® The source of this uncorrected OCR text may be viewed as a digital facsimile at: http://fax.libs.uga.edu/ ' ,'"',«. • •••&••• ••*/!*- fi // ;' V». •'&•»'' . » »;;•. 1^:. y«,,; ^v .<*•• »/ Great Br'ta \ n Ireland Henry John Elwes, F.R.S. AND Augustine Henry, M.A. Privately Printed THE TREES OF GREAT BRITAIN AND IRELAND ^ >"^ \ / . v/ "c •*.» * -\ "^ .*'.A V V & ^\ l \\ ' ; * »-^^_ te 4- \, * Y • ^ji' l. ;• - i- M ; - « . * V A* *> V , ,. -/ P - • x^r^ /^ ' 3 i- H^ "^5 • \ \ « i '* — >• >v U". Jt-t ^"t V ' f V, l'INT LAI' », lOKKdr OK BAVKLLA, CORSICA •*>. , ..t. Great Britain BY Henry John Elwes, F.R.S. AND Augustine Henry, M.A. Ireland VOLUME II Edinburgh : Privately Printed MCMVII CONTENTS TAGE LIST OF ILLUSTRATIONS ......... v THUJOPSIS ......... .201 ^ÎÎSCULUS ........... 206 COMMON HORSE-CHESTNUT ....... 210 RED HORSE-CHESTNUT . . . . . . . . .217 INDIAN HORSE-CHESTNUT . . . . . . . . 219 JAPANESE HORSE-CHESTNUT. . . . . . . . .221 OHIO BUCKEYE ....... . 223 SWEET BUCKEYE ........ 224 CALIFORNIAN BUCKEYK ...... . 226 TSUGA ........ . 227 HOOKER'S HEMLOCK . ... . . 229 WESTERN HEMLOCK ......... 2 34 HEMLOCK OR HEMLOCK SPRUCE ....... 239 CAROLINA HEMLOCK ........ 243 HIMALAYAN HEMLOCK ....... 244 SIEBOLD'S HEMLOCK ..... . . 246 JAPANESE HEMLOCK ........ 247 JUGLANS ......... . 249 COMMON WALNUT ....... 251 BLACK WALNUT ........ 262 BUTTERNUT. . ....... 271 TEXAN WALNUT . ...... 274 MANCHURIAN WALNUT . . . . . . . .276 CORDATE WALNUT . . . . . . . . . .277 iii IV The Trees of Great Britain and Ireland NARROW-FRUITED WALNUT . SIEBOLD'S WALNUT . COMMON OR STALKED-CUPPED OAK . SESSILE OR DURMAST OAK . PUBESCENT OAK LARIX COMMON LARCH RUSSIAN LARCH SIBERIAN LARCH KURILE LARCH JAPANESE LARCH SIKKIM LARCH CHINESE LARCH TAMARACK . WESTERN LARCH LYALL'S LARCH PINUS LARICIO HERZEGOVINIAN PINE GYMNOCLADUS CHINENSIS GYMNOCLADUS CANADENSIS . CEDRELA SINENSIS . PTEROCARYA PTEROCARYA CAUCASICA PTEROCARYA RHOIFOLIA PTEROCARYA STENOPTERA CLADRASTIS . YELLOW-WOOD CLADRASTIS AMURENSIS CLADRASTIS SINENSIS PAGE 279 282 29I 294 345 349 374 379 383 384 388 391 392 395 4°3 407 424 427 428 433 436 438 442 443 445 446 449 45° ILLUSTRATIONS Pinus Laricio, Forest of Bavella, Corsica (from an original sketch by the late Robert Elwes) Frontispiece PLATE No. Weeping Beech at Endsleigh. . ...... 580 /Esculus, leaves . . . . . . . . . 61 ^Esculus, twigs and buds . ....... 62 Horse-Chestnut at Colesborne ....... 63 Weeping Horse-Chestnut at Dunkeld ....... 64 indica at Barton ........ 65 turbinata in Japan ....... .66 Hooker's Hemlock at Murthly ....... 67 Western Hemlock at Dropmore ...... 68 Western Hemlock at Murthly ... ... 69 Hemlock Spruce at Foxley . ....... 70 Hemlock Spruce at Strathfieldsaye ....... 71 Himalayan Hemlock at Boconnoc . . . . . . . . 72 Juglans, leaves . . ...... 73 Walnut at Barrington Park ......... 74 Walnut at Gordon Castle . . . . . . . . . 75 Black Walnut at Twickenham . . . . . . . . 76 Black Walnut at the Mote . . . . . . . . . 77 Quercus, twigs and buds ....... .78 Quercus pedunculata, sessiliflora, and lanuginosa ; leaves ... .79 Cypress Oak at Melbury . . ..... 80 Self-sown Oaks at Thornbury Castle . . . . . . . . 81 Champion Oak at Powis Castle . .... 82 Oak at Powis Castle .......... 83 Lady Powis' Oak at Powis Castle . ...... 84 Tall Oak at Whitfield ....... .85 Oak at Kyre Park . . . .... 86 Tall Oaks at Kyre Park . .... .87 Billy Wilkin's Oak at Melbury ........ 88 v vi The Trees of Great Britain and Ireland Beggar's Oak in Bagot's Park ........ Oaks at Bagot's Park ........ King Oak at Bagot's Pai k ........ Sessile Oak at Merevale Park ........ Oak at Bourton-on-the-Water ...... Oak at Althorp ..... . Major Oak in Sherwood Forest ...... Brown Oak at Rockingham Park ........ Umbrella Oak at Castle Hill .... ... Larches in Oakley Park ....... Larches at Sherborne, Gloucestershire . ...... Larches at Colesborne ........ Champion Larch at Taymouth ....... Forked Larch at Taymouth ......... Mother Larch at Dunkeld ......... Forked Larch at Gordon Castle ..... . . Larch in the Alps .......... Dahurian Larch at Woburn ....... Larch in Kurile Islands ......... Japanese Larch at Tortworth......... Sikkim Larch at Strete Raleigh ........ American Larch at Dropmore ........ Western Larch in Montana ... Lyall's Larch in Alberta ........ Corsican Pine in Corsica ...... Corsican Pine in Corsica ......... Pinus Laricio on sandhills at Holkham ....... Pinus Laricio at Holkham ....... Pinus Laricio at Arley .... Crimean Pine at Elveden ....... Pinus leucodermis in Bosnia ......... Gymnocladus canadensis ........ Pterocarya caucasica at Melbury ........ Pterocarya caucasica at Claremont ...... Pterocarya caucasica at Cambridge . . . . Yellow-Wood at Syon . . Pterocarya, Gymnocladus, Cladrastis, and Cedrela ; leaves .... Liriodendron, Cedrela, Ailanthus Cladrastis, and Corylus ; twigs and buds Pl A PF NO. 89 90 9» 92 93 94 95 96 97 98 99 100 101 102 103 104 !°5 106 107 108 109 no in 112 "3 114 "5 116 117 118 119 120 121 122 123 124 125 126 THUJOPSIS TTiujofsis, Siebold et Zuccarini, FÎ. Jap. ii. 32 (1842). Thuya, Bentham et Hooker, Gen. PI. iii. 427 (1880). Cttfressus, Masters, Journ. Linn. Soc. (Bot.\ xxx. 19 (1893) an(^ xxxi. 3^3 (1896). THIS genus is considered by many authorities to be merely a section of Cupressus or of Thuya. The foliage and cones, however, are remarkably distinct, and justify its retention as a separate genus. Evergreen trees, belonging to the tribe Cupressinese of the order Coniferse, with reddish bark scaling off in longitudinal shreds. Branches in false whorls or scattered, giving off secondary branches, which terminate in very flattened branch-systems, disposed in horizontal planes. These resemble in their general arrangement those of Thuya and Chamsecyparis, and are mostly tripinnate, all the axes being covered with small coriaceous leaves, adnate in part of their length, and arranged in four ranks in decussate pairs. The leaves on the main and ultimate axes differ only in size. The ventral and dorsal leaves are flattened and ovate or spathulate, with rounded apices ; the lateral leaves are carinate, more or less spreading, with a slightly acute apex, which is bent inwards. The dorsal flat leaves are shining green, and marked with a central ridge, which is often hollowed in the middle line. The ventral flat leaves have a central green ridge, with a concavity white with stomata on each side. The lateral leaves, green on the dorsal side, exhibit a single stomatic concavity on their ventral side. Flowers monoecious, solitary, and terminal, the male and female flowers borne on separate lateral branchlets as in Thuya. Male flowers cylindric, £ inch long, with six decussate pairs of stamens. Female flowers with five ovules on each scale. Cones globular, almost erect, with eight clavate, woody scales, in decussate pairs from a central axis, the upper pair abortive. Seeds three to five on a scale, laterally winged, the wing not notched at the summit. The seedling* resembles that of Thuya plicata, but has broader and very blunt cotyledons, with shorter and broader primary leaves. 1 See Tubeiif, Samen, Früchte, u. Keimlinge, 103, fig. 143 (1891). II 2OI 202 The Trees of Great Britain and Ireland THUJOPSIS DOLABRATA Thujopsis dolabrata, Siebold et Zuccarini, Fl. Jap. ii. 34, tt. ng, 120 (1842) ; Franchet et Savatier, Enum. PI. Jap. i. 469 (1875); Shirasawa, Icon. Essences Forest. Jap., text 27, t. xi. 18-34 (1900). Thuya dolabrata, Linnaeus, Suppl. PI. System, 420 (1781); Masters, Jour. Linn. Soc. (Bot.), xviii. 486 (1881), and Gard. Chron. xviii. 556, fig. 95 (1882); Kent, in Veitch's Man. Conif. 236 (1900). The species has been described in detail above. Two well-marked geographical forms occur, both confined to the main island of Japan:— ' ' 1. Var. australis (yar. nova). A small tree 40 to 50 feet in height, or a shrub growing as underwood in the dense shade of forests. As a tree it has a slender trunk, with drooping branches and a narrow pyramidal top. Branchlets very flat and only slightly overlapping, the lateral leaves ending in acute points bent inwards. Cones broadly ovoid, with scales thickened at the apex, which is prolonged externally into a blunt triangular process. This is the form which is known in cultivation in Europe, and described and figured in the works cited above. 2. Var. Hondai, Makino.1 A larger tree, attaining 100 feet in height, with a stem of over 3 feet in diameter. The branch-systems are more densely ramified, the branchlets being placed close together and overlapping one another by their edges more than is the case in the preceding variety. The leaves also are smaller, whiter underneath, and crowded more closely on the shoots ; those of the lateral ranks being usually blunt and not curved inwards at the apex. The cones are globular, with scales not thickened at the apex, which is devoid of the process so conspicuous in the other form, or merely shows it as an obsolete transverse minute mucro. The seeds appear to be more broadly winged, the wings being more scarious in texture. This form has not yet been introduced. Elwes has brought home excellent specimens of it in fruit from the Uchimappe Forest, near Aomori, in the extreme north of Hondo. These differ in the characters given above from specimens of the ordinary form obtained by him in the forest of Atera, Kisogawa, and Yumoto (4000 to 5000 feet altitude) in Central Hondo. The smaller leaves, set more closely on densely ramified branchlets in this variety, may be due to the influence of dense shade. The difference in the cone is paralleled by what occurs in the fruit of the different geographical forms of Cryptomera japonica. I am inclined to think that var. Hondai is not a distinct species ; but as it is very different, from the point of view of cultivators, it may conveniently bear the name Thujopsis Hondai. 1 Tokyo Botanical Magazine, 1901, xv. 104. Thujopsis 203 Several horticultural varieties have been introduced, viz. :— 3. Var. lœtevirens, Masters, Jour. Linn. Soc. (Bot.), xviii. 486. Thujopsis lœtevirens, Lindley, Gard. Chron. 1862, p. 428. Thoujopsis dolabrata nana, Gordon, Pinetum, ed. 2, p. 399. A dwarf shrub having no definite leader, with slender and much-ramified branchlets, and very small and bright green leaves. This variety often shows acicular leaves, spreading all round the shoot, and is apparently a fixed seedling form. It was introduced in 1861 from Japan by J. Gould Veitch. 4. Var. variegata. This only differs from the ordinary cultivated form in having the tips of many of the branchlets pale yellow or cream colour. It was introduced by Fortune in 1861. DISTRIBUTION Thujopsis dolabrata was discovered by Kaempfer,1 who mentions it in his Amcenitates Exotica, p. 884, as " a kind of Finoki." His specimen is still preserved in the Natural History Museum at South Kensington, and was figured by Lambert2 in his account of the species. ' Thunberg long afterwards (about 1776) sent specimens to Linnaeus, who first gave a scientific description of the tree. Thun berg8 cites the locality as follows :—" Crescit in regionibus Oygawœ et Fakoniœ, inter Miaco et ledo." (A. H.) Thujopsis dolabrata in Japan is known under the name of Hiba, and is found in a wild state north of about lat. 35°, and in the southern part of this area is a mountain tree only, occurring in the forest of the Kisogawa district from about 3000 to 5000 feet. In the vicinity of Nikko it is common between about 4000 and 6000 feet according to Sargent, but I only saw it here near Lake Yumoto where it did not appear to attain such large dimensions as farther north. The variety found in the forests of Atera is distinct in its fruit from the northern form. The excellent figure on Plate xi. in Shirasawa's Essences Forestières appears to be taken from the southern variety. The northern form has been described by Makino as var. Hondai, but the latter is not mentioned either by Goto or Shirasawa, nor is it recognised as specifically distinct in any of the Japanese collections which I saw. Though the tree usually occurs in mixture with Tsuga at Nikko, and with Sciadopitys at Atera, yet in the extreme north of Japan, on the hills north of Aomori, it is found in pure forest on hills of volcanic formation from near sea-level up to about 3000 feet. An excellent account of the forest of Uchimappe is given in Forestry and Forest-Products of Japan, where it is stated that the mountains are of Tertiary formation, and the under-lying rock composed of tufa, sandstone, and slate. Pieces of this rock which I brought home have been examined by Mr. Prior of the British Museum of Natural History, who considers that in all probability they represent a rather basic andésite or basalt, but owing to the weathered and decomposed state of the specimens, satisfactory sections could not be made. I visited this forest in the 1 See Salisbury, Jour. Science and Arts, 1817, ii. 313. z Genus Pinus, ed. 2, ii. tab. 68 (1842). 3 Flora Japonica, 266 (1784), sub Thuya dolabrata, Linn 204 The Trees of Great Britain and Ireland company of Mr. Shirasawa in June, and after passing through the flat rice-fields which extend from the sea to the foot of the hills, entered the forest, which consists mainly of Thujopsis naturally reproduced, though here and there, trees of Quercus glandidifera, Magnolia hypoleuca, and other species occur, whilst Cryptomeria and Cupressus obtusa are planted in the valleys, and Larix kptolepis on those parts of the hills where the natural forest has been destroyed by fire. From observations taken at the meteorological observatory of Aomori, it appears that the climate of this part of Japan is cold in winter and the snowfall heavy, the thermometer falling in February to -15° Centigrade, and rising in September to 32-5° Centigrade; the average temperature for the whole year being 9°, and the average moisture 78 per cent. The average height of the trees here is about 70 to 80 feet, attaining in deep shady valleys 100 feet or perhaps more, and about 2 feet in diameter when closely grown, at the age of 150 to 180 years when it is considered ripe for felling. The stems are often much curved at the butt from the pressure of the snow on the young seedlings, which require eight to ten years to get above its surface in winter, and these butts are usually cut separately and used for special purposes. The tree does not seem to have the power of reproducing itself from the stool, but produces abundant seed, which in dense shade germinates freely, though the growth of the seedlings is very slow at first. The undergrowth of the forest is very different from what I saw in other parts of Japan, bamboo-grass (Arundinaria Veitchit) being much less prevalent, but in the damp places tall herbaceous plants were numerous, with Aucuba, Skimmia, and Ilex, and other evergreen shrubs on the drier ground, and many pretty liliaceous plants and orchids in places. Goto says of this tree,1 that it formed under the old regime, together with Cupressus pisifera, C. obtusa, Thuya japonica, and Sciadopitys, the so-called " Goboku " or Five Trees, which enjoyed careful protection at the hands of the feudal autho rities ; he also says that it is rarely planted, being regenerated naturally by seed, and that it forms extensive forests in a mixture with other conifers such as Thuya japonica and Pinus parviflora, in the mountains on the northern frontier of the province of Rikuchu, in Goyosan, and in the mountains of the Tone districts, Kozuké. It has lately come to be in great demand for railway sleepers. Plate 60 (in Vol. I.) represents a dense growth of trees of this species in the forest of Uchimappe very similar to what I saw in the Kisogawa district at about 3000 feet. I am indebted to the Japanese Forest Department for the negative from which it was made. The wood of Thujopsis is highly valued in those parts of Japan where it grows, on account of its great durability. This is proved by specimens shown at the St. Louis Exhibition, one of which had been used as a gate-post for eighty-three years, another as a plank in a fishing-boat for eighty-four years, others as railway sleepers in use for fourteen years. The wood has an aromatic smell, takes a fine lustrous polish when planed, and is yellowish white in colour, showing a fine grain, which makes selected planks from the butt length very ornamental. Exceptional 1 Forestry of Japan, 18(1900). Thujopsis 205 cases occur in which the wood is curiously mottled and freckled. A ceiling and a screen made of such wood, which I saw in the Forestry Bureau at Aomori, were very beautiful. The wood weighs about 30 Ibs. per cubic foot, and is worth at Aomori from 40 to 50 yen per 100 cubic feet, or about is. per cubic foot. It is much valued not only for joinery and building purposes, but for foundations, ship and boat building, as it is stronger and more resinous than other woods of the same character. The bark also, which is thin, tough, and durable, is much used for roofing and for partitions and walls of out-houses, fences, etc. CULTIVATION T. Lobbsent a plant from the Botanical Garden at Buitenzorg in Java, to Exeter in 1853, which died; and soon after, Capt. Fortescue, a cousin of Earl Fortescue, brought a plant from Japan which was planted at Castlehill in 1859. But this tree, as I learn from Mr. Pearson, the head gardener, has been dead for some time, though plants raised from its cuttings are still growing at Castlehill and elsewhere. In 1861 Mr. J. G. Veitch and R. Fortune sent seeds from Japan to the Chelsea and Ascot Nurseries, from which plants were raised and generally distributed, so that the tree is now common in England. From what I have said of its habitat in Japan it is clear that though this tree is hardy as regards frost in winter, it requires conditions which are rarely found in England to bring it to any size, and, as a matter of fact, it has not yet become a tree anywhere except in Devonshire and Cornwall, though perhaps if seeds from North Japan are obtained the results might be better. Though no doubt it has ripened seeds elsewhere, 1 have never obtained any which germinated, except from a tree planted about 1881 by Queen Alexandra in the Earl of Northbrook's grounds at Stratton Park, Hants, which I gathered in October 1900. One of these grew, and is now a healthy plant about 9 inches high. It seems to suffer less from spring frost than many Japanese and Himalayan conifers. The finest tree that I have seen in England is at Killerton, which in 1902 measured 35 feet 6 inches in height and 2 feet 4 inches in girth. It is growing on a slope facing south-west in a peculiar soil, which Sir C. T. D. Acland describes as " Trap, soft below the surface, but hard after exposure. This trap overlies red sandstone, but is rather darker and more porous." This soil evidently suits most conifers admirably, as I have seen no other collection which contains so many fine specimens as this. At Boconnoc, at Carclew, and at other places in Cornwall there are trees approaching this in height, but we have not seen any specimen above 15 to 20 feet in other parts of England, though as a bushy shrub 12 feet high it exists in most modern gardens. In Scotland it seems hardy in the west and in Perthshire, whilst at Castlewellan in Ireland it has attained 30 feet in height. At Powerscourt and Kilmacurragh, Wicklow, there are trees with the lower branches layering and forming numerous independent stems. (H. J. E.) iESCULUS ^Esculus, Linnœus, Gen. PI. 109 (1737); Bentham et Hooker, Gen. PI. i. 398 (1862). Pavia, Boerhave, ex Miller, Gard. Dût. éd. 6 (1752). DECIDUOUS trees and shrubs, belonging to the natural order Sapindaceae, some authorities, however, making the genus the type of a distinct order Hippocastaneae. Leaves in opposite decussate pairs, without stipules, stalked, digitately compound ; leaflets five to nine, serrate in margin, pinnately veined. Branchlets stout, terete, with large triangular leaf-scars. Buds large, of numerous decussately opposite scales which are homologous with leaf-bases, the outer deciduous, dry or resinous, the inner accrescent and often brightly coloured. Flowers in large terminal racemes or panicles, appearing later than the leaves, of two kinds, hermaphrodite and staminate, on the same plant ; placed in the axils of minute caducous bracts on stout jointed pedicels. Calyx imbricate in bud, five- or two-lobed, the lobes unequal, united with an hypogynous annular disc in the hermaphrodite flowers. Petals four to five, imbricate in bud, alternate with the calyx lobes and inserted on the disc. Stamens five to eight, usually seven, inserted on the inner margin of the disc, unequal in length ; filaments filiform ; anthers two-celled, sometimes glandular at the apex. Ovary three-celled, rudimentary in the staminate flowers, each cell containing two ovules. Style slender, elongated, generally curved. Fruit a capsule ; prickly, roughened, or smooth ; coriaceous ; three-celled, three- seeded, and three-valved, or by abortion one- to two-celled and one- to two-seeded, the remains of the abortive cells and seeds usually remaining visible. Seeds without albumen, rounded or flattened by mutual pressure ; seed-coat brown and coriaceous, marked by a large whitish hilum. Cotyledons thick and fleshy, unequal, cohering together by their contiguous faces, remaining in the seed-coat during germination. About twelve species of ^Esculus * are known to occur in the wild state. They are natives of North America, Europe, and Asia. The genus was formerly divided into two sections, Pavia, with smooth fruit, and Hippocastanum, with spiny fruit ; but this division is not a natural one. The following synopsis groups the species under sections, which are more natural, being dependent on the characters of the flowers and buds :— I. HIPPOCASTANUM. Buds viscid. Calyx irregularly campanulate, four- to five- 1 The two Mexican species, which have tri-foliolate leaves, are now separated as a distinct genus, Billia. 2O6 -^Esculus 207 lobed. Petals four or five, claws not longer than the calyx; stamens exserted. This section includes all the old-world species. 1. j£Lsculus Hippocastanum, Linnaeus. Greece. 2. sEsculus indica, Colebrooke. Afghanistan, north-western Himalaya. 3. ^Escuhis punduana, Wallich, List 1189 (1828). Sikkim, western Duars, Khasia Hills, Upper Burma, Tenasserim, Siam, Tonking. Large tree. Leaflets six to seven, very large, thinly coriaceous, stalked, acuminate, serrate. Panicles 12 to 15 inches or more, flowers white or yellow. Fruit brown, smooth. Not introduced and not likely to be hardy. 4. jfLsculus ckinensis, Bunge, Enwn. PI. Chin. Bor. 10 (1835). Northern and Central China. A tree, 40 to 50 feet high. Leaflets five to seven, large, stalked, obovate-oblong, rounded at the base, abruptly acuminate at the apex, finely serrate, shining above, glabrescent below except for pubescence along the nerves, petioles pubescent. Panicles, 8 inches long, pubescent. Flowers small, white ; sepals shortly and unequally five-lobed, pubescent. Petals four, minute. Filaments glabrous. Fruit * pear-shaped or globular, small (f inch diameter), one-celled, three- valved, brown, covered with warts, not spiny. This species has been much confused with the next, from which it differs in every way. The flowers, though small, are numerous in the large panicle, and the foliage is very handsome. It is common enough in the mountains of central China, in Shansi, and in the hills to the west of Peking ; and when introduced is likely to prove hardy in England. 5. ^Isculus turbinata, Blume. Japan. II. PAVIA. Buds not resinous. Calyx tubular, five-toothed. Petals four, yellow or scarlet. 6. sEsculus glabra, Willdenow. North America. 7. Aïsculus octandra. Marshall. North America. 8. /Esculus Pavia, Linnaeus, Sp. PI. 344 (1753); Bot. Reg. t. 993 (1826). Middle United States. A shrub. Leaves with slender grooved petioles, the edges of the grooves jagged. Leaflets five, obovate, acute at the base, acuminate at the apex, finely serrate without cilia, slightly pubescent beneath. Flowers in loose panicles, 4 to 7 inches long. Petals red, meeting at the tips ; upper pair longer, with claws about three times as long as the small spathulate limb; lateral pair shorter, with claws as long as the calyx, and rounded limb equalling the claw in length ; margin of petals beset with minute dark glands. Stamens as long as the upper pair of petals. Fruit brown, without spines. This species, though only a shrub, is mentioned here at some length, as it closely resembles sEsculus octandra, and moreover enters into such important hybrids as tâsculus carnea, versicolor, etc. All its hybrids may be recognised by the red colour of the flowers and the glandular margin of the petals. It is readily distinguished from ^sculus octandra by its smaller leaves and peculiar petioles. In winter it shows the following characters :—Twigs slender, glabrous, shining, with numerous lenticels. 1 Cf. Hance mjourn. Bot. viii. 312 (1870), 1 2,08 The Trees of Great Britain and Ireland Leaf-scars obovate or crescentic on slightly prominent cushions, with three groups of bundle-dots ; opposite scars joined by a linear ridge. Terminal buds long oval or fusiform, pointed ; scales numerous, the upper rounded, the lower pointed at the apex and keeled on the back, minutely ciliate in margin. Pith wide, circular, green. 9. jEsculus aiistrina, Small, Bull. Torrey Bot. Chib, 1901, xxviii. 359 ; Sargent, Man. Trees N. America, 647 (1905); jfâscuhts Pâma, ß discolor, Torrey and Gray, FI. N. Amer. i. 252 (1838), in part. A small tree, attaining 30 feet in height, occurring in Tennessee, S. Missouri, E. Texas, and north-western Alabama. This resembles the last species. The leaflets, however, are usually more irregularly but finely serrate, and pale tomentose beneath. Panicles pubescent, 6 to 8 inches long. Petals bright red, meeting at the tips, unequal, oblong-obovate, rounded at the apex, glandular, those of the upper pair about half as wide as those of the lateral pair, with claws much longer than the calyx. Stamens longer than the petals. Fruit brown, slightly pitted. Not introduced. III. MACROTHYRSUS. Buds not viscid. Calyx five-toothed. Petals four to five, white, claws longer than the calyx. Stamens exserted, very long. ID. sEscîtlus parvißora. Walter, Flora Caroliniana, 128 (1788). South-eastern North America. A shrub. Leaflets five to seven, elliptical or oblong-ovate, densely grey-tomentose beneath, finely serrate. Panicles erect, 8 to 10 inches long, slender, narrow. Flowers white, faintly tinged with pink. The long and thread like stamens are pinkish white and very conspicuous. This is a valuable shrub, as it flowers late, in July or August, some five or six weeks later than any of the other species except californica. Occasionally it forms a short single trunk, but generally it sends up a crowd of stems from the ground. It is figured in Gard. Chron. 1877, via. fig. 129; and is often known in gardens as Pavia macrostachya, Loiseleur, or sEscuhts macrostachya, Michaux. See Bot. Mag. t. 2118 (1820), where it is stated that the species was introduced by Mr. John Fräser in 1785. Canon Ellacombe reported in 1877' that he had at Bitton a specimen, which was at least forty years old, but that it remained a bush, not exceeding 8 or 10 feet in height. IV. CALOTHYRSUS. Buds viscid. Calyx two-lipped or five-lobed. Petals four, pink or white, claws not longer than the calyx. Stamens exserted. 11. ^sculus californica, Nuttall. California. 12. jfëscïihis Parryi, A. Gray, Proc. Amer. Acad. xvii. 200 (1881); Sargent, Garden and Forest, 1890, p. 356, fig. 47. Lower California. A small shrub, resembling the preceding species ; but differing in the five-lobed calyx, and in the leaflets, which are small, obovate and hoary pubescent beneath. It has not been introduced. V. HYBRIDS. The most important is sEsculus carnea, Hayne, which is a cross between the common horse-chestnut and A. Pavia. This is described fully below. 1 Card. Chrcn. 1877, v"i- *>9'- ^Esculus 2,09 plantierensis, André, a supposed hybrid between sEsculus carnea and Aîsculus Hippocastamtm, will be mentioned under the former species. sEscttlus versicolor, Dippel, a hybrid between sEsculus Pavia and sEsculus octandra, will be treated under the latter species. The following key to the species in cultivation is based on the characters of the leaves and buds. In Plate 61 the leaves of all these species are shown ; and in Plate 62 are represented the twigs and buds of six species, viz., Hippocastanum, carnea, indica, glabra, octandra, and californica :— A. Leaflets sessile or nearly so ; buds very viscid. 1. /Esculus Hippocastanum. Petioles glabrescent. Leaflets obtusely and irregularly serrate. 2. sEscîtlus turbinata. Petioles pubescent, especially towards their tips. Leaflets regularly and crenately serrate. B. Leaflets stalked. *Buds viscid. 3. sEsculus indica. Leaflets finely and sharply serrate, pale beneath. Buds very viscid. 4. sEscuhts carnea. Leaflets obtusely and irregularly serrate. Buds only slightly viscid, the brown scales having a dark-coloured margin. 5. sEscuhts californica. Leaflets shallowly and crenately serrate, pale beneath. Buds viscid, glisten ing with white resin. ** Buds not viscid. 6. sEscuhts parvißora. Leaflets densely grey-tomentose beneath, finely serrate in margin. Buds minutely pubescent. 7. sEsculus octandra. Leaflets pubescent beneath, broadly lanceolate, shortly acuminate, with twenty or more pairs of nerves in the terminal leaflet ; margin finely serrate but not usually ciliate. Petioles without jagged marginal ridges. 8. sEsculus glabra. Leaflets glabrous beneath, except for a slight pubescence along the midrib and tufts in the axils, long-acuminate, with about fifteen pairs of nerves in the terminal leaflet, finely serrate with ciliate tufts in the bases of the serrations. Petioles with smooth marginal ridges. 9. ^Escîihts Pavia. Leaflets slightly pubescent beneath, narrowly lanceolate, finely serrate but not ciliate in margin. Petioles flattened on the upper side, with marginal sharp ridges, usually jagged. ii 210 The Trees of Great Britain and Ireland ^SCULUS HIPPOCASTANUM, COMMON HORSE-CHESTNUT Hippocastanum, Linnaeus, Sp. PI. 344 (1753); Loudon, Arb. et Frut. Brit. i. 462, iv. 2543 (1838); Gard. Chron. 1881, xvi. 556, figs. 103, 104. A large tree, attaining in England a height of over 100 feet and a girth of 15 or even 20 feet. Bark smooth and dark brown in young trees, becoming greyish and fissured longitudinally in old trees, at the same time scaling off in thin plates. Leaves palmately compound, digitate, on a long stalk widened at its insertion. Leaflets five to seven, sessile, obovate, cuneate at the base, abruptly acuminate at the apex, unequally and coarsely serrate ; green above ; beneath pale, tomentose at first, but ultimately glabrous, except for small tufts of hairs in the axils of the veins and a few scattered hairs over the surface ; middle leaflet the largest, with twenty-four or more pairs of nerves, lower pair smallest ; venation pinnate ; petiole glabrous. The leaflets as they emerge from the bud are at first erect, but soon bend downwards on their stalks. When nearly full grown they rise up and become horizontal. In autumn they turn yellow or brownish and fall early, each leaflet disarticulating separately from the petiole. Flowers in large upright pyramidal panicles, the primary branches of which are racemose, the lateral branches cymose. Upper flowers staminate and opening first ; lower flowers hermaphrodite. Calyx greenish, five-toothed. Petals four to five, crumpled at the edge, white, with yellow spots at the base, which ultimately become pink. Stamens seven, longer than the petals, the filaments bent down when the flower opens and the stigma protrudes, later moving up on a level with the style. Fruits few on each panicle, large, globular, green, with stout, thick conical spines, three-valved, usually one-seeded, occasionally two- to three-seeded. Seed large, shining-brown, with a broad whitish hilum. Cotyledons two, large, fleshy, distinct below, blended into one mass above. SEEDLING 1 The cotyledons are large and fleshy and remain in the seed, which frequently germinates on the surface of the soil or only slightly buried beneath it. The cotyledons have long petioles (f-i inch), which are broad and flattened, with a concavity on their inner surface. The caulicle, very variable in length ( i to 4 inches), is stout, brownish, pubescent, and ends in a stout tap-root, which gives off numerous branching fibres. The young stem is stout, terete, brownish, striated and marked with numerous lenticels, puberulent or glabrous ; it has no scale-leaves, differing in this respect from the young stem of the oak. In other respects the germination of the oak and of the horse-chestnut are almost identical. At a varying height 1 Cf. Lubbock, Seedlings, i. 356 (1892), where it is stated that the seed is carried a considerable height above ground during germination owing to the great length of the caulicle. So far as I have observed, the seed does not change its position during germination. ./Esculus 211 above the cotyledons the first pair of true leaves are produced, which are opposite, compound, digitately five-foliolate, and closely resemble the adult foliage except that they are smaller in size. Successive pairs of similar leaves follow on the stem, each pair being placed decussately with reference to the pair immediately below it. ABNORMAL FLOWERING The horse-chestnut sometimes produces a second crop of flowers in autumn, which appear in much smaller panicles than those of spring. This is due to the premature fall of the leaves in July or August, usually following an excessively dry season. The buds are stimulated to premature energy and put forth young leaf-shoots, which are terminated by flowers. This phenomenon, which is equivalent to an anticipation of the opening of the buds by several months, as they would normally open in the following spring, is frequently observed in the trees planted in the boulevards of Paris.1 In the dry season of 1884, a single tree at Kew produced small panicles of flowers in September, after previously shedding nearly all its leaves. In the following year it produced a few panicles of the ordinary size. At Hythe,2 near Southampton, a horse-chestnut is reported to have bloomed and fruited three times in 1868, once in spring, again after the rain which succeeded the long drought, and a third time in September. IDENTIFICATION In summer the common horse-chestnut is unmistakable. The only other species with large sessile leaflets, ^Esculus turbinata, is easily distinguished by their regular crenate serration. In winter the twigs and buds show the following characters :—Twigs stout, brown, glabrous or minutely pubescent towards the tip ; lenticels numerous. The large opposite leaf scars, flat on the twigs with no prominent cushion, are joined by a linear ridge, and vary in shape, the larger being obovate with seven bundle-dots, the smaller semicircular or crescentic with usually only five dots. Buds very viscid, larger than in the other cultivated species ; the terminal much exceeding the lateral buds in size, occasionally absent, and replaced by the saddle-shaped scar of the previous year's inflorescence ; scales imbricate, the external ones in four vertical ranks, rounded at the apex, glabrous, not ciliate, dark red-brown. The buds contain the next year's shoot in an advanced state of development, flowers being visible in them in October. The scales are morpho logically equivalent to leaf-bases. In the interior of the bud, scales are observable with traces of leaf-blades, which gradually pass into the true leaves, visible in the upper part of the bud. VARIETIES i. Var. flore pleno, Lemaire, Illust. Horticole, 1855, ii. t. 50. A variety with double flowers, the pistil even in some cases becoming petaloid. Mr. A. M. 1 See article by Roze, translated in Gard. Chron. 1898, xxiii. 228. 2 Gard. Chron. 1868, p. 1116. 212 The Trees of Great Britain and Ireland Baumann discovered in 1822, near Geneva, a horse-chestnut tree, of which a single branch bore double flowers ; and from this branch the variety was propagated at the Bollweiler nursery in Alsace.1 The flowers last longer than those of the single kind,2 and no fruits are formed, which renders it useful as a tree in streets, where the fall of fruits is an inconvenience. This variety is very hardy, and resisted well the severe winter of 1879-80 in France.8 2. Var. laciniata (var. asplenifolia, var. incisa). Leaflets cut up into narrow lobes. According to Beissner4 this variety has been in cultivation for over forty years ; and a form of it was found by Herr Henkel of Darmstadt, which keeps its foliage much longer than the typical form ; but this is not the case in some localities. 3. Var. crispa. Leaves short-stalked, with broad leaflets. Tree compact in habit. 4. Var. pyramidalis. Upright in habit. 5. Var. umbraculifera. Crown densely branched, and globular in outline. 6. Var. tortuosa. Branches bent and twisted. 7. Var. Memmingeri. Leaves yellowish in colour, looking as if powdered with sulphur. 8. Var. aureo-variegata. Leaves variegated with yellow. Several other varieties of slight interest, which do not seem to be in cultivation in this country, are mentioned by Schelle.6 DISTRIBUTION AND HISTORY The horse-chestnut occurs wild in the mountains of northern Greece. Halacsy,6 the latest authority, gives many localities in Phthiotis, Eurytania, Thessaly, and Epirus ; but states that it is not found wild on Mount Pelion or in Crete. Baldacci,7 in 1897, found the tree growing wild on almost inaccessible precipices below the lower limit of the coniferous belt near Syrakou in the district of Janina in Albania. The native country of the tree was long a matter of doubt ; but the whole question was satisfactorily elucidated by Heldreich8 in a paper, from which we extract most of the following account. Linnaeus considered the habitat of the tree to be northern Asia, and De Candolle thought that it came from northern India. The tree is, however, not known wild in India, where it is replaced by ^Sscu&s indica. Boissierfl states that it is recorded from Greece by Sibthorp, from Imeritia (Caucasus^by Eichwald, and from Persia by various authors. It is, however, unknown in the wild state in Persia ; and Radde10 mentions it only as a planted tree 1 Rev. Belgique Horticole, 1854, iv. 216. 2 See Garden, 1890, xxxviii. 601, where some observations are recorded on the periods of flowering of the single and double horse-chestnuts, and of sEsculus cornea. 3 Rev. Horticole, 1884, p. 98. 4 Mitt. Deut. Dendrol. Gesell. 1905, pp. 13, 14. and 1906, p. 10. 6 Handbuch Laubholz-Benennung, 321 (1903). e Cotisp. Fl. Grceca, i. 291 (1900). 7 Rivista Collez. Botan. in Albania, 23 (Florence, 1897). 3 Verlumd. Bot. Vereins Prov. Brandenburg, 1879, P- 139- The British Minister at Athens, Sir F. E. H. Elliot, K.C.M.G., who kindly made inquiries, has sent us a letter from Professor Miliarakis of the University of Athens, dated April 2, 1904, which confirms Heldreich's statements. 9 flora Orientalis, i. 947 (1867). I0 Pßatizenverbreitung in Kaukasusländern, 433, 434 (1899). ^sculus 213 in the Caucasus. All the evidence goes to show that it is confined to northern Greece and Albania. Heldreich states that the horse-chestnut was first found wild in Greece by Dr. Hawkins.1 In his own travels in Greece in 1897 ne observed it in many stations, all lying in the lower fir region, between 3000 and 4000 feet altitude, where it grows in shaded moist gulleys, in company with alder, walnut, plane, ash, several oaks, Ostrya carpinifolia and Abies Apollinis. These stations, situated in remote uninhabited spots, establish the fact that the tree is really wild. Plants introduced into Greece by the Turks are always found in the neighbourhood of towns. Whether the tree was known to the ancient Greeks is doubtful. The horse-chestnut was first mentioned2 by the Flemish doctor Quakleben, who was attached to the embassy of Archduke Ferdinand I. at Constantinople,—in a letter to Matthiolus in 1557. The latter received a fruit-bearing branch, and pub lished the first descriptions of the tree as Castanea equina, because the fruits were known to the Turks as At-Kastane (horse-chestnut), being useful as a drug for horses suffering from broken wind or a cough. The tree was introduced into western Europe from Constantinople, the first tree being raised by Clusius at Vienna from seeds sent by the Imperial Ambassador, D. Von Ungnad, in 1576. This tree quickly grew, and was mentioned by Clusius4 in 1601. The horse-chestnut was introduced into France6 in 1615 by Bachelier, who brought the seeds from Constantinople. Gérard mentions it in his Herbal of *579i P- 1254> as a tree growing "in Italy and sundry places of the eastern countries"; and in Johnson's edition of this work, published in 1633, the tree was stated to be growing in Tradescant's garden at South Lambeth. It was probably introduced into England about the same time as into France. (A. H.) CULTIVATION No tree is easier to raise from seed than the horse-chestnut. Its large fleshy fruit are so little hurt by frost and damp that they germinate freely where they fall, and do not seem to be eaten by mice like acorns and beech-mast. Seeds which have been exposed all winter germinate more readily in spring than those which have been kept dry, and should be sown early and covered with about two inches of soil. Though it is advised by French writers that the extremity of the radicle should be pinched off before sowing in order to prevent a strong tap-root from forming, as is done in the case of walnuts and chestnuts, I have not observed that they suffer from removal if this is not done ; and if transplanted at one or at latest two years after sowing there are abundance of fibrous roots which make the tree an easy one 1 Sihthorp et Smith, Fl. Graca Prodromus, i. 252 (1806). Hawkins' observation has been disputed, as he records it from Pelion, where the tree does not, so far as we know now, occur wild. Orphanides was the first to establish beyond doubt that the tree is indigenous to the mountains of northern Greece. Cf. Grisebach, Vegetation der Erde, French ed. i. 521. 2 Matthiolus, Epistol. Medicin. Libri Quincue (Prague, 1561). 3 Matthiolus, Comment, in Dioscorid. Mat. Med. 211 (Venice, 1565). 4 Clusius, Rar. Plant. Hist. J (1601). 6 Tournefort, Relation d'un Voyage au levant, i. 530 (»717)- 214 The Trees of Great Britain and Ireland to move, even when five or six feet high. As the tree is liable to form large side branches, the buds should be rubbed off the stem early in order to form a clean trunk, though it bears pruning well as a young tree. Though somewhat liable to suffer from cold winds and spring frost, which injure the foliage and flowers, the tree is hardier in this respect than many of our native trees, though coming from a warm southern country. As regards the chemical nature of the soil it is quite indifferent, for though it grows faster on a good loam and does not come to perfection on sandy soil, it attains a large size on dry, rocky, calcareous soils, and even at an elevation of 800 feet and upwards resists wind better than many trees. I have seldom seen horse-chestnuts blown down, though large heavy branches are often torn off by violent winds. As an ornamental flowering tree for parks, lawns, and avenues it has no superior, though on account of its branching habit it requires considerable attention in order to form tall shapely trees. Its principal defect is the tendency of the leaves to become brown and ragged early in the autumn, but they fall quickly, and being easily removed make less litter than the leaves of the beech, oak, or sycamore. The large branches when allowed to rest on the ground in damp situations frequently take root and become naturally layered, the best instance of this that I have seen being at Mottisfont Abbey, Hants. For town planting, on account of its beautiful flowers and dense shade during the hottest months, the horse-chestnut is perhaps, next to the plane, one of the best trees we have, and does not seem to suffer much from smoke. In parks it is valuable for its fruit, which are so much liked by deer that they are eaten as fast as they fall, and would perhaps be worth collecting for winter food. The extraordinary hardiness of this southern tree is proved by the fact that it will grow to a large size as far north as Trondhjem in Norway, lat. 63° 26', a tree figured by Schubeler near this place being 37 feet by 8 feet 9 inches. Another in the Botanic Garden at Christiania, which is considered the largest in Norway, measured in 1861, 16.62 metres by 2.45 metres, and when I saw it in 1903 had increased to no less than 28 metres high by 3 in girth, though it has been exposed to as low a temperature as — 18° to — 20° Reaumur. As regards the age which the horse-chestnut attains we have few exact records, but it does not seem a very long-lived tree. J. Smith states1 that an avenue running south-east from the front of Broadlands House, near Romsey, Hants, was planted in 1735; but in 1887 only two trees remained, which were u feet and 12 feet 4 inches in girth. REMARKABLE TREES There are so many fine trees in almost every part of Great Britain that I need not go into great detail as to their dimensions, but though it is possible that in Bushy Park, or other places near London, taller trees exist, I have only at 1 Trans. Scot. Art. Sac. xi. 540 (1887). ./Esculus 215 Petworth measured one which exceeds in height the group of three which grow near my own house at Colesborne, of which I give an illustration (Plate 63). The height of these as measured in 1902 by Sir Hugh Beevor and myself was 105 feet, and the girth of the largest 11 feet. They grow in a sheltered situation, on damp, cold soil. One of these trees being inclined to split at the base, owing to the great weight and length of one of its principal limbs, was chained up many years ago, and though the iron band which was put round it has become buried in the wood the limb has not broken off. At Dynevor Castle, Carmarthenshire, the seat of Lord Dynevor, where the park contains a greater number of fine trees than any I have seen in South Wales, there is a very large tree which the Hon. W. Rice measured in 1906 and found to be 109 feet by 17 feet 9 inches. For height and girth combined this seems to be the largest tree in Great Britain. The tallest tree I have seen is in a grove of beech, chestnut, oak, and silver fir, which grows near the house at Petworth Park, the seat of Lord Leconfield in Sussex, on a deep greensand formation. This tree, though forked at six feet from the ground, has been drawn up to a great height by the trees surrounding it, and though difficult to measure exactly, probably exceeds 115, and may be 120 feet. The two stems are 9 feet 8 inches and 8 feet respectively in girth. In Bushy Park most of the horse-chestnuts are past their prime ; many of the old trees are dead and have been replaced by young ones. The largest, seen in 1906, was growing near the gate ; it had a bole of 20 feet giving off four great stems, and measured 100 feet high by 16 feet 5 inches in girth. Another near the pond was ici feet by 16 feet i inch. At Birchanger Place, near Bishop Stortford, the seat of T. Harrison, Esq., there is one of the largest and finest trees in England, which measures about 80 feet by 20 feet, with a bole about 15 feet high and a spread of 32 yards; a beautiful photograph was taken in 1864 when the tree was in flower, but it is now partially decayed on the north side, and has lost some large branches. At West^Dean Park, Sussex, the seat of W. D. James, Esq., there is a large tree about 70 feet by 16 feet, with branches spreading over an area no less than 36 yards in diameter. At Hampton Court, Herefordshire, the seat of John Arkwright, Esq., there is a very fine tree growing on deep alluvial soil in the big meadow south of the house. Measured by T. Hogg in iSSi1 it was 93 feet by 16 feet 6 inches. When I saw it in 1905 it had increased about three feet in height and was 18 feet 7 inches in girth, and still handsome and vigorous. The largest trees I have seen as regards girth and spread of branches are in Ashridge Park, on a bank near the lodge on the Berkhampstead road. The largest of these is about 80 feet high and 20 feet in girth, with extremely wide-spreading branches, and there are several others of 16 to 17 feet girth in the row. These trees are growing on a dry, flinty, calcareous loam. 1 Trans. Scot. Arb. Sac. ix. p. 151 (1886). ai6 The Trees of Great Britain and Ireland There is a fine tree at Syon, which in 1905 was 93 feet high by 15 feet 4 inches in girth ; and at Broom House, Fulham, there is a tree 95 feet high. In the courtyard at Burleigh, near Stamford, the seat of the Marquess of Exeter, there is a large and very beautiful tree, figured by Strutt, plate 37, which was in 1822 60 feet high by 10 feet in girth, with a spread of 61 feet diameter. When I saw it in 1903 it was still in perfect health, and was about 80 feet by 12 feet 6 inches. It had remarkably spiny fruit, and its trunk was covered with small twigs. At Trebartha Hall, near Launceston in Cornwall, Mr. Enys reports in 1904 a tree 15 feet 6 inches in girth, with an estimated height of 70 feet. In Scotland the horse-chestnut seems as much at home as in England, and thrives in most places as far north as Gordon Castle, where there is a tree, measured in 1881 by Mr. Webster, 65 feet high by 13 feet 4 inches in girth, and 274 feet in circumference of its branches. At Newton Don, Kelso, the seat of Mr. C. B. Balfour, there is a tree which was in 1906, 13^ feet in girth with a spread of branches of 165 feet in circumference. In Perthshire there is a very beautiful tree, remarkable for its weeping habit, in the park at Dunkeld, which measures 80 feet in height by 17 feet 6 inches in girth (Plate 64). At Kilkerran, Ayrshire, Mr. J. Renwick has measured a fine tree 84 feet high by 14 feet in girth, with a bole 22 feet high. At Pollok, near Glasgow, a tree measured, in 1904, 63 feet high by 13 feet 6 inches girth at 2^ feet from the ground, with a bole of 5 feet, giving off four great stems. None of these are equal to a tree in a group of seven standing at the west end of Moncreiffe House in Perthshire, which Hunter1 describes as the largest in Scotland, and which then measured 19 feet in girth at five feet from the ground. At ten feet it divides into three great limbs, one of which has become firmly rooted in the ground, and extends so far from the trunk that the total spread of the tree is 90 feet in diameter. The remarkable hardiness of this tree is shown by the existence of one, reported by Mr. Farquharson of Invercauld, as growing at an elevation of 1110 feet, which was supposed to be 177 years old in 1864, when it was 8 feet 7 inches in girth.2 In Ireland the horse-chestnut attains a great size, the largest we know of occurring at Woodstock in Co. Kilkenny, on an island in the River Nore. One tree measured in 1904, 93 feet in height by 18 feet i inch in girth, and according to the careful records which have been kept of the growth of the many fine trees on this property, measured in 1825, 10 feet 2 inches in girth ; in 1846, 13 feet 2 inches; in 1901, 17 feet 9 inches. Another about the same height, in a meadow near the river, measured in 1825, n feet in girth; in 1834, 12 feet; in 1846, 12 feet n inches; in 1901, 14 feet 4 inches. Woods and Forests of Perthshire, 1883. Old and Remarkable Trees of Scotland, p. 115. 217 TIMBER The wood of the horse-chestnut is one of the poorest and least valuable we have, on account of its softness and want of strength and durability. Though it has a fine close and even grain, white or yellowish-white colour, and is not liable to twist or warp so much as most woods, it does not cut cleanly, decays rapidly, and is only used as a rule for such purposes as cheap packing-cases and linings. It burns so badly that it is of little use as firewood, and though occasionally cut into veneers or used as a cheap substitute for sycamore, poplar, and lime, in making dairy utensils, platters, and brush backs, it cannot be said to have a regular market. From 4d. to 8d. a foot is about the usual value in most parts of England, though Webster says that it was worth a shilling in Banffshire some years ago. Holtzapffel says that it is one of the white woods of the Tunbridge turner, a useful wood for brush backs and turnery, preferable to holly for large varnished and painted works on account of its great size. I am not aware whether it has been tried for pulp-making, but it would seem to be a suitable wood for that purpose on account of its softness, and could, if required, be produced in quantity at a low price. (H. J. E.) ^ESCULUS CARNEA, RED HORSE-CHESTNUT Aisculus carnm, Hayne, Dendrol. Flora, 43 (1822). ;£sculus rubicunda, Loiseleur, Herb. Amat. vi. t. 357 (1822); Loudon, Arb. et Frut. Brit. i. 467 (1838); Carrière, Rev. Horticole, 1878, p. 370, coloured figure of var. Briotii. sEsculus Hippocastanum, L. x ^Esculus Pavia, L., Koch, Dendrologie, i. 507 (1869). A small tree, occasionally 50 feet, but rarely exceeding 30 feet in height. Leaves resembling those of the common horse-chestnut, but darker green with an uneven surface, the leaflets being shortly stalked and more or less curved and twisted. Flowers red, showing as they open an orange-coloured blotch at the base of the petals, which afterwards becomes deep red. Petals five, standing nearly erect, their limbs not spreading horizontally at right angles to the claws, as occurs in the common horse-chestnut ; edges of the petals furnished with minute glands, like those present in sEsculus Pavia. Fruits with slender prickles. IDENTIFICATION In winter, the species is distinguished as follows : — Twigs rather stout, grey, shortly pubescent ; leaf-scars as in jZLsculus Hippocastamim. Buds slightly viscid and smaller than in that species ; scales brown, edged with a dry membranous dark-coloured rim. Lateral buds small, oval, pointed, arising from the twig at an acute angle. II 218 The Trees of Great Britain and Ireland VARIETIES 1. Van Briotit. Flowers in larger panicles and more brilliantly coloured, the filaments, calyx, and style being red. Fruits never developing fully, falling soon after the flowers. This variety1 was obtained in 1858, by M. Briot at the State Nurseries of the Trianon, Versailles, as a seedling of sEscuhis carnea. 2. Several variegated forms are known, as var. a^treo-mac^tlata and aureo- marginata. Var. alba is a form with white flowers, \zx.pendula is pendulous in habit. 3. j&Lscuhtsplantierensis, André, Rev. Horticole, 1894, p. 246, is supposed to be a cross between A. carnea and the common horse-chestnut, as it is intermediate in character. This variety arose in the nursery of Messrs. Simon-Louis Frères at Plantières-lès-Metz, from a seed of sEsciifas Hippocastanum. Other intermediate forms, named by André s£sc^fl^ts intermedia and jfâsculus balgiana, were derived from seeds of SEsathts carnea. HISTORY Nothing is known for certain concerning the origin of sEsculus carnea. Loise- leur received the plant from Germany in 1818, and there are no earlier accounts of it. Its parentage, however, is undoubted : it possesses characters of both the supposed parents. The leaves and slightly spiny fruit are derived from the common horse-chestnut. The colour of the petals and the glands on their margins come from ^Esc^^lus Pavia. According to André2 the seeds when sown usually produce plants which bear whitish flowers and are of no horticultural value. The species is accordingly always propagated by grafting. Koch,8 however, reports that while some seedlings are like those of the common horse-chestnut, others produce smooth fruits. At Kew, according to Mr. Bean, it has come true from seed. The largest specimen of this tree that we have seen occurs at Barton in Suffolk. It was 50 feet high in 1904, with a bole, however, of only 2 feet, girthing 7 feet 9 inches at a foot above the ground, and dividing into three stems. It does not seem to live long or to attain any great size in England, and is often supposed to be a red-flowered form of the common horse-chestnut. (A. H.) 1 Rev. Hort. loc. cit. 2 André, Rev. Hort. loc. cit. 3 Verhand. Ver. Beförd. Gart. König. Pieuss. Staat, 1855. /Esculus 21 ^ÎSCULUS INDICA, INDIAN HORSE-CHESTNUT indica, Colebrooke, VVallich, List 1188 (1828); Bot. Mag. t. 5117 (1859); Hiern, in Flora British India, i. 675 (1875); Bean, in Gard. Chron. 1897, xxii. 155 and 1903, xxxiii. 139, Suppl. Illust.; Collett, Flora Simla, 97 (1902); Gamble, Man. Indian Timbers, 193 (1902); Brandis, Indian Trees, 185, 705 (1906). Pavia indica, Wallich, ex Jacquemont, Voyage dans l'Inde, iv. 31, t. 35 (1844). A large tree, attaining in India 150 feet in height and 40 feet in girth of stem. Bark in old trees peeling off in long strips. Leaves large, glabrous, dark green above, pale, almost glaucous beneath ; leaflets five to nine, stalked, obovate- lanceolate, acuminate, finely and sharply serrate, with about twenty pairs of nerves in the terminal leaflet. Panicles 12 to 15 inches long, loose, narrow, erect. Flowers large, about i inch long ; calyx ^ inch long, irregularly lobed, often splitting so as to appear two-lipped. Petals four, white, of two unequal pairs ; the upper pair narrow and long with a red and yellow blotch at the base, the lower pair flushed with pink. Stamens seven or eight, scarcely longer than the petals, spreading. Fruit brown, rough, without spines, irregularly ovoid, one to two inches long, containing one to three dark brown shining seeds. IDENTIFICATION In summer the viscid buds and the large stalked leaflets with finely serrate margins distinguish it from the other species in cultivation In winter the twigs show the following characters : — Branchlets coarse, shortly pubescent ; lenticels like brown raised warts, numerous; pith circular, white; leaf-scars on slightly prominent cushions, each pair wide apart and joined by a raised linear ridge, obovate or semicircular with a raised rim and three groups of bundle-dots. Buds viscid, greenish, the lower scales only being brown ; terminal buds ovoid, pointed, the two lowest scales having projecting beaks ; scales not ciliate, the outermost four pubescent ; lateral buds small, arising at an acute angle. DISTRIBUTION It is a common tree in the north-west Himalayas from the Indus to Nepal, occurring at elevations of from 4000 to 10,000 feet, and also occurs in Afghanistan. Sir George Watt informs me that he has measured many trees 150 feet in height with trunks of enormous size, a girth of 40 feet not being uncommon. The wood is used in building and for making water-troughs, platters, vases, cups, packing-cases, and tea-boxes. The twigs and leaves are lopped for use as fodder. The fruit is given as food to cattle and goats ; ground and mixed with ordinary flour, it is part of the dietary of the hill tribes. The bark of old trees is very remarkable in appearance, exfoliating in long flakes, which remain attached at their upper ends and hang downwards and outwards. (A. H.) 220 The Trees of Great Britain and Ireland CULTIVATION Colonel Henry Bunbury brought seeds from India in 1851, from which plants were raised by Sir Charles J. F. Bunbury1 at Barton in Suffolk. The large tree2 now flourishing on the lawn at Barton (Plate 65) is one of the original seedlings, and measured, in 1904, 66 feet high by 7 feet 9 inches in girth. Another tree in the arboretum at Barton measured 65 feet high by 7 feet 2 inches in girth ; and divides into two main stems at 7 feet above the ground. This tree flowered for the first time in 1858, producing twelve panicles, being then only seven years old from seed, and 16 feet in height. It did not suffer in the least from the terrible winter of 1860, and flowered as usual in the summer following. In 1868 it ripened fruit, and four thriving plants were raised from its seed. There are no records of the tree on the lawn, which is now the finer of the two. Other trees were planted apparently at Mildenhall,3 which is about fifteen miles distant from Barton ; but these never throve, and none remain. The soil at Mildenhall is a light loam on chalk, and probably did not suit the tree. I saw the beautiful tree at Barton in full flower on June 24, 1905, when it did not seem to have received the least injury from the severe frosts and cold north-east winds which had occurred a month previously, and which ruined the flowers and destroyed the fruit of the common horse-chestnut in many places. It seems incredible that this species should be so rare and have remained so little known in England, where it ought to be planted generally in the south and west. Mr. Bean says that the seeds soon lose their vitality if kept dry, and that of some scores received in ordinary paper packets from India in recent years, not one has germinated at Kew. He recommends that the seeds should be gathered as soon as ripe, and be sent packed in fairly moist soil. Mr. Walker, the gardener at Barton, informed me that it ripens seed in good years, and showed me several seedlings raised from them which appeared to grow as well as the common horse-chestnut. The only other place except Kew, however, where we have seen it, is at Tortworth, where the Earl of Ducie planted in 1890 a few seeds which were sent to him by the late Duke of Bedford. The seedlings were planted at first in sunny places in the open, but did not thrive until moved to a sheltered dell in 1900, where they are now growing well, the best being about 12 feet high. At Kew there are two or three small trees which have flowered a few times. It seems, therefore, that it only requires a good deep soil and a sheltered situation to succeed as well as it has done at Barton. The late Lord Morley informed me that there was a tree recently planted, but growing very well at Saltram, his place in Devonshire. According to Jouin,4 this tree is quite hardy at Metz. (H. J. E.) 1 Arboretum Notes, 73 (1889). * Gard. Chron. 1903, xxxiii. 188. 2 Figured in Gard. Chron. 1904, xxxvi. 206, Suppl. Ilhtst. 4 Mitt. Deut. Dendrol. Gesell. 1905, p. 12. /Esculus 221 ^SCULUS TURBIN AT A, JAPANESE HORSE-CHFSTNUT jEsculus turbinata, Blume, Rumphia, iii. 195 (1847); André, Revue Horticole, 1888, p. 496, figs. 120-124; Bean, Gard. Chron. 1897, xxii. 156, and 1902, xxxi. 187, fig. 58; Shirasawa, Icon. Essences Forestières du Japon, text 113, t. 71, ff. 16-28. ^sculus chinensis, Masters (non Bunge), Gard. Chron. 1889, v. 716. fig. 116. A tree attaining in Japan, according to Shirasawa, ioo feet in height and 20 feet in girth of stem. Bark thick and scaly. Leaves resembling those of the common horse-chestnut, but much larger, mainly differing in the serration, which is finely crenate. Leaflets five to seven, sessile, obovate-cuneate, occasionally as much as 15 inches long, abruptly acuminate, pubescent beneath. The terminal leaflet has fifteen to twenty-two pairs of nerves. Petiole remaining pubescent towards the tip. Panicles 6 to 10 inches long, dense, somewhat narrow. Flowers yellowish-white, smaller than those of ^.sculus Hippocastanum. Fruit slightly pear-shaped, if to 2 inches in diameter, four to five on a verrucose rhachis, brown, warty, without spines ; valves three, thick ; seeds usually two. IDENTIFICATION In summer only liable to be confused with the European species, from which it is distinguished by the character of the serration of the leaflets. In winter the twigs closely resemble those of that species, but are not so stout; they are similarly pubescent towards the tip, and are marked with smaller but similar five to seven dotted leaf-scars. Buds smaller, equally viscid, the scales, however, not being uniform in colour, but partly light chestnut brown and partly dark brown. Pith large, irregularly circular in cross-section, and yellowish in tint. DISTRIBUTION The tree is known in Japan as Tockinoki, and is common in the forests at 1500 to 5500 feet elevation in the mountains of the main island, descending to lower levels in Yezo. It is recorded by Debeaux, Fl. Shanghai, 22, from the provinces of Kiangsu and Chekiang ; but no one else has seen the tree in China, and Debeaux's identification is probably incorrect. The exact date of the introduction of the tree into Europe is uncertain, but it is supposed to be about thirty years ago. It has often passed under the name of ^Esculus chinensis, an entirely different species. It first produced fruit in 1888 in the arboretum at Segrez in France. It flowered in 1901 at Coombe Wood. As only small trees are known to exist in England, the hardiness of the tree and its suitability for garden decoration are as yet unproved; but at Tortworth it is growing vigorously, and has ripened its buds well whilst still quite small; 222 The Trees of Great Britain and Ireland and the great size of the leaves on the young trees give it a striking and distinctive appearance. (A. H.) In Japan I saw this tree planted in gardens and parks near Tokyo, where it does not seem to grow so large as in its native forests and in higher, colder situations. Sargent says1 that in the forests of the interior of Hondo, at 2000 to 3000 feet, it attains 80 to 100 feet high, with trunks 3 to 4 feet in diameter, and that these were perhaps the largest deciduous trees that he saw growing wild in the forest. It reaches its most northern point of distribution near Mororan in Hokkaido at sea-level, and I did not see it near Sapporo, in the Aomori district, or near Nikko. At a tea-house called Hideshira, near the village of Sooga on the Nakasendo road, Central Japan, I saw the largest trees of this species growing in a dense grove with Zelkova acuminata. They attained over 80 feet high, with clean trunks 40 to 50 feet long, and a girth of 14 feet. On the Torii-toge Pass, between Wada and Yabuhara, at about 3300 feet, there were many fine trees growing by the side of the road, of one of which I give an illustration from a photograph taken for me by Masuhara of Tokyo in November (Plate 66). TIMBER The timber of this tree, though not highly valued in Japan on account of its softness and want of strength, is used for boat and bridge building,- furniture making, house-fittings, and for the groundwork of lacquer. It often shows a waved figure, and when old assumes a pale reddish - brown colour, which makes it very orna mental. Such wood, which I procured at Aomori, has been used with good effect in my Japanese wardrobe, and takes a good polish. It is also much used for trays, and from the large burrs and swellings near the root very handsome trays, as much as 18 or 20 inches square, are carved by the Japanese and sold in the villages at a low price. Its value in Tokyo is given at 60 to 100 yen per 100 cubic feet. I saw a plank of this wood in a timber merchant's shop in Osaka measuring 15 feet long and 58 inches wide, showing wavy figure all through. For this plank 90 yen, equal to about £<), was asked, these immense planks being much valued by Japanese connoisseurs for house decoration. (H. J. E.) 1 Forest Flora of Japan, 28. /Esculus 223 GLABRA, OHIO BUCKEYE ^Esfulus glabra, Willdenow, Enum. PI. Hort. Berol. 405 (1809) ; Loudon, Arl. et Frut. Brit. i. 467 (1838), Sargent, Suva N. America, ii. 55, tt. 67, 68 (1892), Man. Trees N. America, 644 ALsadus pallida, Willdenow, loc. at. 406 (1809). A tree attaining 70 feet in height and 6 feet in girth in America. Bark dark brown and scaly, becoming in old trees f inch thick, ashy-grey, densely furrowed and broken into thick plates roughened on the surface by numerous small scales. Leaves with long slender stalks; leaflets five, oval or obovate-cuneate, long -acuminate, finely serrate in margin, with tufts of hairs in the bases of the serrations, glabrous underneath except for a few hairs along the midrib and tufts in the axils ; petiolules short. Terminal leaflet with about fifteen pairs of nerves. Flowers in pubescent panicles, 5 to 6 inches long ; calyx campanulate ; petals four, pale yellow ; claws shorter than the calyx ; limbs twice as long as the claws, broadly ovate or oblong in the lateral pair, oblong -spathulate, much narrower and sometimes red-striped in the upper pair. Stamens usually seven, long, exserted, pubescent. Ovary pubescent. Fruit ovate or obovate, brown, i to 2 inches long, roughened by prickles. The species is distinguished in summer by the glabrous leaves, which always show some cilia in the bases of the serrations. In winter the following characters of the twigs and buds may be recognised : — Twigs glabrous, shining, with orange- coloured lenticels. Leaf-scars slightly oblique on obscure leaf-cushions, crescentic or semicircular, with three groups of bundle-dots, the opposite scars wide apart and often not joined by any linear ridge. Pith large, circular, greenish. Buds not viscid ; terminal much larger than the lateral, the latter arising from the twig at an angle of 45°; ovoid, acuminate ; scales keeled on the back, ciliate in margin, acuminate, the pointed tips being raised outwardly, dark brown. Var. Buckleyi, Sargent (^Esculus arguta, Buckley, Proc. Acad. Phil. 1860, p. 448), is a geographical form, occurring in Iowa, Missouri, Kansas, and Texas, and characterised by six to seven leaflets, which are sharply and unequally serrate. No well-marked horticultural varieties are known. The type occurs in alluvial soil in Atlantic North America, from Pennsylvania to N. Alabama, and westward to S. Iowa, Central Kansas, Indian Territory, and S. Nebraska. Sargent says that it is nowhere very common and from an ornamental point of view very inferior to jEsculus octandra. This species was introduced, according to Loudon, in 1812, but appears to be very rare in this country. At Devonshurst, Chiswick, a tree cut down in 1905 was 60 feet in height by 6 feet in girth, but though the tree probably exists in some nurseries and old gardens, where it is mistaken for ^Esculus octandra, more commonly than is supposed, we cannot mention any which are remarkable. (A. H.) 224 The Trees of Great Britain and Ireland ^ESCULUS OCTANDRA, SWEET BUCKEYE sEsculus octandra, Marshall, Arbust. Am. 4 (1785); Sargent, Siva N. America, ii. 59, tt. 69, 70 (1892), and Man. Trees N. America, 646 (1905). jEsculus lutea, Wangenheim, Schrift. Gesell. Nat. Fr. Berlin, viii. 133, t. 6 (1788). ^Esculusßava, Aiton, Hort. Keui, i. 403 (1789). sEsculus neglecta, Lindley, Bot. Reg. xii. t. 1009 (1826). Paviaßava, Moench, Method. 66 (1794); Loudon, Art. et Frut. ßrit. i. 471 (1838). A tree attaining in America 90 feet in height and 9 feet in girth of stem. Bark of trunk f inch thick, dark brown, slightly fissured, separating on the surface into thin small scales. Leaves with long slender petioles. Leaflets five, occasionally seven, elliptical or obovate-oblong, cuneate at the base, acuminate, finely serrate, pubes cent beneath ; petiolules short. Terminal leaflet with twenty or more pairs of nerves. Flowers in pubescent panicles, 4 to 6 inches long ; calyx campanulate ; petals four, yellow, coming into contact at the tips, very unequal, the upper pair much longer than the lateral pair, claws villose within and much exceeding the calyx, limb of lateral pair obovate or round with a subcordate base, limb of upper pair spathulate, minute. Stamens usually seven, shorter than the petals, villose. Ovary pubescent. Fruit 2 to 3 inches long, brown, smooth or slightly pitted. IDENTIFICATION In summer distinguished from ^Ssculus glabra by the leaflets being pubescent beneath and devoid of cilia in the serrations ; from ^Ssculus Pavia, by the larger leaves, which have petioles with smooth ridges on their upper surface. In winter the twigs show the following characters:—Branchlets glabrous, shining, with a few scattered lenticels. Leaf-scars flat on the twigs (there being no cushion), obovate, with usually three groups of bundle-dots ; opposite scars joined by a linear ridge. Pith large, circular, green or white. Buds not viscid, terminal much larger than the lateral, the latter arising at an angle of 45°, long-oval, pointed at the apex ; scales brown, the cilia on the exposed margins minute or absent, upper scales rounded at the apex and on the back, lower pair pointed at the apex and keeled on the back. VARIETIES 1. Var. hy brida, Sargent (Var. purpurascens, A. Gray; ^Esculus discolor,* Pursh). This is a form occurring wild in the Alleghany mountains. The flowers are purple or red in colour, and the under surfaces of the leaves, as well as the petioles and panicles, are clothed with a dense pale pubescence. 2. ^Esculus versicolor, Dippel. This is a hybrid between sEsculus octandra and ^Esculus Pavia, and is intermediate in character, the flowers varying in 1 Figured in Bot. Reg. iv. 310 (iSiS). ^sculus 2,2,5 colour from yellowish to pink. The edges of the petals show a few glands and are tufted ciliate. A considerable number of forms of this variety are known in cultivation in which slight differences occur in the length and shape of the petals. ^ZLsculus Lyoni and jfcsculus Whitleyi are apparently sub-varieties of this hybrid. The forms with red flowers are often known in gardens as Pavia rubra, a name which belongs properly to sEsculus Pavia. DISTRIBUTION This tree occurs in alluvial soil of river valleys and on moist mountain slopes, from Pennsylvania southward to Georgia and N. Alabama ; and westward to S. Iowa, Indian Territory, and W. Texas. Sargent says that when at its best on the slopes of the Tennessee and Carolina mountains, it sends up a straight shaft sometimes free of branches for 60 to 70 feet, and reaches a total height of 90 feet. (A. H.) CULTIVATION According to Loudon this species was introduced into England in 1764, but though more common in cultivation than any sEsculus except A. Hippocastanum, and apparently not particular about soil, it does not attain any great size. It is perfectly hardy at Colesborne, and ripens fruit in most years, from which I have raised seedlings, which, however, do not grow so fast or well as those of the common horse-chestnut. A seedling raised from a tree at Tortworth in 1905 was 6 inches high in the first year, and some raised from seed which I gathered in the Arnold arboretum, which germinated earlier, were much injured by the frost of May 21-22. At Syon there are two trees, probably of a great age, both grafted on the common horse-chestnut. One is 65 feet high by 4 feet 4 inches in girth ; the other is 56 feet high by 6 feet 4 inches in girth, with a bole of 7 feet, dividing into three stems, which form a wide-spreading crown. A tree at Belton Park, Lincolnshire, was, in 1904, 50 feet high by 3 feet 4 inches in girth, with a fine straight stem, drawn up in a wood. Another, crowded by other trees near the Broad Water at Fairford Park, Gloucestershire, measures about 60 feet by 4 feet 5 inches. A self- sown seedling was growing near it in 1903. There is also a tree, measuring about 50 feet by 5 feet 6 inches, at Charlton Kings, near Cheltenham. (H. }. E.) II 2,26 The Trees of Great Britain and Ireland yESCULUS CALIFORNICA, CALIFORNIAN BUCKEYE californica, Nuttall, in Torrey and Gray, Fl. N. America, i. 251 (1839); Bot. Mag. t. 5077 (1858); Sargent, Suva N. America, ii. 61, tt. 71, 72, and Man. Trees. N. America, 648 (1905); Bean, in Gard. Chron. 1902, xxxi. 187, fig. 57. A tree, attaining in America 40 feet in height, with a short trunk occasionally 9 feet in girth. Bark smooth, grey or white. Leaves with slender grooved petioles. Leaflets five to seven, stalked, oblong lanceolate, acuminate at the apex, cuneate or obtuse at the base, shallowly and crenately serrate, pale glabrescent beneath. Terminal leaflet, with ten to twelve pairs of nerves. Flowers in dense pubescent panicles, 3 to 8 inches long. Calyx two-lipped, upper lip with three teeth, lower lip with two teeth much shorter than the four narrow oblong petals, which are white or pale rose in colour. Stamens five to seven, long, erect, exserted. Ovary pubescent. Fruit pear-shaped, two to three inches long, smooth, pale brown. In summer it is readily distinguished from the other species with viscid buds by the small leaves, pale beneath. In winter the twigs are slender, grey, glabrous, with numerous lenticels. Leaf-scars wide apart, joined by a linear ridge, flat on the twig, without a leaf-cushion, crescentic or semicircular, with a row of five to seven bundle-dots. Pith large, circular, white. Terminal buds, larger than the lateral buds, which arise at an acute angle, oval, pointed, glistening with white resin ; scales gaping at the apex of the bud, broadly ridged on the back, ciliate in margin, with a tuft of hairs at the apex. The species is a native of California, where it grows on the banks of streams. A very striking picture of a tree, at San Mateo, California, is given in Garden and Forest, iv. 523. It shows a very short forked bole, nearly 20 feet in girth at 2 feet from the ground, and an immense umbrella-shaped head only 32 feet high and 60 feet in diameter, densely covered all over with flowers. It was introduced in 1855 by Messrs. Veitch, and flowered in their nursery at Exeter in 1858. It fruited1 at the Bath Botanic Gardens in 1901, and again in 1905, though it remains a shrub. It is perfectly hardy in the south of England, and is remarkable for the beauty of its flowers, which appear in June and July. The best specimen we know of in the country is one which Elwes found growing in a shrubbery at Hutley Towers near Ryde, Isle of Wight. It is about 30 feet high, and was in flower on June 22, 1906. (A. H.) 1 Gard. Chron. 1902, xxxi. 187. T S U G A Tsuga, Carrière, Traité Conif. 185 (1855); Bentham et Hooker, Gen. PI. iii. 440 (1880); Masters, Journ. Linn. Soc. (Bot.) xxx. 28 (1893). Hesperopeuce, Lemmon, Rep. Calif. State Board Forestry, iii. ni (1890). EVERGREEN trees belonging to the natural order Coniferse. Branches horizontal or pendulous, pinnately and irregularly ramified. Buds, one terminal and a few lateral, arising irregularly in the axils of some of the leaves of the current year's shoot, most of the leaves being without buds in their axils. Leaves linear, arising from the branchlets in spiral order, and usually thrown by a twisting of their petioles into a pectinate arrangement, or in one species spreading radially. Petioles short, arising from prominent leaf-bases on the branchlets, appressed against the twigs, a sharp angle being formed by the leaf with the stalk at the point of junction. The leaf has one resin-canal, lying in the middle line between the vascular bundle and the epidermis of the lower surface. The leaves persist for several years ; and all the species have in consequence of this and their numerous and fine branchlets very dense foliage. Flowers monoecious. Male flowers in the axils of the leaves of the previous year's shoot near its apex, composed of numerous spirally arranged, short-stalked, two-celled anthers, with glandular-tipped connectives. Female flowers terminal on lateral shoots of the previous year, short-stalked or sub-sessile, erect, composed of spirally arranged, nearly circular scales, and membranous, usually shorter bracts. Ovules, two on each scale. Cones solitary, small, composed of concave woody imbricated scales, which persist on the axis of the cone tafter the escape of the seeds, and of inconspicuous bracts, which, except in one species, are concealed between the scales. The cones, ripening in one season, allow the seeds to fall out in the first autumn or winter, but remain on the tree until the summer or autumn of the second year. The seeds, two on each scale, are minute, furnished with resin vesicles and winged. The seedling has three to six cotyledons, which bear stomata on their upper surface. Tsuga is confined to temperate North America, Japan, China, and the Himalayas. The genus consists of nine species, and is divided into two sections :— I. Hesperopeuce, Engelmann, in Brewer and Watson, Bot. California, ii. 121 (1880). Leaves rounded or keeled above, bearing stomata on both surfaces, and radially arranged ; the shorter and lateral branchlets standing in a plane at right angles 227 228 The Trees of Great Britain and Ireland to the longer and terminal ones. Cones oblong-cylindrical, large, composed of numerous (about seventy) scales. This section includes one species :— i. Tsuga Pattoniana, Sénéclauze. Western North America. II. Micropeuce, Spach, Hist. Vég. xi. 424 (1842), identical with Eutsuga, Engel- mann, loc. cit. 120 (1880). Leaves flat, grooved above, bearing stomata on the lower surface only, pectinately arranged on the branchlets, which are all in one plane. Cones ovoid, small, com posed of few scales, rarely more than twenty-five. This section comprises the remaining species, of which six are in cultivation in this country. These may be conveniently arranged as follows :— A. Leaves serrulate in margin. Shoots pubescent. 2. Tsuga Canadensis, Carrière. Eastern North America. Leaves, -J to § inch long, usually tapering from the base to the acute or rounded apex ; lower surface marked with two narrow well-defined white stomatic bands, the part of the leaf external to them being pure green in colour. Buds brown, ovoid, pointed, composed of pubescent, keeled acute scales. 3. Tsuga Albertiana, Sénéclauze. Western North America. Leaves, \ to f inch long, usually rounded at the apex and uniform in breadth ; lower surface with two ill-defined broad white stomatic bands, which are indistinctly continued to the margins, there being no distinct bands of pure green. Buds greyish, ovoid, apex obtuse and flattened ; scales keeled, pubescent. 4. Tsuga Brunoniana, Carrière. Himalayas. Leaves, i to i^. inch long, gradually tapering from the base to the acute apex ; lower surface silvery white, stomatic bands well-defined and extending almost to the margins. Buds globose, flattened on the top, surrounded at the base by a ring of modified leafy scales, the other scales ovate, acute, pubescent. B. Leaves entire in margin. Shoots glabrous. 5. Tsuga Sieboldii, Carrière. Japan. Leaves, \ to i inch long, oblong, rounded and notched at the apex, shining above ; lower surface with two narrow well-defined white stomatic bands. Buds red, ovoid, slightly acute at the apex ; scales glabrous and ciliate. C. Leaves entire in margin. Shoots pubescent. 6. Tsuga diversifolia, Maximowicz. Japan. Shoots pubescent, both on the leaf-bases and in the furrows between them. Leaves, \ to \ inch long, oblong, rounded and notched at the apex ; lower surface with two narrow well-defined white stomatic bands. Buds red, pyriform, flattened above ; scales obtuse, minutely pubescent. 7. Tsuga Caroliniana, Engelmann. Southern Alleghany Mountains. Shoots pubescent in the furrows between the leaf-bases, which are glabrous. Tsuga 229 Leaves, i to f inch long, oblong, rounded at the apex, which is entire, minutely notched or mucronate ; lower surface with two narrow well-defined white stomatic bands. Buds reddish, ovoid, sharp-pointed ; scales indistinctly keeled. In addition to the preceding, two species of Tsuga, belonging to this section, occur in China. They are as yet imperfectly known. Tsuga chinensis, Masters,1 a native of the high mountains of Szechuan, is closely allied to Tsuga diversifolia, and, like it, has pubescent young shoots. It differs in the cones, which are quite sessile, and have very lustrous scales. The leaves are described as being green beneath ; but this is probably an inconstant character. Tsuga yunnanensis, Masters,2 which was discovered by Père Delavay in the mountains near Likiang in Yunnan, is unknown to me. Franchet considers it to be closely allied to T. Sieboldii. TSUGA PATTONI AN A, HOOKER'S HEMLOCK Tsuga Pattoniana, Sénéclauze, Conif. 21 (1867); Engelmann, in Brewer and Watson, Bot. California, ii. 121 (1880); Masters, Gard. Chron. xii. 10, fig. i (1892). Tsuga Hookeriana, Carrière, Traité Conif. 252 (1867); and Lemmon, Erythea, vi. 78 (1898). Tsuga Mertensiatia, Sargent, Silva N. Amer. xii. 77, t. 606 (1898), and Trees N. Amer. 51 (1905); Kent, Veitch's Man. Coniferce, 468 (1900). Pinus Mertensiatia, Bongard, Végét. de Sitcha, 54 (1832). Pinus Pattoniatia, Pariatore, D. C. Prod. xvi. 2, p. 429 (1864). Abies Pattoniana, Balfour, Rep. Oregon Assoc. i (1853); Murray in Lawson, Pin. Brit. ii. 157 (1884). Abies Hookeriana, Murray, Edin. New Phil. Jaum. 289 (1855); and in Lawson, loc. cit. 153. Abies Williamsonii, Newberry, Pacific R'. R. Report, vi. pt. iii. 53, t. 7, fig. 19 (1857). Hesperopeuce Pattoniana, Lemmon, Rep. Calif. State Board Forestry, iii. 128 (1890). A tree, occasionally attaining in America 150 feet in height, with a girth of 15 feet. Bark dark cinnamon in colour, deeply divided into rounded connected scaly ridges. Shoots brownish-grey, and densely pubescent. Branchlets in different planes, the shorter and lateral ones usually arising on the upper side of the longer and terminal ones, and disposed at right angles to them, giving a tufted appearance to the branch. Leaves radially arranged on the branchlets, not markedly different in size, f to i inch long, curved, linear ; apex usually rounded and obtuse, rarely acute ; upper surface convex and keeled towards the apex ; lower surface rounded with a median groove ; both surfaces with about eight lines of stomata, which are sparse and do not form conspicuous white bands ; margin entire. Buds brownish, ovoid, acute at the apex, composed of a few closely imbricated, strongly keeled scales. Cones sessile, about two inches long, oblong cylindrical, tapering at the apex and slightly narrowed at the base, composed of five series of scales, each series with 1 Journ. Linn. Soc. (Bot.) xxvi. 556 (1902) ; Abies chinensis, Franchet, Journ. de Bot. 1899, p. 259.. 2 Journ. Linn. Suc. (Bot.) loc. cit.; Abies yunnanensis, Franchet, loc. cit. p. 258; and cf. also Masters,./cw?-». Linn. Soc. (Bot.) xxxvii. 421 (1906), who identifies the specimens from Szechuan with this species; but judging from Franchet's description, they are the other species. 23° The Trees of Great Britain and Ireland fourteen to fifteen scales. Scales thin, broader than long, semicircular with a wedge-shaped base, convex, margin irregularly denticulate, pubescent on both surfaces. Bract oblong, abruptly tapering at the apex, which is visible between the scales. Seed with terminal asymmetrical wing, and two resin-vesicles on the side next the scale. The name Pattoniana is adopted as being the first published under the correct genus Tsuga. The tree is known to American botanists as Tsuga Merten- siana, which is unfortunate, as this name was for many years in use for the western hemlock. There is no confusion possible if Pattoniana be selected, as no other hemlock has been known at any time by this name. VARIETIES The preceding description is drawn up from living specimens of the form with bluish entire leaves, cultivated in this country, and applies, in all essential charac ters, to dried specimens from trees growing wild in America. I have examined the material in the Kew herbarium and also specimens collected by Elwes on Mount Shasta at 7500 feet elevation ; and there do not appear to be two distinct varieties of the tree in the wild state, as the presumed alpine form is only a stunted shrub which agrees in botanical characters with the trees from lower levels. In England, however, there is a form in cultivation, distinguished by its green serrulate leaves, which differs in many respects from the other form. Concerning its origin, we only know, on the authority of Murray,1 that it was raised at Edinburgh from seeds collected by Jeffrey in 1851 on the Mount Baker range in British Columbia. Jeffrey found trees growing there from 5000 feet elevation to the snow line, varying in size from 150 feet in height and 13^ feet in girth at lower levels to a stunted shrub not more than 4 feet high close to the timber line. Specimens at Kew from Mount Baker gathered by Jeffrey all have entire leaves and belong to the ordinary wild form. Engelmann,2 who visited the Mount Baker range, states that the trees growing there are the ordinary forms of Tsiiga Pattoniana and Tsuga Albertiana. He suggests that the plants raised from Jeffrey's seed may be a mountain form of the latter species ; but this cannot be admitted, as they do not resemble that species in botanical characters (buds, leaves, etc.). It is possible that these plants are only a seedling variation of Tsuga Pattoniana, and do not correspond with any distinct species or geographical form in the wild state. Murray,3 believing that he had two species to deal with, named the bluish form Abies Hookeriana, and assigned the name Abies Pattoniana, Balfour, to the other form. The original figure of Balfour's species represents, however, the same plant as Abies Hookeriana of Murray; and much confusion has resulted in consequence in the use of the two names Hookeriana and Pattoniana. It is most convenient to 1 Edin. New. Phil. Jour. 289 (1855) and Proc. Hort. Soc. ii. 202 (1863). 2 Card. Chron. xvii. 145 (1882). 8 The distinctions relied on by Murray in the cones are trifling ; and in the Kew Herbarium there are wild specimens showing these differences, but all belonging to the form with blue entire leaves. I have not seen cones belonging to the other form. Tsuga 231 apply the name Pattoniana to the bluish form, as it is the earliest name of the wild plant, and to consider the green-foliaged plant to be a variety of it, which may be called var. Jeffreyi. The two forms are distinguished as follows :— 1. Var. typica. The form distinguished in cultivation by its bluish foliage. Introduced in 1854 by William Murray, who found the tree on Scots Mountain, in California. Leaves, though radially arranged, tending on the lower side of the shoot to be in the plane of the branch and not spreading ; those on the upper side of the shoot curved and directed outwards and forwards. They are long and narrow, \ to \ inch long, and -£$ inch wide, entire in margin, convex on both surfaces, the groove in the median line above being very short or absent and never continued to the apex of the leaf, which is rounded or acute ; both surfaces marked with conspicuous lines of stomata extending from the base to the apex of the leaf. 2. Var. Jeffreyi. Only known in cultivation, distinguished by its greenish foliage. Leaves spreading radially and directed outwards (never forwards) on all sides of the shoot ; straight, short, and broad, less than ^ inch long and about ^ inch in width, serrulate in margin ; upper surface flattened and distinctly grooved, the groove continued to the rounded apex ; lower surface convex, with lines of stomata the whole length of the leaf. On the upper surface the stomata only occur in four to six broken lines towards the apex. This form agrees with the typical form in the character of the buds and pubescence of the branchlets ; the shoots, however, are not so slender. (A. H.) Mr. Gorman gives the following account1 of the supposed Alpine form, alluded to above :—" Among the hardy alpine trees Hooker's hemlock stands pre-eminent, having a northern range far beyond that of even the white-barked pine. It is a small, dwarfed and stunted tree compared with the type, and seldom exceeds 12 inches diameter or 30 feet in height. It usually ranges in altitude from 5500 to 6400 feet, but is occasionally found up to and beyond 7000 feet where it can find sufficient moisture. Though generally favouring the heads of moist valleys it is sometimes to be found on the leeward side of peaks and slopes, where snowbanks of sufficient size have formed in winter to maintain an adequate supply of moisture during the rest of the year. It is in the latter situations where the tree reaches its highest altitude. In addition to its smaller size and more alpine habit it further differs from its nearest congener in having thinner bark and small erect cones, all the other hemlocks having pendent cones. The tree is too small and inaccessible to have any economic value." This seems to be distinguished principally by its erect cones. Sargent,2 who alludes to Gorman's account, does not consider this variation to be worthy of distinc- i Survey E. Part Washington Forest Reserve, p. 336 (igth Ann. Report of the .Survey, Part v. 1899). * iSilva ff. Amer. xii. 78, note I. 232, The Trees of Great Britain and Ireland tion, and explains it by saying that the position of the cones "is evidently due to the thickness of the short lateral branchlets, on which they are terminal and which are sometimes so rigid that the weight of the cones does not make them pendent." 'DISTRIBUTION This tree is only found at high elevations, where it has much the same geographical range as the western hemlock, but it extends farther south in California and reaches its southern limit at 9000 to 10,000 feet on the south fork of King River in the Sierra Nevada. In the north it descends to sea level on Baranoff Island, and on the shores of Yes Bay in Alaska, lat. 55° 54' N., where Mr. Martin Gorman collected it. As a rule it is a tree of high altitudes, growing on exposed ridges and slopes near the upper limit of the forest, in company with Abies lasiocarpa, Picea Engelmanni, and Pinus albicaulis. In the Rocky Mountains of British Columbia Mrs. Nicholl found it as a good-sized tree near Glacier up to 7000 feet, though Wilcox,1 in his excellent account of the trees ofthat region, pp. 61-65, does not mention it. Though usually a more or less stunted and ragged tree, it attains a large size on the Cascade Mountains, where I saw it in perfection on the road from Longmire Springs to Paradise Valley, on the south side of Mount Tacoma,2 in August 1904, first at about 4000 feet, where it was only a scattered tree, and higher up it mixed with the western hemlock in a splendid forest. I was not able to distinguish the two species by their bark, though when not crowded, the habit of Hooker's hemlock is very distinct ; but they could be identified by the fallen cones under the trees. The largest that I measured here was about 150 feet by 13 feet 8 inches. Higher up, where the forest3 opened out into glades at the bottom of the Paradise Valley, which is, in Professor Sargent's opinion, one of the most interesting in America for its alpine flora, it assumed a different and more flat-topped habit ; the largest here that I measured was 108 feet by 13 feet 3 inches. It grew in company with Abies lasiocarpa, and seedlings of both were numerous on rotten logs on the shady sides of the clumps in which they always grew. The tree in a very stunted state reaches the timber line—about 7500 feet—in company with Abies lasiocarpa and Cupressus nootkatensis ; but in California, J. Muir4 measured a specimen at 9500 feet, near the margin of Lake Hollow, which was 19 feet 7 inches in girth at 4 feet from the ground. Mr. Gorman gives an excellent account of the tree in his Survey of the Eastern Part of the Washington Forest Reserve, pp. 335-336, from which I quote as follows :— " This hemlock is confined to the moist valleys and vicinity of the passes. It is the prevailing tree in Cascade Pass, 5421 feet, and is quite common about the 1 I'he Rockies of Canada, 61-65 (1900). 2 The local name is Mount Tacoma, but in maps and wiitings it is usually called Mount Rainier. 3 An account of this forest, with two beautiful illustrations of " Patton's spruce," is given in Garden and Forest, x. I, figs. I, 2 (1897). 4 Mountains of California, p. 20. Tsuga 2-33 sources of the Stehekin, where it attains a very fair size for this region, ranging from 50 to 90 feet in height and from 12 to 27 inches in diameter. The altitudinal range is greater than was expected, from 3100 feet to 5800 feet, and a tree supposed to be of this species was found as low as 2109 feet in the Stehekin Valley. " The tree is sometimes taken for the western hemlock, but may be distinguished by the erect top of the sapling, the cones long, purple, and more or less massed about the top of the tree ; and the mature tree has an unusually thick, roughly corrugated bark : while in the western hemlock the top is generally drooping, the cones small, oval, and brown, and well distributed over the branches, and the mature tree has a comparatively thin bark. The wood is close grained and of fine texture, and is quite suitable for lumber or fuel, but is not much used on account of its growing usually in inaccessible situations." Near Crater Lake in Southern Oregon, Mr. Leiberg {Cascade Forest Reserve Report, pp. 245, 259), says :—" A few scattered groves of Patton hemlock occur in the southern tracts, some of which are of large size, occasional individuals reaching six to seven feet in diameter. Occasional stands of Patton hemlock 200 to 300 years old exhibit fine proportions at this elevation, 6000 feet ; the species usually grows in close groups, composed of ten or twenty individuals, collected together on what appears to be a common root ; such close growth develops clear trunks, though not commonly of large diameter. Stands of this character sometimes run as high as 25,000 feet per acre." REMARKABLE TREES Though now introduced for about fifty-five years this tree has made but little show in our gardens, as the climate of most parts of England is probably too warm for it. I have seen flourishing specimens of no great size in several places, the best, perhaps, being one at Tyberton Court, Herefordshire, the seat of Chandos Lee Warner, Esq., where there is a tree of the typical form 43 feet high by about 35- feet in girth, said to be fifty years old, and perhaps one of those introduced by William Murray, and sent out by Lawson. In Scotland it seems to thrive even better, especially at Murthly Castle, where there is a fine group of trees on a lawn (Plate 67). When measured for the Conifer Conference in 1892 the best of these was 35 feet by 3 feet 10 inches, another 30 feet by 4 feet. When I last saw them in September 1906 the tallest tree on the left of the row was 47 feet by 3 feet 8 inches, the tree in the middle with weeping branches 43 feet by 4 feet 2 inches, and the thickest between these two was 6 feet 7 inches in girth. The difference in the habit'of these three is well shown in the plate. They produced seed in 1887, from which a number were raised and planted at Murthly. These have grown slowly, .ind the tallest in 1906 were six or seven feet high, though quite healthy ; and the growth of seedlings which 1 raised from seed gathered on Mount Rainier is extremely slow. At Keillour, Henry measured, in 1904, a specimen which was 40 feet by 3 feet 9 inches ; and at the Cairnies, near Perth, the seat of Major R. M. Patton, there were in 1892 two specimens little inferior to those at Murthly. (H. J. E.) n F 234 The Trees of Great Britain and Ireland TSUGA ALBERTIANA, WESTERN HEMLOCK Tsuga Albertiana, Sénéclauze, Conif. 18 (1867); Kent, Veitch's Man. Coniferce, 459 (1900). Tsuga Mertensiana, Carrière, Traité Conif. 250 (1867); Masters, Gard. Chron. xxiii. 179, fig. 35 (1885). Tsuga heterophytta, Sargent, Situa JV. Amer. xii. 73, t. 605 (1898), and Trees N. Amer. 50(1905). Abies heterophylla, Rafinesque, Atlantic Jour. i. 119 (1832). Abies Mertensiana, Gordon, Pinetum, 18 (1858). Abies Albertiana, A. Murray, Proc. Roy. Hort. Soc. iii. 149 (1863). A large tree, attaining in America 200 to 250 feet in height and 20 feet or more in girth, narrowly pyramidal in habit. Bark of old trees reddish brown, and deeply divided into broad, flat, connected scaly ridges. Young shoots whitish grey, and covered with short pubescence, intermixed with scattered long straggling hairs. Leaves pectinately arranged, the shorter leaves on the upper side of the branchlets, those in the median line above often parallel to the twig and directed forwards, exposing their stomatic surfaces. The leaves are \ to f inch long, linear-oblong, uniform in width, serrulate in margin, dark green above, with a median groove continued up to the rounded apex ; under surface with inconspicuous midrib and two broad white stomatic bands, which are ill defined on the outer side, there being no distinct marginal green bands. Buds greyish brown, ovoid, with an obtuse and flattened apex ; scales keeled and pubescent. Cones sessile, about one inch long, ovoid, composed of five series of scales, each series with six to seven scales. Scales spathulate, nearly twice as long as broad, wider in the upper half, abruptly narrowed below, rounded with a slightly acute apex, entire and slightly bevelled in margin, striate and slightly pubescent on the outer surface. Bract small, concealed, lozenge-shaped, pubescent and keeled. Seed with a very long wing, decurrent on the outer side of the seed to the base ; seed with wing about three-fourths the length of the scale. The young seedling has three to four cotyledons, which are a little more than \ inch in length, gradually tapering to an acute apex, sessile, flattened beneath, the upper surface two-sided and bearing stomata, margin entire. The young stem is pubescent and bears first two to three whorls of true leaves (three in each whorl), which are serrulate, shortly stalked, and bearing stomata on their upper surface. These are succeeded by leaves borne spirally. The cotyledons are supported by a caulicle, reddish and glabrous, about an inch in length, which terminates in a very slender flexuose tap-root. The name Albertiana has been chosen, as it appears to have been published as early as that of Mertensiana under the correct genus Tsuga. Tsuga Mertensiana is now the name given by American botanists to Tsuga Pattoniana, and its adoption would cause considerable confusion. Albertiana, never having been applied to any other species, is correct on the grounds of common sense as well as of priority. (A. H.) Tsuga 2-35 DISTRIBUTION On the west coast of North America it extends southwards from south-eastern Alaska, where it forms the greater part of the great coast forest, which reaches from sea-level up to about 2000 feet, and is associated with Menzies's spruce. In British Columbia it is very abundant on the coast, and extends as far inland as the heavy rainfall reaches up the valley of the Frazer, on the Gold and Selkirk ranges, and east of the Columbia valley nearly up to the continental divide.1 In Vancouver's Island it forms with the Douglas fir and red cedar a large though not economically important part of the forest. In Washington and Oregon it is also one of the principal elements of the forest, of which, in the Cascade Forest Reserve, it forms about nine per cent of the timber,2 and extends up to 5000 feet, crossing the watershed of the coast range in lat. 45°. In the drier parts of southern Oregon it becomes rare, and though it occurs in the redwood forests of northern California as far south as Cape Mendocino, I did not see it on the Siskyou mountains or on Mount Shasta. In the interior it is found in the wetter parts of northern Montana, Idaho, and in southern British Columbia, where, in company with Douglas spruce, Picea Engelmanni, Abies grandis, and Larix occidentals, it sometimes forms a considerable part of the forest, and reaches up to 6000 feet in the Cœur d'Alêne mountains, though I did not see it in the valley of the Blackfoot river, near Missoula, where the climate is drier. It attains its finest development on the coasts of Washington and Oregon, where Sargent says that it attains 200 feet in height, with a stem 20 to 30 feet in girth. Plummer, in his Report on the Mount Rainier Forest Reserve,8 says (p. 101) that it attains an extreme diameter of 6 feet, with a height of 250 feet, of which half to two-thirds is crown. The largest that I actually measured, however, on my visit to Mount Rainier in August 1904, were under 200 feet, with a girth of 12 to 14 feet, and these were growing mixed with Tsuga Pattoniana at an elevation of 4000 to 5000 feet. In the Cascade Reserve Forest of northern Oregon, near Bridal Veil, at about 3500 feet elevation, I measured and Mr. Kiser photographed a tree 175 feet high and 16 feet 6 inches in girth, with a clean bole of about 60 feet, but I am unable to reproduce this, as the negative has not arrived. The growth of seedlings in all the forests that I saw was exceptionally good. Mr. H. D. Langille says,4 p. 36 :— "Certain cone-bearers are better adapted for restocking than others, though the reasons are not apparent. For example, young lovely firs (A. amabilis] are abundant everywhere within the zone of that species, whilst noble fir (A. nobilis), having a cone and seed of very similar size and nature, seldom germinates, and a seedling of that species is rarely seen. 1 Mrs. Nicholl, who explored the Rocky Mountains in 1904 and 1905, tells me that it is a large tree at Glacier, on the Canadian Pacific Railway, and grows up to about 5000 feet. 2 forest Conditions of Cascade Reserve, p. 25, Washington, 1903. 3 Twenty-first Annual Report of the U.S. Geological Survey, part v. Washington, 1900. 4 forest Conditions in Cascade Reserve, U.S. Geological Survey, Washington, 1903. 236 The Trees of Great Britain and Ireland " From many observations made in the zone of the hemlock and lovely fir, it is apparent that these trees, from their ability to thrive under the most adverse con ditions, are rapidly superseding the others, and will, under natural conditions, be the sole components of the alpine forests. It is a striking fact that, upon many areas where from 50 to 100 per cent of the present forest is red fir (Douglas), the repro duction is entirely hemlock and lovely fir. Should these forests be destroyed by fire it is probable that red fir would rival these species in restocking the burn ; but under natural conditions it is evident that the red fir will be displaced, and the limits of the alpine trees become much lower than at present. "The yellow pine (P. ponderosa), in some instances, does good work in stocking open spots in the timber, but seldom extends far beyond the parent tree. In the yellow pine forests most of the young growth is red or white fir (A. grandis], which, taking advantage of the shade and moisture afforded by the yellow pine cover, is growing rapidly, and will in time form a larger percentage of the forest than it has in the past." I can confirm this from my own observation both in the Cascade Forest and in Vancouver's Island. The seedlings germinate most freely when they fall on the moss-covered rotting trunk of a fallen tree, along which a complete row of young trees often grows ; and Plate 59, vol. i. shows a tree of this species, probably 150 years old, whose roots had completely enclosed the still sound trunk of a red cedar ( Thuya plicata). A valuable paper1 by Mr. E. T. Allen, dealing with the western hemlock from a forestry point of view, has been published by the U.S. Bureau of Forestry. CULTIVATION It was introduced in 1851 by Jeffrey, and named in 1863 by Murray, at the request of Queen Victoria, in memory of the late Prince Consort, who was a patron of the Oregon Association, and President of the Royal Horticultural Society.2 In grace, freedom of growth, and adaptability to varied conditions of culture, in England this, as an ornamental tree, is second to none, and much superior to any other hemlock. Though it has been in cultivation little over fifty years it has already attained a height of about 90 feet in such widely distant counties as Kent, Devon shire, and Perthshire. The only soils on which it will not thrive are chalk, limestone, and heavy clay, and though it enjoys all the moisture that the wettest parts of England afford, it wants, like all its congeners, a well-drained soil and a sheltered situation. It ripens seed abundantly in England, and has sown itself in several localities, especially at Blackmoor, the seat of the Earl of Selborne, where there are several self-sown trees, of which the best, growing on the lower greensand formation, is, at about fifteen years old, 10 to 12 feet high, though the parent trees do not exceed about 65 feet. In Fulmodestone Wood, on Lord Leicester's estate in Norfolk, I have also seen self-sown seedlings ; and though they are very slow in growth for the first four or five 1 "The Western Hemlock," U.S. Dept. Agric. Forestry Bulletin, No. 33 (1902». " Hunter, Woods of Perthshire, p. 359. Tsuga 237 years, yet if kept moist and shaded in a mixture of sand and leaf-mould they may be planted out at five to six years old, with every hope of success. So far as my experience goes, trees grown from cuttings are not so satisfactory, and there is no excuse for this practice except the saving of trouble, as seedlings are raised in quantity at a very low cost from home-grown seed in Scotland, as I have seen in the nursery at Murthly Castle. REMARKABLE TREES Among so many fine trees of this species, all of about the same age, it is hard to choose, but perhaps the largest1 which we have measured is at Hafodunos, in Denbighshire, which in 1904 was found by Henry to be 94 feet 6 inches by 8 feet 5 inches, and this tree has also produced self-sown seedlings. At Dropmore there is a very beautiful tree of the spreading type (Plate 68), about 70 feet by 6 feet. At Hemsted, in Kent, I was shown by Lord Cranbrook, in 1905, a tree which is perhaps as tall as any in England, but which, growing in a hole and surrounded by other trees, it was not possible to measure accurately. It is, however, about 90 feet by 4 feet 11 inches, well shaped and growing fast. At Penllergare, near Swansea, the seat of Sir J. T. D. Llewellyn, Bt., are several fine trees growing in a sheltered valley, which were planted about fifty years ago in company with Tsuga canadensis. They are now from 70 to 80 feet high, whilst the best of the eastern hemlock is only 50 feet, and the difference in habit of the two trees is very well shown. A very large tree, reported2 to be no feet high, is growing at Singleton Abbey, near Swansea, the residence of Lord Swansea, but I have been unable as yet to get confirmation of the height stated. At Castlehill, JM. Devon, are several fine trees, the best of which, on a steep bank above a waterfall, where it is somewhat drawn up by beeches, is 90 feet by 6 teet 7 inches. At Carclew, Cornwall, is a fine tree, which in 1902 was 80 feet by 6 feet 3 inches, and in 1905, 82 feet by 6 feet 6 inches, both measurements taken by myself. At Barton, Suffolk, a young and very thriving tree, shut in by tall beeches and conifers, in 1905 was 80 feet by 4 feet 3 inches, a remarkable instance of height as compared with girth. In Scotland the tree flourishes exceedingly, and has been planted in many places. Perhaps the tallest is one at Castle Menzies, which in 1904 I made about 90 feet by 7 feet 8 inches, though the gardener thinks it is taller ; but one of the most beautiful for its shape, graceful habit, and situation, grows by a deep shady burn on the road from Dunkeld to Murthly Castle, and is about 70 feet by 5 feet (Plate 69), and there are many other fine trees in the grounds there. A tree at Riccarton, near Edinburgh, planted in 1855, measured in 1905, 73 feet by 7 feet i inch. A very large tree, measuring in 1907, 10 feet in girth, is reported by Major P. J. Waldron, to be growing at Hallyburton, Coupar-Angus, the seat of Mr. W. Graham Menzies. 1 This tree was in 1868, 28^ feet high by 2 feet 3 inches in girth at the base. In 1883 it measured 65 feet by 4 feet II inches at 3 feet from the ground (Card. Chron. 1868, p. 657, and 1885, xxiii. 179). According to the owner, Colonel Sandbach, it was planted probably in 1856. 2 Car,/. Chron. xxxvii. 136(1905). 238 The Trees of Great Britain and Ireland The only place where the tree is reported to have been killed by frost is in the plantations at the Cairnies, Perthshire, where Hunter says (p. 364) that in the severe winter of 1880-81 many were injured and some killed. Two of the finest specimens in Scotland are, however, growing in the grounds at this place.1 In Ireland the best specimen we know of is one at Glenstal, Co. Limerick, which measured in 1903, 78 feet high by 7! feet in girth. One of exactly the same height by 6 feet in girth is growing at Kilmacurragh, Co. Wicklow ; and around it are several self-sown seedlings. At Mount Usher, in the same county, there is a fine specimen, 28 years old, from seed, which was 57 feet high by 4 feet 5 inches in 1903. TIMBER The timber of the western hemlock has not until recently been much valued, or cut for lumber, on account of its supposed inferiority to that of the Douglas spruce, and is often left standing by loggers, but the increasing scarcity of lumber in some districts has led to its being converted into boards, and it is now largely used for the construction of buildings. Sargent says that it is light, hard, and tough, stronger, more durable, and more easily worked than the other American hemlocks. Allen2 says that in strength it cannot be classed with oak, red fir, or longleaf pine, nor is it suitable for heavy construction, especially where exposed to the weather ; but it possesses all the strength requisite for ordinary building material. It is largely used in Washington for mill frames. At Mr. Bradley's sawmill at Bridal Veil, Oregon, I saw it being manufactured, and brought away a sample which quite bears out Sargent's high opinion of it. If such timber existed in Japan or in Europe, I am sure it would be highly valued for joinery, but so far as I can learn none has yet been shipped to Europe. Hemlock timber2 has been exported to Manila, and is likely to prove of considerable value in the tropics for housebuilding and indoor finish, as it appears to be free from the attacks of white ants. The wood is distasteful to rodents, and is used on that account by farmers for the construction of oat-bins. The bark, according to Sargent, forms the most valuable tanning material produced on the west coast of North America, and the inner bark is eaten by the Indians of Alaska. James M. Macoun3 says of it—"The abundance of other wood of better quality has prevented the hemlock from coming into general use, and the same prejudice exists in British Columbia against the western tree that prevailed until very recently against hemlock in eastern Canada. Though its grain is coarse, western hemlock is for many purposes just as serviceable as other woods which cost more. The bark is rich in tannin, but is too thin to be extensively used while there is such an abundance of Douglas fir in the same region." (H. J. E.) 1 These are trees growing in peat soil at 635 feet altitude. The seeds were sown in 1853, and in 1868 one tree was 29 feet by I ft. II in., and the other 26 feet by 2 feet at three feet from the ground (Gai-ti. Chroti. 1868, p. 518). 2 Alien, "Western Hemlock," 20, 21 (U.S. Forestry Bulletin, No. 33, 1902). 3 Forest Wealth of Canada, 82 (1904). Tsuga 239 TSUGA CANADENSIS, HEMLOCK OR HEMLOCK SPRUCE Tsuga canadensis, Carrière, Traité Conif. 189 (1855); Sargent, Silva N. Amer. xii. 63, t. 603 (1898), and Trees N. Amer. 48 (1905); Kent, Veitch's Man. Coniferce, 463 (1900). Pinus canadensis, Linnœus, Sp. PI. 1421 (1763); Lambert, Genus Pintis, \. t. 32 (1803). Abies canadensis, Michaux, FI. Bor. Am. ii. 206 (1803), and Hist. Arb. Amer. i. 137, t. 13 (1810); Loudon, Arb. et Frut. Brit. iv. 2322 (1838). Picea canadensis, Link, Linncea, xv. 523 (1841). A tree attaining in America over 100 feet in height, but usually only 60 to 70 feet, with a girth of 12 feet as a maximum. Bark of old trees brownish and deeply divided into narrow rounded ridges, covered with appressed scales. Young shoots greyish in colour and covered with short stiff pubescence. Leaves pectinately arranged, the shorter ones on the upper side of the shoot ; those on the median line above pointing forwards, appressed to the twig, and displaying their white under surfaces. They are ^ to f inch long, linear, usually broadest towards the base and tapering to the apex, which is rounded or acute ; distinctly and sharply serrulate in margin ; dark green above with a median groove often not continued to the apex ; lower surface with distinct midrib and two narrow well-defined white stomatic bands, the edges being pure green in colour. Buds brown, ovoid, pointed ; scales ciliate, pubescent, keeled, acute. Cones, \ to f inch long, ovoid, on slender puberulous stalks nearly £ inch long, composed of five series of scales, with about five scales in each series. Scales orbicular oblong, nearly as broad as long, entire and slightly bevelled in margin, striate, glabrescent in the exposed part. Bract small, concealed, lozenge-shaped. Seed with an oblong wing, decurrent half-way on its outer side. The seed with wing about two-thirds the length of the scale. VARIETIES A considerable number of horticultural varieties are known, no less than fourteen being described by Beissner. Some of these are variegated forms, as var. argentea or albo-spica, in which the tips of the young shoots are whitish. Others differ in habit and stature, as var. pendula, with pendulous branches, and var. Sargentii? a flat-topped bushy form of compact habit with short pendulous branches. The latter was found about forty years ago on the Fishkill Mountains in New York, and was first cultivated and made known by Mr. H. W. Sargent. One of the original plants, growing on the Howland estate, in Matteawan, New York, is now about 25 feet across. Grafted plants of this variety form in a few years an erect stem, and lose the dense low habit which is the charm of the original seedlings.1 Var. parvifolia, as cultivated at Kew, is a shrub, with stout branchlets, and very short leaves, about ^ inch long, which spread radially outwards from the shoot. (A. H.) 1 Sargent, Garden and Forest, x. 490 (1897). 240 The Trees of Great Britain and Ireland DISTRIBUTION In the colder parts of New England and Canada the hemlock is one of the most characteristic trees of the virgin forest, and extends, according to Sargent, from Nova Scotia and New Brunswick westward through Ontario to eastern Minnesota, southwards through Delaware, southern Michigan, and central Wisconsin, and along the Appalachian Mountains to north-western Alabama. He says that it attains its largest size in the south, in the mountain valleys of North Carolina and Tennessee, and gives its size as usually 60 or 70 and occasionally 100 feet in height, with a trunk 2 to 4 feet in diameter; but Pinchot and Ashe (loc. cit. p. [34) give no feet with a diameter of 6 feet as its extreme size, with a beautiful picture of it (pi. xix.). When, however, I was at Ottawa in September 1904 I visited, in company with Mr. James M. Macoun of the Geological Survey, a forest near Chelsea, in the Gatineau valley, where several hemlocks of nearly 100 feet were standing, mixed with birches, maples, and other hardwoods, and found a fallen tree which must have been at least 125 feet, and perhaps 135 feet long, though the top was too rotten to follow it out to the end. Mr. Macoun, however, said he had never seen one so large before. It often grows on rocky ridges, where it forms dense groves on the north side, and loves the steep banks of river gorges. Henry visited in 1906 Pisgah Mountain, near Hinsdale, in New Hampshire, where there remain on the estate of Mr. Ansell Dickinson about 700 acres of virgin forest. This mainly consists of a mixture of hemlock and hardwoods, with white pine occurring here and there singly and in small groups ; though on one or two areas of a few acres the white pine and hemlock form a pure coniferous stand. The largest hemlock seen measured 113 feet by 7 feet 10 inches, with a clean stem of only 30 feet, being much branched though densely crowded by other trees. A great many small hemlocks throughout the forest formed an undergrowth, and had been suppressed in growth, one which was f inch in diameter and 10 feet high showing 65 annual rings. In the Arnold Arboretum, near Boston, is a fine natural grove of this tree, called Hemlock Hill, which gives a very good idea of its normal growth in New England. The average height here is 60 to 70 feet by 3 to 4 feet, and the best that I measured at the bottom of the hill was 80 feet by 4 feet 6 inches. These trees were rather crowded, and had clean boles for 15 to 30 feet up. The growth of the tree is very slow, and Sargent says that the specimen of its timber in the Jessup Collection in the American Museum of Natural History at New York (which is the most complete that has ever been formed of the woods of any country) is only 13^ inches in diameter inside the bark, though it shows 164 annual rings, of which the sapwood, 2 inches thick, has twenty-nine. It seeds freely, but the seedlings do not germinate well in the open or on land which has been recently burned over, and seem to succeed best on a mossy stump or fallen log, where they must often remain eight to ten years before their roots reach the earth. According to Sargent they are only three or four inches high at four years old, under favourable conditions, and are easily destroyed. Tsuga 241 CULTIVATION Though introduced by Peter Collinson about 1736/ and at one time planted in almost every garden as an ornamental tree, the hemlock is rarely seen in Europe in a condition to remind the American of it as he knows it at home. Of late years it has been superseded by more modern and faster growing introductions. I cannot exactly say what are the conditions which suit it best in this country, because I have not seen it planted in the shady, damp, and rocky gorges which it likes at home ; but a deep light soil, free from lime and well drained, and a northern aspect, seem to suit it best in gardens. Its graceful habit and perfect hardiness should recommend it to all lovers of trees. It has a general tendency to fork near the ground, and this can only be checked by crowding it when young, or perhaps to some extent by careful pruning, as Loudon says that it bears the knife well, and is used for hedges in American nurseries ; though I should consider either common spruce or arbor vitae much better suited for the purpose here. It ripens seed freely, but the plants I have raised were so small that frost and March winds destroyed them before I learned the necessity of protecting them ; and in future I would imitate nature, and sow them on a mossy piece of half-rotten wood, or in a mixture of sand and leaf mould in a shaded frame. REMARKABLE TREES By far the most remarkable specimens of this tree which exist in England, or, as I believe, in Europe, are at Foxley, Herefordshire, the seat of the Rev. G. H. Davenport, which are believed to have been planted by Sir Uvedale Price, who was once the owner of this place. He was born in 1747, and died in 1828. In Nash wood, about half a mile from the house, on a rich soil of old red sand stone formation, in a dell facing south-west, a number of these trees are growing, which, though not quite so large as the tree at Studley, average about 55 feet high by 8 to ID in girth, and although their trunks are not so straight and clean as in an American forest, are nearly all sound and healthy. I measured twenty of these trees in July 1906 and found the largest, the only one which was forked near the ground, to be 10 feet in girth. Another was 9 ft. 10 in., and had a trunk which would contain from 120 to 130 cubic feet. The others ranged from 7 to 9^ feet at 5 feet from the ground, averaging over 8 feet, and were mostly clear of branches, or nearly so, for 15 to 30 feet from the ground. The dense shade of these trees keeps the soil quite free from vegetation below them, but I saw no seedlings in the grove. Though Mr. Davenport was good enough to have a considerable clearing made in order to get a better view of the trees, and Mr. Foster went to Foxley on purpose to photograph them, the difficulty of the subject was so great that the prints taken (Plate 70) do not show them as well as I could wish. The largest tree which I have seen in England is at Studley Royal, not far below 1 A tree said to be the original one planted by him at Mill Hill still survives, but was, when I saw it in 1906, in poor condition, the soil being too dry for it. H G The Trees of Great Britain and Ireland Fountains Abbey, and close to two very tall spruce. This, though hard to measure correctly owing to its crowded position, which makes a satisfactory illustration impossible, is over 80 feet hi