros guajacana, Romans, Nat. Hist. Florida, 2 0 (1775).Diospyros concolor, Moench, Meth. 4 71 (1794).Diospyros pubescens, Pursh, Fl. Am. Sept. i. 265(1814); Loudon, Arb. et Frut. Brit. ii. 1196(1838).Diospyros caroliniana, Rafinesque, Fl. Ludovic. 1 39 (1817).Diospyros Persimmon, Wikstrom, Jahr. Schwed. 1 830, p. 92 (1834).A deciduous tree, attaining occasionally in America 115 feet in height and 6feet in girth, but usually smaller. Bark* deeply divided into square corky plates.Young shoots with a minute dense erect pubescence, persistent usually in the secondyear. Leaves (Plate 199, Fig. 3) oblong or elliptical; rounded and unequal orbroadly cuneate at the base; shortly acuminate at the apex; margin entire andciliate; upper surface dull, light green, and glabrous except for some pubescence onthe midrib at the base ; lower surface pale, glabrous; veins pinnate, arcuate, andlooping near the margin ; petiole pubescent, ^ to i inch long.Flowers appearing, when the leaves are more than half-grown, on the currentyear's shoot, dioecious. S laminate flowers in two- to three-flowered pubescentpedunculate cymes; calyx with four broadly ovate acute ciliate lobes; corollatubular, slightly contracted below the very short acute reflexed lobes; stamenssixteen, in two series, with pubescent filaments. Pistillate flowers, solitary, on shortrecurved peduncles; stamens eight, usually with aborted anthers; ovary pilosetowards the apex, eight-celled; styles four, two-lobed at the apex, pubescent at thebase.Fruit solitary, on short woody peduncles, persistent on the branches duringwinter; depressed, globose; surrounded at the base by the persistent calyx, which hasfour broadly ovate pointed recurved lobes. The fruit is variable in size, from that ofa small cherry to a large plum ; and its flavour is very different in different localitiesand even on trees growing close together sometimes sweet without the action offrost, or ripening after frost, or at other times acid and never edible. Seeds oblong,flattened, £ inch long. Seedless forms occur, and experiments are being made inAmerica with these and other good varieties.The leaves on trees, growing in the Southern States, are strongly pubescentbeneath; and this variety, which we have not seen in cultivation in England, isscarcely to be distinguished by the foliage alone from D. Lotus, which has pubescentleaves, and differs in this respect from the ordinary form of D. virginiana, with1 The bark is well figured in Card. CAron. iv. 504, fig. 7 1 ( 1888).Diospyros 997glabrous leaves. The buds and leaf-scars of the two species are very different, andconstitute the best marks of distinction.In winter, the American Persimmon (Plate 200, Fig. 4) shows the followingcharacters: Twigs slender, covered with a minute dense erect pubescence, with twobroadly ovate scales (of the previous season's bud) persisting at the base of the shoot.Leaf-scars, oblique on prominent pulvini, small, semicircular, with a deep transverselunate depression, showing indistinctly the coalesced cicatrices of the vascularbundles. True terminal bud not formed, the tip of the branchlet dying off insummer and leaving at the apex of the twig a short glabrous stump with a terminalscar, which subtends the uppermost axillary bud. Buds ovoid, slightly compressed,small, brown, shining, glabrous, usually with a minute curved beak tipped by a fewhairs; outer scales two, imbricate, ovate, acute, non-ciliate, concave interiorly,pubescent at the tip, glabrous elsewhere.This species is widely distributed in the United States, its most northerly pointbeing at Newhaven in Connecticut. It is not uncommon in Long Island, and extendssouthwards to Alabama and Florida, and westwards through Ohio and Iowa to Missouri,Arkansas, Louisiana, eastern Kansas, Indian Territory, and the valley of the ColoradoRiver in Texas. It usually grows on light, sandy, well-drained soil, but attains itslargest size in the deep alluvial lands of the Mississippi basin, where it sometimesreaches a height of more than 100 feet, with a slender trunk free from branches for70 or 80 feet. It is exceedingly common in the south Atlantic and Gulf States,often covering with its suckers abandoned fields, and springing up by the sides ofroads and fences. Sargent* gives a figure of a tree with wide-spreading branches,not unlike the specimen at Kew in size and appearance, which is growing in an oldcorn-field near Auburn in Alabama.Elwes saw a fine tree of this species in a damp river bottom near Mount Carmel,Illinois, in 1904, which measured 100 feet high by 6 feet in girth, with a clean,straight trunk 60 feet high; but the late Dr. Schneck, who showed it to him,measured one as much as 115 feet high, 80 feet to the first limb, and only 5 % feet ingirth at the base. When growing in open fields or along road-sides, where it ismost frequently seen, it forms a more spreading tree, usually 30 to 40, and rarelymore than 60 feet high. (A - H-)CULTIVATIONThis tree is easy to raise from seed, and perfectly hardy in England, but requiresa warm, dry soil, and a much hotter summer than usual to make it thrive. Theseedlings which I have raised grow very slowly and do not root freely in my soil.Judging from the extreme rarity of the tree in cultivation, it is hardly likely to beworth planting generally, and, so far as we know, has never borne fruit in England.Even in the climate of central France it fruits, according to Parde, 2 very rarely, andgrows slowly, having only attained about 20 feet in height at Les Barres. Neither' Garden and Forest, viii. 262, fig. 38 (1895). Arboretum Nat. des Barres, 2 15 ( 1906).

998 The Trees of Great Britain and IrelandMouillefert nor Mayr mention this tree; and I have seen none in Europe worthmentioning, except a fine tree in the Botanic Garden at Padua, which appears to bethe pubescent variety of this species, although it is labelled D. Lotus. According toProf. Saccardo, it was planted in 1760, and measured 30 metres high by 2 metres ingirth in 1887.The American Persimmon was introduced into England some time before 1629,when an account of a cultivated tree appeared in Parkinson's Paradisus published inthat year.It is extremely rare in this country at the present day, and Loudon in 1838only mentioned six trees, two of which still survive. One of these, which is astaminate tree, growing in Kew Gardens, now measures 64 feet in height by 5 feet3 inches in girth, and, according to Sargent, 1 is apparently as thriving as if it werein its native habitat. It is one of the denizens of the original Kew Arboretum,which was laid out by W. Aiton, and in all probability was one of the numeroustrees presented in 1762 to the mother of George III. by Archibald, Duke ofArgyll, who was a great introducer and cultivator of rare trees at Whitton, nearHounslow. 2 (Plate 261.)Another, mentioned by Loudon as being 24 years planted and 18 feet highin 1838, is growing at the Wilderness, White Knights, near Reading, and is now45 feet high by 4 feet i inch in girth. At Barton, Suffolk, another is 40 feethigh by 2 feet i inch in girth. At Bushey Lodge a tree with a broken top was, in1904, 30 feet high by 5 feet 8 inches in girth. Suckers are growing from its rootsas far away as 50 feet, and one of these, 10 feet high, is said to be about tenyears old.TIMBERThe wood is very hard and heavy, of a pale yellowish-white colour, with blackheartwood, which, however, usually shows only in old trees. Hough states thathe felled one 14 inches in diameter for the specimens in his work,8 but thoughthere were over sixty rings of sapwood, only two or three in the heart wereblack. It is used in America for shoe-lasts and shuttles, for which latter purpose itis imported to a small extent to Liverpool. Michaux 4 states that it was used atCharleston for shafts, and preferred for that purpose to ash or any wood exceptlancewood, but the quantity available is too small to give this timber much economicimportance. The fruit is little valued as human food, though eaten by animals.(H. J- E.)1 Garden and Forest, loc. cit.2 Cf. J. Smith, Records of Kew Gardens, 258; and Nicholson, in Card. Chron. iv. 504 (1888), in which is given a goodpicture of the tree (fig. 72). Cf. also Kew Bulletin, 1 891, p. 292.* American Woods, iii. No. 61.4 Midi.//., N. Am. Sylva, ii. 222.Diospyros 999DIOSPYROS LOTUS, DATE-PLUMDiosfyros Lotus, Linnaeus, Sp. PL 1 057 (1753); Loudon, Arb. et Frut. Brit. ii. 1194 (1838); C.B. Clarke, in Hooker, Fl. Brit. India, iii. 555 (1.882); Hemsley, in Journ. Linn. Soc. (£ot.)xxvi. 70 (1889); Shirasawa, Icon. Ess. Forest. Japan, 1 23, t. 79 (1900).Diospyros microcarpa, Siebold, in Ann. Soc. Hort. Pays.-Bas. 1 844, p. 28.Diospyros japonica, Siebold et Zuccarini in Abh. Bayer. Acad. iv. 3, p. 136 (1846).A tree attaining 60 feet in height and 6 feet in girth. Bark remaining a longtime smooth, finally rough and with plate-like scales. Young shoots with a moderately long dense pubescence, often persistent in the second year. Leaves (Plate 199,Fig. 4) oblong or elliptical, base rounded or broadly cuneate, apex acuminate or acute;margin entire, ciliate ; upper surface dark green, shining, usually becoming glabrousexcept at the base of the midrib, but often with scattered minute hairs on the veinsand veinlets; lower surface pale and pubescent throughout; veins pinnate andlooping towards the margin; petiole, \ to \ inch, pubescent.Flowers dioecious. Staminate flowers, two to three together in subsessile cymes ;calyx with four short ovate acute ciliate lobes ; corolla urceolate, with four short obtuselobes ; stamens sixteen, in pairs in two series ; filaments glabrous. Pistillate flowers,solitary, subsessile; staminodes, eight; ovary eight-celled, one ovule in each cell;styles, four. Fruit subsessile, almost globose, yellow or blackish, \ to f inch indiameter; fruiting-calyx spreading, with a ring of short dense silky hairs on theinside beneath the fruit. The fruit varies considerably in size, and is astringent inflavour.The leaves pubescent beneath, and the different buds and leaf-scars distinguishthis species in summer from D. virginiana. I n winter the following characters(Plate 200, Fig. 3) are available: Twigs slender, usually with scattered long hairs,occasionally glabrous; two long acuminate scales of the previous season's budpersist at the base of the shoot. Leaf-scars small, nearly parallel to the twig onprominent pulvini, semicircular, marked with a raised transverse crescentic ridge,composed of the coalesced bundle cicatrices. True terminal bud absent, stump atthe apex of the twig pubescent. Buds long, ovoid, acuminate, blackish, pubescent;outer scales two, imbricate, long, acuminate, pubescent, ciliate, concave interiorly.This species has been long in cultivation, and its exact distribution in the wildstate is difficult to define. It appears to be indigenous in Asia Minor; in theCaucasus, 1 where it occurs wild throughout the whole territory between sea-level and3500 feet; in Afghanistan; in the north-west Himalaya 2 at 2000 to 6000 feet inHazara and Kashmir ; and in central and northern China.It has been cultivated for centuries in the countries bordering on the Mediterranean, and has become naturalised in many places, as in the south of France andin Dalmatia. It is not wild in Japan, but is often planted there, either for its ownKadde, Pflanzetwcrb. Kaukasusldnd. iSl ( 1889). 2 Gamble, Indian Timbers, 455 (1902).

iooo The Trees of Great Britain and Irelandfruit or as a stock on which to graft the persimmon. In China it is largelycultivated, and the fruits, known as hei-tsao or "black dates," are an article ofcommerce. - (A. H.)The date-plum was early introduced into England, being cultivated by Gerard 1in 1633 or earlier. It grows easily from seed, and, according to Loudon, at the rateof 12 to 18 inches annually, if planted in rich soil, 2 but requires a warmer climate thanours and never attains a large size, so far as we know, in England ; though trees of20 feet or so in height are sometimes seen in botanic gardens and in parks. Itripens fruit usually every year at Kew.Mayr 3 figures the wood, which is very remarkable on account of thecontrast in colour between the black heart - wood and the pale sap - wood.Judging from this and the wood of the Japanese kaki, this wood if procurablewould be valuable for cabinet-making; but, so far as I know, it is nowhere commonenough to have acquired any recognised commercial value. (H. J. E.)1 Herball, Johnson's edition, 1495 (1633).2 The seedlings which I have raised from seed collected in France, seem liable to injury by frost, and do not ripen theirshoots when young.3 Fremdldnd. Wald- u. Parkbaume, 464, t. xvii. fig. 29 (1906).END OF VOL. IVPrinted by R . & R. CLARK, LIMITED, Edinburgh.

SILVER FIR AT COWURAYPi ACE 2oS.

SILVER FIRS AT I.ONCI.EATTI.ATK 200.

SILVER FIR AT ROSENEATHPLATE 210.

SILVER FIR AT TULLYMOREPLATE 211.

SPANISH FIR IN ANDALUSIAPLATE 212.

\SPANISH FIR AT LONGLEATPLATE 213.

GREEK EIR AT BARTON

IIIMALAY \N I-'IR IN SIKKI.MI'l.All. 215.

JAPANESE FIR IN JAPANPLATE 216.

*--»*:>££••*e ->* . .**GIANT FIR AT EASTNOR CASTLEPLATE 2 17.

GIANT KIR IN VANCOUVER'S ISLANDPLATE 218.

CAUFORNIAN FIR AT I.TNTONPLATE 219.

LOVELY FIR IN BRITISH COLUMBIAPLATE 220.

NOBLE FIR IN OREGONPLATE 221.

RED OR SHASTA FIR AT BAYFORDBURYPlJVTK 222

RED OR SHASTA FIR AT BONSKEIDPLATE 223.

BRISTLE-CONE FIR AT EASTNOR CASTLEPLATE 224.

ROCKY MOUNTAIN FIR IN MONTANAPLATE 225.

MEXICAN FIR AT FOTAI'l ATE 226.

DOUGLAS FIR ON BARKLEY'S FARMPLATE 227.

DOUGLAS FIR FOREST IN VANCOUVER'S ISLANDPLATE 228.

DOUGLAS FIR AT EGGESFORDPLATE 229.

DOUGLAS FIR AT LYNEDOCHI'l A I k 230.

DOUGLAS FIR AT TORTWORTHPLATE 231.

SPANISH CHESTNUT GROVE AT BICTONPLATE 232.

SPANISH CHESTNUT AT ALTHORPPLATE 233.

ajjSPANISH CFIESTNUT AT THORESBVK 234.

••*• -,-• -Vt*- '*SPANISH CHESTNUT AT RYDALPLATE 235.

SPANISH CHESTNUT AT ROSSANAGIIPLATE 236.

JAPANESE CHESTNUT AT ATERA, JAPANPLATE 237.

WEEPING ASH AT ELVASTON CASTLEPLATE 238.

TALL ASH AT COBHAM PARK239.

TWISTED ASH AT COBHAM PARKPLATE 240.

TALL ASH AT ASHRIDGEPLATE 241.

ASH AT WOODSTOCK, KILKENNYPLATE 242.

ASH AT CASTLEWELLAN

DEFORMED ASH AT CIRENCESTER DISEASED ASH AT COLESBORNEPLATE 244.

NARROW-LEAVED ASH, ROUGHAM HALLPLATE 245.

WHITE ASH AT KEWI'LAMC 246.

•Rz a .i.J.>mOD A5TIAVV.1 XV IISV H>IOK.L'IIiT

ZELKOVA CRENATA AT WARDOUR CASTLEPlATE 248.

ZELKOVA CRENATA AT GLASNEVINPLATE 249.

ZELKOVA ACUMINATA AT CARLSRUHEPLATE 250.

CELTIS OCCIDENTALIS AT WEST DEAN PARKPLATE 251.

ALDERS AT LILFORDPi ATE 252.

ALDERS AT KILMACUKRAGHPLATE 253.

-?.--*•- ' • -, * ' %^;—.-vsr- • ». " - .-'.-TilEf-- , "~.^, j. iTî ;-»#•>.,'''.^ ; - — -.^./-,---- ' -,W- '->-J '&?£*- - %•' --.^" "---.• -i-- :'^-~"* "•' "**"•-•:•.---•••*-• -^-sITALIAN ALDER AT TOTTENHAM HOUSE, SAVERNAKEPLATE 254.

-*.•BIRCH AT SAVERNAKE FORESTPLATE 255.

BIRCH AT MERTON HALL

BIRCH IN SHERWOOD FORESTPLATE 257.

GNARLED BIRCHES IN GLENMOREPLATE 258.

Pftsw^*/"ki' ^l*..-JL-r.l,1 ^lL4«. ••'> ike' , .'-.ii'. f^JSP* r 7>J >iS^-*i1*«^7'-''*^*'V^\ •?• •. ^.i^-^Â^kvH*--^;*' - •"^•"^*At yJt5E*.V'.AN''l:r5IPHffi^hH05

S£-* " - -. i • ^'{tyzLfK \JN^?^^%OM 3^^plîH^^î%^.?9^YELLOW BIRCH AT ORIEL TEMPLEPLATE 260.

•U..DIOSPYROS VIRGINIANA AT K.EWPLATE 261.

excelsior monophyllnanguslifolialuonoplivlla .Hvutt del PLATE 2 6 ZFRAXI NUS

FRAXINUSPLATE 363

Huitt.del. Hu.th,lith PLATE 264F RAX I N US .

25. G-oin TiCLis O r n u s .HiLrtt dd. HuthjithPLATE 265F RAX INUS .

iltinorL-una .Haiti, dd HutKlitJn PLATE s eeFRAXINUS.

Z c-ronaia. /' \t. d«t Ilutll L-iliC mississippirnsis.PLATE £67MORUS, CELTIS , AND ZELKOVA

niarithma. I :\"- • K A•J*^*'tenmiohjv ires* enIfuitt del Hut]. hLhA L- N U SPLATE 268.

HL±Jm

vemicosa daJeeai-lJca.15.utilis cTacqriemcmtucorylifolia.Uiutt, del, HuUililK.PLATE E7OB E T U LA.

THE TREES OF GREAT BRITAIN & IRELAND - VOL. IV

You also want an ePaper? Increase the reach of your titles

Delete template?

Save as template?