The source of this uncorrected OCR text may be viewed in the DjVu format at: http://fax.libs.uga.edu/QK488xE4/7tgbi or http://purl.galileo.usg.edu/ugafax/QK488xE4/7tgbi VOLUME VII Great Britain Ireland BY Henry John Elwes, F.R.S. AND Augustine Henry, M.A. Edinburgh : Privately Printed THE TREES OF GREAT BRITAIN AND IRELAND A. HENRY „. j. **• < *B if *.. ' Ift* , f » > ,< \ .• 'C \ - T! in . ;• \ »>. l1-*, •;, . M ' '*' t, A The Trees rf Great Britain 4 ^* Ireland BY Henry John Elwes, F.R.S. * < .. ? Si '< * •* .. J > AND Augustine Henry, M.A. VOLUME VII Edinburgh: Privately Printed MCMXIII CONTENTS LIST OF ILLUSTRATIONS TILIA . ... TlLIA CORDATA, SMALL-LEAVED LlME TILIA PLATVPHVLLOS, LARGE-LEAVED LIME . TILIA VULGARIS, COMMON LIME TILIA EUCHLORA .... TILIA TOMENTOSA, WHITE LIME TILIA PETIOLARIS, WEEPING WHITE LIME . TILIA MONGOLICA .... TILIA PAUCICOSTATA .... TILIA OLIVERI TlLIA MANDSHURICA TILIA MAXIMOWICZIANA. TILIA MIQUELIANA .... TlLIA AMERICANA, AMERICAN LlME, BASS-WOOD TlLIA HETEROPHVLLA TILIA MICHAUXII .... TRACHVCARPUS .... TRACHYCARPUS FORTUNEI, CHUSAN PALM ACANTHOPANAX .... ACANTHOPANAX RICINIFOLIUM ACACIA ..... ACACIA DEALBATA, SILVER WATTLE . ACACIA MELANOXVLON, BLACKWOOD . LAURELIA ..... LAURELIA SERRATA .... ILEX ..... ILEX AQUIFOLIUM, COMMON HOLLY Buxus ..... BUXUS SEMPERVIRENS, COMMON Box CRAIVEGUS ..... 111 vu 1653 1656 1661 1664 1674 1675 1677 1679 1680 1681 1682 1683 1684 1685 1688 1689 1690 1690 1694 1694 1697 1697 1699 1700 1701 1702 1704 1721 1724 I73I IV The Trees of Great Britain and Ireland CRATVEGUS MONOGYNA, HAWTHORN, WHITETHORN CRAT.ŒGUS OXYACANTHA, HAWTHORN, WHITEHORN SALIX ...... SALIX CAPREA, SALLOW, GOAT WILLOW SALIX PENTANURA, BAY WlLLOW SALIX BABYLONICA, WEEPING WlLLOW SALIX FRAGILIS, CRACK WILLOW SALIX ALBA, WHITE WILLOW SALIX CŒRULEA, CRICKET-BAT WILLOW SALIX VITELLINA, GOLDEN WILLOW . POPULUS ...... POPULUS ALBA, WHITE POPLAR POPULUS CANESCENS, GREY POPLAR . POPULUS TOMENTOSA .... POPULUS TREMULA, ASPEN .... POPULUS TREMULOIDES, AMERICAN ASPEN . POPULUS GRANDIDENTATA .... POPULUS SlEBOLDII ..... POPULUS FREMONTII .... POPULUS NIGRA, BLACK POPLAR POPULUS MONILIFERA, CANADIAN BLACK POPLAR POPULUS ANGULATA, CAROLINA POPLAR POPULUS SEROTINA, BLACK ITALIAN POPLAR POPULUS REGENERATA .... POPULUS EUGENEI ..... POPULUS MARILANDICA .... POPULUS HENRYANA .... POPULUS ROBUSTA .... POPULUS LLOYDII ..... POPULUS ANGUSTIFOLIA .... POPULUS BALSAMIFERA, BALSAM POPLAR POPULUS CANDICANS, ONTARIO PoPIAR POPULUS TRICHOCARPA, WESTERN BALSAM POPLAR . POPULUS MAXIMOWICZII .... POPULUS SIMONII ..... POPULUS TRISTIS ... POPULUS SUAVEOLENS POPULUS LAURIFOLIA POPULUS LASIOCARPA ULMUS ...... FAGE 1733 1739 1745 1747 1749 1754 1763 1768 1770 1777 1780 1786 1787 1791 1792 1794 '794 1807 1810 1816 1824 1826 1828 1829 1829 1830 1831 1832 r836 1838 1840 1841 1842 1846 1847 Contents ULMUS PEDUNCULATA, EUROPEAN WHITE ELM ULMUS AMERICANA, AMERICAN WHITE ELM ULMUS RACEMOSA, ROCK ELM " . . ULMUS FULVA, SLIPPERY ELM . ULMUS MONTANA, WYCH ELM . ULMUS VEGETA, HUNTINGDON ELM, CHICHESTER ELM ULMUS MAJOR, DUTCH ELM . ULMUS NITENS, SMOOTH-LEAVED ELM ULMUS MINOR, GOODYER'S ELM . ULMUS CAMPESTRIS, ENGLISH ELM .... ULMUS JAPONICA ... . ULMUS ALATA, WAHOO, WINGED ELM ULMUS CRASSIFOLIA ...... ULMUS PUMII.A ...... ULMUS PARVIFOLIA ...... KOEI.REUTERIA ...... KOELREUTERIA PANICULATA ..... PAGE l85I 1860 1862 1864 1879 1883 1887 1901 1903 1923 1924 1925 1926 1928 1930 1932 ILLUSTRATIONS Portrait of the Authors ...... . Frontispiece PLATE No. Small-leaved Lime at Sprowston, Norwich . . . . . . . 372 White Lime at Albury .......... 373 Weeping White Lime at Hatherop Castle . . . . . . -374 Chusan Palm at Lamorran . . . . . . . . -375 Acacia dealbata at Derreen . . . . . . . -376 Holly at Gordon Castle . . . . . . . . -377 AVhitethorn at Hethel . . . . . . . .378 Weeping Willow at Cheltenham ... • • 379 White AVillow at Haverholme . . . . . 380 Cricket-bat Willow at Hertford ...... .381 Grey Poplar at Colesborne .... . . 382 Female Lombardy Poplar in Brunswick . . . 383 Carolina Poplar at Danny Park ........ 384 Fastigate Black Poplars in Belgium ........ 385 Black Italian Poplar at Belton ........ 386 Balsam Poplar at Bute House, Petersham ....... 387 Western Balsam Poplar in Vancouver Island ...... 388 European White Elm at Syon ....... 389 European White Elm at Ugbrooke ........ 390 American White Elm in Massachusetts . ..... 391 American White Elm at Hargham ........ 392 Weeping Wych Elm at Glasnevin .... . . 393 Wych Elm at Studley Royal ......... 394 Huntingdon Elm at Magdalen College, Oxford . . • • 395 Elms in Kensington Gardens ..... . 396 Cornish Elms at Coldrenick . . . . . . . . -397 Wheatley Elm at Richmond . . . ... 398 Smooth-leaved Elm at Sharpham ... . . 399 Wych Elm at Cassiobury ......... 400 vii The Trees of Great Britain and Ireland Smooth-leaved Elm at Briggins Smooth-leaved Elm at Saling Goodyer's Elm at Weston Birt . English Elm at Powderham . English Elms at Well Vale . Fastigiate Beech at Dawyck . Tilia ; leaves, etc. Populus; leaves, etc. Populus ; leaves, etc. Populus ; leaves, etc Ulmus ; leaves, etc. . Ulmus ; leaves, etc. . PLATE No. 401 402 403 404 405 406 407 408 409 410 411 412 TILIA Tilia, Linnœus, Hort. Cliff. 204 (1735), Sp. PI. i. 514 (1753), and Gen. PL 267 (1767); Bentham et Hooker, Gen. PL i. 236, 986 (1862-7); V. Engler, Monog. Gatt. Tilia, pp. 1-159 (1909). DECIDUOUS trees, belonging to the order Tiliaceae, with tough fibrous inner bark. Leaves simple, long-stalked, alternate, arranged on the branchlets in two rows ; unequal* and cordate or truncate at the base ; acute or acuminate at the apex ; serrate or toothed ; venation pseudo-palmate, the midrib giving off secondary nerves pinnately on both sides, the lower two pairs of which arise together at the base, and give off tertiary nerves on the outer side only. Stipules ligulate, membranous, caducous. Flowers white, fragrant, regular, perfect, in cymes ; peduncle connate with the axis of a membranous elongated persistent bract, from the middle of which it apparently arises ; inflorescence and bract springing from the axil of a leaf, alongside a bud, which develops into a branch in the following year.2 Sepals five, distinct ; petals five, imbricate in bud. Staminodes either absent or present as petaloid scales, one opposite each petal, and united with the base of the stamens. Stamens numerous, free, or in five clusters united together at the base. Filaments unbranched, or forked at the apex, with each branch bearing an extrorse half-anther. Ovary sessile, five- celled ; style erect, dilated at the apex into five spreading stigmatic lobes ; ovules two in each cell. Fruit nut-like, dry, indéhiscent, one-celled, and one- to two-seeded by abortion. Seeds obovate, with fleshy albumen. Cotyledons reniform or cordate, palmately five-lobed, raised above ground on germination. In winter the twigs are zig-zag and bear lateral buds, disposed alternately in two ranks ; each bud with two to three scales visible externally, ovoid, obliquely displaced to one side of the semicircular leaf-scar, which is set on a prominent pulvinus. Stipule-scars small, linear or oblong, one on each side of the leaf-scar. Terminal bud absent ; a circular scar at the apex of the twig, opposite the uppermost leaf-scar, indicating where the tip of the branchlet fell off in early summer. Base of the branchlet girt with a ring of scars, due to the fall of the bud-scales in the previous spring. About twenty species of Tilia can be distinguished. These are widely distri buted in the temperate regions of the northern hemisphere, extending southward in North America as far as the highlands of Mexico ; but in the old world, while common in Europe and in northern and eastern Asia, no species is known in 1 Cf. Van Tieghem, in ^4«». Set. Not. (Bot.) iii. 378 (1906), on the peculiar asymmeliy in the leaves and stipules of the lime. 2 The occurrence of an inflorescence and a normal bud in one and the same axil is unusual ; and is explained by the feet that the former is the result of the very early development of a flower-bud under the first scale of the normal bud, the other scales of the latter remaining dormant until the following season. VII 1653 B 1654 The Trees of Great Britain and Ireland northern Africa or in the Himalayas. A large number of hybrid forms have arisen, some of which are common in cultivation. The following key, based on the characters of the branchlets and leaves, will serve to distinguish the species and hybrids which are cultivated in this country. I. LEAVES GREEN BENEATH, WITH AXIL-TUFTS OF PUBESCENCE, BUT WITHOUT ANY STELLATE TOMENTUM. (a) Axil-tufts^ present at the base of the leaf and elsewhere. * Branchlets glabrous, or nearly so. Leaves glabrous, except for axil-hifts beneath. 1. Tilia mongolica, Maximowicz. North China and Mongolia. Seep. 1679. Branchlets quite glabrous, reddish. Leaves 2\ in. wide, often trilobed, glaucous beneath with non-prominent tertiary veins, coarsely toothed. 2. Tilia cordata, Miller. Europe, Caucasus. See p. 1656. Branchlets slightly pubescent at first, speedily becoming glabrous. Leaves 2 to 2 J in. wide, bluish green beneath with non-prominent tertiary veins, finely serrate. 3. Tilia vtilgaris, Hayne. A hybrid, occasionally wild in Europe. See p. 1664. Branchlets quite glabrous. Leaves 3 in. wide, dull green above, pale green beneath and with prominent tertiary veins, finely serrate with short points to the teeth. 4. Tiliaeuchlora, Koch. A hybrid, occasionally wild in the Caucasus. Seep. 1674. Branchlets usually quite glabrous. Leaves 2\ to 3 in. wide, dark shining green above, pale green beneath and with prominent tertiary veins, finely serrate with long points to the teeth. ** Branchlets densely ptibescent -with long hairs. 5. Tilia platyphyllos, Scopoli. Europe. See p. 1661. Leaves 3 to 4 in. wide, upper surface with short scattered pubescence, lower surface covered with long hairs. (b) Axil-tufts absent at the base of the leaf, present elsewhere. 6. Tilia americana, Linnaeus. North America. See p. 1685. Branchlets glabrous. Leaves 5 to 6 in. long and 3^ to 4^ in. wide, broadly ovate, cordate at the base, glabrous beneath and with numerous prominent parallel tertiary veins ; margin with long-pointed coarse serrations. 7. Tiliapaucicostata, Maximowicz. Western China. See p. 1680. Branchlets glabrous. Leaves 2\ in. long and 2 in. wide ; ovate, usually truncate at the base, glabrous beneath with few prominent irregular tertiary veins ; margin with long-pointed fine serrations. II. LEAVES GREEN OR GREYISH BENEATH, WITH SCATTERED STELLATE TOMENTUM. (a) Under surface of the leaves without axil-tiifts, but with long hairs on the midrib. 8. Tilia Moltkei, Schneider. See p. 1686. A hybrid, with large leaves similar to those of T. americana in shape and serrations. Buds and branchlets glabrous. 1 These are tufts of hairs at the junctions of the midrib and lateral nerves on the under surface of the leaf, which are now often termed domatia ; they are the abodes of mites, and serve a useful purpose in the economy of the tree. They were fully studied and described by Lundström, in Nov. Act. Keg. Sac. Sei. Ufsala, xiii. pt. 2, pp. 3-10 (1887). Cf. Lord Avebury, Brit. Flffivering Plants, 123 (1905). Tilia l6S5 9. Tilia spectabilis, Dippel. See p. 1686. A hybrid similar to T. Moltkei, but with smaller leaves, which have long hairs on the principal nerves, as well as on the midrib. Buds pubescent in their upper half. Branchlets with traces of stellate pubescence. (b) Under surface of the leaves with axil-tufts. 10. Tilia Michauxii, Nuttall. North America. See p. 1689. Leaves usually large,1 5 to 7 in. in length and 4 to 6 in. wide ; ovate-cordate, very oblique at the base, with long-pointed large triangular serrations. Buds and branchlets glabrous. III. LEAVES WHITE OR GREY BENEATH, AND COVERED WITH A DENSE STELLATE TOMENTUM. (a) Branchlets glabrous. * Axil-tufts present. IDA. Tilia Michauxii, Nuttall. In some forms of this species the leaves are densely greyish tomentose beneath. See above, No. io. * Axil-tufts absent. it. Tilia Jieterophylla, Ventenat. North America. Seep. 1688. Leaves ovate-cordate, very oblique at the base, 4 to 5 in. long, 3 to 4 in. wide, covered beneath with a silvery white tomentum ; serrations coarse and short-pointed. 12. Tilia Oliveri, Szyszylowicz. China. See p. 1681. Leaves orbicular-ovate, 3 to 4 in. long, z\ to 3 in. wide, silvery white beneath ; serrations minute, crenate. (b) Branchlets pubescent. *'Axil-tufts present. 13. Tilia Maximowicziana, Shirasawa. Japan. See p. 1683. Leaves orbicular-ovate, averaging 5 in. long and broad, covered beneath with a greyish tomentum ; axil-tufts and tomentum on midrib and nerves brownish. ** Axil-tufts absent. 14. Tilia tomentosa, Moench. South-eastern Europe, Asia Minor. See p. 1675. Leaves orbicular-ovate, 3 to 5 in. across, greyish or snowy white beneath, with stout or slender short petioles ; serrations fine, regular, ending in short points. Buds, branchlets, and petioles grey tomentose. 15. Tilia petiolaris, Hooker. See p. 1677. Possibly a sport of T. tomentosa, from which it differs in the pendulous habit of the tree, the long" slender petioles, and the peculiar fruit. 16. Tilia mandstmrica, Ruprecht and Maximowicz. Manchuria, North China, Korea. See p. 1682. Leaves orbicular-ovate, 4 to 5 in. across, white beneath ; margin often one- to two-lobed, with coarse serrations, ending in long awn-like points. Branchlets, buds, and stout petioles brown tomentose. 1 In native specimens the leaves are smaller, averaging 4 to S in. long anil 3 to 4 in. wide. 1656 The Trees of Great Britain and Ireland 17. Tilia Miqueliana, Maximowicz. Cultivated in Japan. See p. 1684. Leaves remarkably variable in shape, deltoid or ovate, usually much longer than broad, 3 to 4 in. long and 2 to 2^ in. wide, grey beneath ; serrations irre gular, ending in short points. (A. H.) TILIA CORDATA, SMALL-LEAVED LIME Tilia cordata, Miller,1 Gard. Diet. No. i (1768); Moench, Verz. Ausl. Weissenst. 135 (1785); Schneider, Laubholzkunde, ii. 372 (1909); V. Engler, Monog. Gatt. Tilia, 74 (1909). Tilia europœa, Linnœus, Sp. PL 514 (1753) (in part); Loudon, Arb. et Frut. Brit. i. 364 (1838). Tilia iilmifolia, Scopoli, Fl. Cam. i. 374 (1772); Sargent, in Garden and forest, ii. 256, f. 111(1889). Tilia parvifolia, Ehrhart, Beitr. Naturk. v. 159 (1790); Willkomm, Forstliche Flora, 729 (1887); Mathieu, Flore Forestière, 29 (1897). Tilia microphylla, Ventenat, in Mem. Acad. Se. Paris, iv. 5 (1803). Tilia sylvestris, Desfontaines, TaMe Éc. Bot. Mus. Paris, 152 (1804). A tree, attaining ioo ft. in height and 20 ft. in girth. Bark smooth and grey on young trees ; ultimately on old trunks divided by narrow longitudinal fissures into scaly ridges. Young branchlets green, slightly pubescent at first, speedily becoming glabrous, the pubescence, however, being often retained on short shoots ; older branchlets dark brown. Leaves2 (Plate 407, Fig. 8), membraneous, 2 to 2^ in. wide, usually broader than long, smooth and not wrinkled, cuspidate at the apex, cordate at the base ; margin non-ciliate, regularly serrate, the teeth ending in short cartilaginous points ; upper surface dark green, shining, glabrous except for occasional long hairs on the nerves ; lower surface bluish or glaucous green, glabrous except for the conspicuous dense orange-brown axil-tufts at the base, and at the junctions of the midrib, primary, and secondary nerves ; tertiary veins scarcely prominent on the under surface, and more irregular, and less straight and parallel than in T. vulgaris and T. platyphyllos ; petiole about half as long as the blade, slender, glabrous, or with a few scattered hairs. Cymes directed upwards, five- to seven-flowered ; bract long-stalked, glabrous ; pedicels glabrous or with a few scattered hairs ; sepals pubescent ; petals glabrous ; stamens about thirty, longer than the petals ; staminodes absent ; ovary tomentose, style glabrous. Fruit globose, faintly ridged, apiculate at the apex, covered with long scattered tomentum ; shell thin and fragile. In winter the buds are more globose than those of T. vulgaris or T. platyphyllos, and appear to be composed of two external scales, though the 1 E. G. Baker, \njourn. Bot. xxxvi. 318 (1898), states that Miller's specimen in the British Museum is T. platyphyllos ; but there is no evidence that this is a lype specimen. It is plain from Miller's statement that T. cordata " grows naturally in the woods in many parts of England," and from his identification of it with Tilia foemina, folio minore, C. Bauhin, Finax, 426 (1671), that he meant the small-leaved lime. 1 The leaves on coppice shoots in the first year are remarkably large. Mr. Riddelsdell sent us specimens from Glamorganshire, with leaves 5 to 7 in. long and nearly as broad, coarsely toothed, deeply and narrowly cordate at the base, ending at the apex in a long acuminate point, and on short petioles scarcely an inch in length. As the coppice shoots lengthen year by year, the leaves gradually assume their normal form, small in size, broader than long, and with long petioles. Lees, in Bot. Worcestershire, 16 (1867), argues from the variable appearance of the leaves of coppice shoots of T. parvifolia, that the common lime is only a variety of the latter. The coppice shoots of most broad-leaved trees have peculiar leaves, different from those in the adult state, and more alike in allied species, so that their discrimination is difficult. Tilia 1657 pubescent tip of a third scale may be discerned at the apex of the bud ; the first and second scales are shining, glabrous, ciliate. This species is readily distinguished by the bluish tint of the under surface of the leaves, which are very different in their tertiary venation from the other common limes. The erect and not pendulous cymes of flowers are also a peculiar feature. VARIETIES This species, as limited here to the European and Caucasian small-leaved lime, displays little variation in the wild state, the varieties1 established by Schneider on the shape and size of the leaf and the amount of the pubescence on the fruit, being probably due to soil conditions, and not worth enumerating. A few peculiar sports have been noticed, none of which appear to be known in England :— 1. Var. vitifolia, Schneider, op. cit. Leaves three-lobed. Wild in Hungary. 2. Var. aureo-variegata, Schneider,2 op. cit. Leaves variegated with yellow. 3. Var. cîtcullata, Henry (var. nova). Similar to the variety so named of T. platyphyllos. De Vries, Species and Varieties, 355, 669 (1906) and Mutation Theory, 470, fig. 106 (1910), draws attention to a tree with peltate and pitcher-like leaves, which is growing at Lage Vuursche, near Amsterdam. DISTRIBUTION The small-leaved lime is a native of the greater part of Europe and of the Caucasus, the closely allied forms3 in Siberia, Manchuria, and Japan being now regarded as distinct species. In Europe, it extends from northern Spain to the Ural range, attaining its maximum development in Russia, where it occasionally forms pure woods, but more usually, as is always the case elsewhere, growing as isolated trees or in small groups with other deciduous trees. It occurs as far north as the province of Volgoda, where it disappears after becoming a small shrub at lat. 62°. In the Ural, it reaches as far north as lat. 58° 50'. The finest lime woods 1 Var. Blockiana, Schneider (T. Blockiana, Borbas), and var. ovalifolia, Spach, with leaves larger and less cordate than usual, are possibly of hybrid origin. 2 Tilia iilmifolia, Scopoli, \3.r.foliis variegatis, Petzold and Kirchner, Arb. Muse. 156 (1864), is another name for this variety. 3 The Asiatic forms are distinguished as follows from the European T. cordata :— A. Tilia sibirica, Bayer, in Verh. Z. Bot. Ges. Wien. xii. 23 (1862). Tilia cordata, var. sibirica, Maximowicz, in Bull. Ac. St. Petersb. xxvi. 432 (1860). Indigenous in western Siberia. Not yet introduced. Differs mainly in the leaves, truncate or cuneate at the base, with sharper serrations, and long hairs on the nerves. B. Tilia amurensis, Ruprecht, Fl. Cane. 253 (1869). Tilia Maximowiczii, Baker, injottrn. Bot. xxxvi. 319 (1898). Tilia cordata, var. mandslmrica, Maximowicz, in Mil. Biol. x. 584 (1880). Indigenous in Manchuria and Korea. Not yet introduced. Differs in the larger leaves, with fewer coarser serrations, which are tipped with long points. C. Tilia japonica, Simonkai, in Math. Term. Koezl. xxii. 326 (1888). Tilia cordata, var. japonica, Miquel, in Ann. Mus. Lugd. Bat. iii. 18 (1867) ; Sargent, in Garden and Forest, vi. ill (1893), and Forest Flora Japan, 20(1894); Shirasawa, Icon. Ess. Forest. Japon, i. text 115, t. 72, figs. I-IO ( 1900). Indigenous in Japan, where it is a small tree, rarely higher than 60 ft. It was introduced into the Arnold Arboretum, U.S.A., in 1886, where it is hardy, producing flowers and fruit every year ; but is said to be scarcely distinct from the European species. It appears to differ mainly in the flowers, which are 20 to 40 in each cyme, and possess staminodes. Specimens collected by Elwes at Asahigawa in Yezo show no difference in leaves and branchlets. 1658 The Trees of Great Britain and Ireland in Russia are in the region extending southwards from Kostroma to the edge of the steppe ; and here both this species and T. platyphyllos grow together. In Norway it is found as a wild tree as far north as lat. 62° 9' on the west coast, and in Sweden up to 63° ic/ in Angermanland ; but according to Schübeler, it thrives when planted as far north as 67° 56' in Norway, 65° 50' in Sweden, and 63° in Finland. It appears to be not a native tree in Belgium, Holland, Denmark, and north-western Germany ; and is nowhere very common in central Europe at present, though it is supposed to have been more widely spread in ancient times, as the word linden is very prevalent in German and Slavonic names of places. It is rather a tree of the plains than of the mountains; but it ascends in Bohemia and Bavaria to 2000 ft., and in Switzerland and the Tyrol to 4000 ft. Bolle informed Sargent that very old and enormous trees1 of this species, one being nearly 23 ft. in girth, exist at Paelitzaerder on the Paarestein lake near Eberswalde. In France, it is met with in most of the forests of the plains and low hills, except in the departments bordering on the Mediterranean. It is occasionally treated as coppice, being used for firewood and making charcoal. Bast, which was formerly a product of some importance, is now only produced in the forest of Chantilly, nearly all the bast used being now imported from Russia. Mathieu mentions a tree, planted at Gerardmer in the Vosges, which measured 95 ft. in height and 19 ft. in girth, and was supposed to be at least 250 years old. The small-leaved lime extends southwards to about lat. 41°, occurring in northern Spain, Italy, and the Balkan States ; but is unknown in Greece and Sicily. Huffel2 says that both it and T.platyphyllos are common in the forests of the hills of Dobrudja, Roumania, where they are the dominant trees. The small-leaved lime is a native of England, ranging from Cumberland south ward. It occurs in woods in rather inaccessible positions, where it is a rare tree, and more commonly in coppice, situations in which the indigenous vegetation has often been preserved. Ray8 considered this species to be a true native ; and in his time it was frequent and wild in woods and coppices in Essex, Sussex, Lincolnshire, and especially in Bedfordshire, " where there were thousands of lime trees." He adds that it was less common in the Forest of Dean, and rare in Cranborne Chase in Dorset. Many of the local floras give instances of its occurrence, as J. G. Baker4 for Yorkshire, who states that it occurs " at Slip Gill near Rievaulx, where aboriginal woods composed principally of oak and hazel cover the steeply-sloping rocky banks of one of the loneliest and pleasantest glens in the eastern calcareous range." Ley5 records it for different parts of Herefordshire. Murray6 says it is abundant in the Leigh woods near Bristol. It is said7 to be wild in several localities in Glamorgan shire. Bromfield8 mentions wild trees in one locality in the Isle of Wight, and in aboriginal woods on the chalk at Bordean Hill, near Petersfield, Hants. Bromfield supposes that Lyndhurst, in the New Forest, owes its name to the 1 Bean, in A'ew Bull. 1908, p. 397, mentions a tree in the Grosse Garten, Dresden, branching close to the ground, where the trunk was about 8 ft. through. 2 Les Forêts de la Roumanie (1890). 3 Syn. Meth. 316 (1696) and Philos. Letters, 250 (1718). 4 Flora N. Yorks. 274 (1906). 6 Flora Herefordshire, 54 (1889). ° Flora Smnerset, 64 (1896). 7 Riddelsdell, injourn. Bot. Suffi. 18 (1907). 8 FI. Vect. 83 (1856). Tilia prevalence of this species there in ancient times. Limehouse, in London, according to Stowe, was originally called Limehurst, meaning a grove of linden trees in Saxon times. (A. H.) Some doubt exists among botanists as to whether the small-leaved lime is truly native of England or not. It is not mentioned by Clement Reid as having been found in the fossil state in Britain ; and by some it is supposed to have been introduced at an early epoch, perhaps by the Romans. But in some parts of the West Midlands it is found in woods remote from buildings, where one can hardly believe it was planted, so that it might fairly be considered a native but for one important fact. Notwithstanding many inquiries even in the districts where it now seems most at home, I have found no one who has seen a self-sown small-leaved lime. It seems hardly possible that a native tree should have lost its power of reproduction by seed, in a climate where it succeeds so well even as far north as Ross-shire ; and in the north of France self-sown seedling limes are not uncommon, as I have myself observed in the Forêt de Retz. The tree has a remarkable power of persistence after repeated cutting, and of extending from stools to a considerable distance ; so that in two old coppiced woods on my own property, it is now impossible to say where the stools originated. I have seen limes in remote' rocky woods on the Wye valley near Moccas Court, whose stools had the appearance of very great age ; and in the deep rocky gorge of Castle Eden Dene, on the coast of Durham, there are limes growing on such steep rocks that they could scarcely have been planted. But though rabbits will eat almost anything before they touch lime, I have searched in vain for seedlings in all these places. On the Carboniferous limestone rocks at Pen Moel near Chepstow, the residence of W. R. Price, Esq., I saw the tree growing in situations where it must have grown naturally from seed ; and though Mr. Price has never found ripe fruit he has not the least doubt that it is indigenous here and elsewhere on the cliffs of the lower Wye valley. E. Lees, in Botany of Worcestershire, 16 (1867) gives an excellent account of the occurrence of the lime in that county, where it is, in his opinion, " an undoubted native." He states that Shrawley wood, west of the Severn, which is about 500 acres in extent, is remarkable for a great part of it consisting of an undergrowth of lime, which is regularly cut as coppice-wood, and, therefore, is never in a flowering state.1 On visiting this place, I agreed with Sir H. Vernon, of Hanbury Hall, near Droitwich, the owner of the wood, that the stools are in rows as though they had been planted ; moreover there is not, so far as he knows, any lime in the adjoining woods. He says that this underwood used to be cut every seventeen years, and sent to the Potteries for making crates, but that this demand having ceased, it is now difficult to get rid of. It is now allowed to grow into poles, which are sometimes sold for copper-smelting in the Black Country, at about six or seven pounds per acre for twenty to twenty-five years' growth. In Sir H. Vernon's opinion, it would now pay better to grub the lime and plant larch in its place. Lees2 goes on to state that " Ockeridge wood, near Holt, though in a lesser 1 Cf. p. 1656, note 2. 2 Cf, also Lees' remarks in Forest and Chace ofMalverii, abstracted in Card. Chron. 1870, p. 1536. 1660 The Trees of Great Britain and Ireland degree, nourishes the same tree, as well as various coppices on the banks of the Severn between Ombersley and Hawford, where T. grandifolia exists in a naturalised state." He mentions "a very old and remarkable pollard tree of T. parvifolia at Hawford, on the ridge not far removed from the Severn. The base is more than 40 ft. round and six large boles arise from this in a semicircular manner. In fact, commencing with the border of Wyre forest and proceeding southward, the lime appears in numerous woods, coppices, and old hedgerows, to the very end of the Malvern range near Bromsberrow. The base of the round hills near Alberley, Ockeridge wood, the western base of the Berrow Hill near Martley, the banks of Leigh brook, Rosebury Rock on the Teme, the Old Storridge Hill, the country about Great Malvern, and ancient woods in the parishes of Castle Morton and the Berrow, may be particularly mentioned. Many of the old lime trees get pollarded, and then, in the course of years, put on a very grotesque appearance." REMARKABLE TREES The small-leaved lime apparently never attains so great a height in England as the common lime, but is occasionally of great girth and is certainly long-lived. A tree (Plate 372), remarkable for its spreading habit, at Sprowston Hall, Norwich, was figured by Grigor in Eastern Arboretum, 200 (1841), where it is stated that it measured 24 ft. 7 in. " near the ground " and was believed to have been planted on 3Oth January 1649. It still survives in a shattered condition. There is a remarkably fine tree of this species at The Hall, Thirsk, the seat of Reginald Bell, Esq., who has kindly sent us photographs. In 1904 the trunk in its narrowest part was 20 ft. in girth, and the spread of the branches was about 250 ft. in circumference. One of the finest small-leaved limes is growing on a flat by the River Teme, at Oakly Park, Ludlow, which, in 1908, as nearly as I could measure it, was about no ft. by 14^ ft. A fine tree, of weeping habit, at Hursley Park, Hants, the seat of Sir G. A. Cooper, Bart., measures about 80 ft. by 15^ ft. Close to it stands the hollow trunk of a much larger tree of the same species, which was blown down some years ago, and measures 19^ ft. in girth. The spread of its branches is said to have exceeded IOO ft. At Arley Castle, Bewdley, a good specimen measured, in 1903, 85 ft. by 9 ft. 9 in. At Woburn Abbey, the largest tree of this species measures 76 ft. by 7 ft. 4 in., but appears to be still young, as the bark is comparatively smooth. In Lincolnshire, the tree is not uncommon in parks and hedgerows. In Burghley Park there are several old trees, one of which measured, in 1908, 80 ft. by 11 ft. 4 in. At Casewick House, another was 82 ft. by 9 ft. 6 in. in the same year. At Syston Park there is a fine specimen, which measured 97 ft. by 11 ft. in 1906. (H. J. E.) Til ia 1661 TILIA PLATYPHYLLOS, LARGE-LEAVED LIME Tilia platyphyllos, Scopoli, Fl. Cam. i. 373 (1772); Sargent, in Garden and Forest, ii. 256, f. 109 (1889); Schneider, LauMwlzkunde, ii. 376 (1909). Tilia grandifolia, Ehrhart, Belt. v. 158 (1790); Willkomm, Forstliche Flora, 733 (1887); Mathieu, Flore Forestière, 33 (1897). Tilia paticiflora, Hayne, Arzn. iii. 48 (1813). Tilia corailina, Smith, in Rees, Cycl. xxxv. No. 2 (1819). Tilia mollis, Spach, in Ann. Se. Nat. ii. 336 (1834). Tilia europœa, Linnxus, Sp. PL 514 (1753) (in part) ; Loudon, Arb. et Frut. Brit. i. 364 (1838). A tree, attaining 130 ft. in height and upwards of 20 ft. in girth. Bark at first smooth and grey, ultimately on old stems with narrow shallow longitudinal fissures and ridges separating on the surface into small quadrangular scales. Young branch- lets moderately covered with long white hairs ; older branchlets glabrescent. Leaves (Plate 407, fig. 6) 3 to 4 in. in width and length, slightly uneven or wrinkled, ciliate in margin, regularly serrate, the serrations ending in short cartilaginous points ; upper surface dull green, covered with short pubescence ; lower surface lighter green, covered with long whitish pubescence, densest on the midrib, nerves, and veinlets, and forming dense axil-tufts at the base of the blade and at the junctions of the primary nerves with the midrib and with the secondary nerves ; tertiary veinlets parallel and prominent on the under surface ; petiole stout, shorter than the blade, whitish pubescent. Flowers in pendulous, usually three-flowered cymes ; about \ in. in diameter, yellowish - white ; sepals slightly pubescent externally, downy within; petals ob- lanceolate, longer than the sepals ; stamens about thirty, longer than the petals ; staminodes absent ; ovary globose, tomentose ; style glabrous. Fruit globose, pyriform, or ovoid, usually1 with three to five prominent ribs, tomentose, apiculate at the apex ; shell woody and hard. In winter this species may be recognised by the twigs being slightly pubescent near the buds, which are minutely pubescent at the tip and show externally three glabrous ciliated scales. VARIETIES This species in the wild state varies considerably in the amount of pubescence on the leaves, branchlets, and petioles ; and has been subdivided into five sub species by Schneider, who acknowledges, however, the great difficulty of limiting them clearly. The most pubescent forms occur in northern Germany, northern France, and Scandinavia ; while nearly glabrous forms are found in southern France, Austria, and the Balkan States. V. Engler disagrees with Schneider's classification ; and considers that the limes occurring in southern France, the Pyrenees, Italy, etc., should be united with T. caucasica ; but this view is hardly tenable. The bract is stalked in most cases, but is occasionally sessile ; and abnormal forms occur 1 A tree at Kew of undoubted T. platyphyllos, bore fruit in 1907, on which no trace of ribs was perceptible. VII C 1662 The Trees of Great Britain and Ireland (var. multibracteata) in which two bracts are borne on one peduncle. The fruit is remarkably variable, both in shape and in the prominence of the ribs. The southern more glabrous forms are rarely cultivated in England, the only specimen which I have seen being a tree at Kew about 25 ft. high, which is named var. obliqua? The branchlets are nearly glabrous ; leaves very oblique and truncate at the base, glabrous above, with scattered pubescence below. It bears flowers similar to those of the type. A large number of sports have arisen both under cultivation and in the wild state, the most noteworthy of which are :— 1. Var. pyramidalis, Simonkai, in Math. Term. Koezl. 334 (1888). Pyramidal in habit ; leaves usually more or less cordate at the base. According to Schneider this is occasionally wild in south-eastern Europe. 2. Var. tortuosa, Bean, m Kew Hand-list Trees, 71 (1902). A peculiar sport, with all the twigs and branches twisted and curved. This2 originated in the Royal Horticultural Society's garden at Chiswick in 1888 as a single specimen out of a bed of 500 large red limes. Grafts were sent to Kew from Chiswick in 1890, and three trees about 18 ft high survive, in the Lime collection. 3. Var. aurea, Loudon (var. aurantia\ Twigs golden yellow. 4. Var. corallina, Solander, in Aiton, Hort. Keiu. ii. 229 (1789). Twigs bright red. Both these varieties are conspicuous in winter, and have been known for more than a century. According to Koch they were probably introduced from England to the Continent. The latter is the red-twigged lime of some English nurseries. 5. Var. laciniata, Loudon (var. asplenüfolia, var. ßlicifolid). Leaves smaller than in the type, deeply and irregularly cut and twisted. This never attains a large size,3 and is only suitable for planting as a curiosity in gardens. It commonly throws out branches on which the foliage is normal. 6. Var. vitifolia, Simonkai, op. cit. Leaves lobulate or weakly three-lobed. 7. Forms with variegated leaves are known, as var. albo-marginata, Van Houtte. 8. Var. cucullata, Schneider (T. c^lcullata, Jacquin,* Frag. Bot. 19, t. ii, f. 3 (1800)). A form with small leaves, of which the edges of the two sides are joined together at the base, making the leaf pitcher-shaped. It is said to occur wild in southern Bohemia, where, according to Willkomm, there are some old trees, with all the leaves showing this peculiarity, at the monastery of Goldenkron, near Krumau. DISTRIBUTION This species is widely distributed throughout central and southern Europe, extending as far eastward as the Ural Mountains. Its northern limit as a wild tree is not known with certainty, and Willkomm considers it not to be indigenous in 1 Probably identical with T. obliqua, Host, in Schmidt, Oestr. £amns. iv. t. 224 (1822), and Host, Fl. Austr. ii. 62 (1831). The Kew tree agrees with a dried specimen collected in Host's garden in 1832. 2 Cf. Gard. Chron. iv. 708 (1888). 3 A. B. Jackson saw a tree at Blenheim, 40 ft. by 3^ ft., in 1908. * Jacquin figures leaves from trees in a cemetery at Sedlitz, near Kuttenberg in Bohemia. Leneck, in Mitt. Nat. Ver. Univ. Wien, 1893, pp. 19-29, figs. I-i I, gives an account of these abnormal leaves ; and records a large-leaved lime growing at Leitmeritz in northern Bohemia, of which 20 to 30 per cent of the leaves were pitcher- or cowl-shaped. Cf. Just, Bot. Jahresb. xxii. pt. 2, p. 219 (1894). Tilia 1663 northern Germany, Denmark, and the Baltic provinces of Russia. Bolle, however, states that it grows sparingly in these countries, and mentions small groups of wild trees growing in the islands on the west coast of Sweden, near Strömstad. It is most common in southern Russia, where in the provinces of Ukraine and Volhynia it often forms pure woods, though it is also seen in mixture with the small-leaved lime and Quercuspedimciilata. It is also frequent in the southern states and Rhenish provinces of Germany, and ascends in the Bavarian Alps to 3300 ft. It is also widely spread in Austria, Hungary, and the Balkan States ; and occurs in Italy and Spain, reaching its most westerly point in Asturias and New Castile. In France it is found scattered in the forests of the plain, except in the Mediterranean region, where, however, it has been observed as a rare tree in the Ravin des Arcs, 15 miles north of Montpellier.1 More common in the hills and mountains, it attains its highest elevation, 4600 ft., in the Pyrénées-Orientales. It is replaced in Greece by the closely allied species T. corinthiaca, Bosc, and in the Caucasus, north Persia, and Armenia by T. ca^^cas^ca, Ruprecht. Neither of these is in cultivation. This species is a doubtful native of England, and was considered by Watson to be only a denizen. Bromfield2 says that the broad-leaved lime, though partly naturalised in hedgerows, is nowhere indigenous in this country. Ley,3 in 1889, however, considered it to be truly wild in rocky woods in the lower valley of the Wye, where, on the Great Doward and at Symonds Yat, it grows on bare limestone rock in company with the small-leaved lime. The occurrence here of Pyms latifolia as an indigenous tree supports Ley's opinion. Baker* also considers it to be a native of Yorkshire, where it grows on the limestone scars of the lower part of Swaledale in a rocky wood, where no trees have ever been planted. Lintons also records it as growing wild on limestone cliffs in Derbyshire. (A. H.) REMARKABLE TREES One of the oldest large-leaved limes in England is the famous tree" planted by Queen Elizabeth during her visit to Burghley Park, Stamford. This is now only about 60 ft. high, having lost many limbs in recent years, but it is 20 ft. in girth, and still bears foliage freely. The tallest trees of this species which I have seen are those on the hill in the park at Longleat, where there are many from 120 to 130 ft. high, and some probably 1 Elwes found a lime wild in the Forêt de Sainte Baume, near Aubagne (Var), which was identified by M. Macler of Nice with T. platyfhyllos. It is recorded for this station by Albert and Jahandiez, Plant. Vase, du Var, 84 (1908). Enormous trees of this species are said to have existed in France, one at Château Chaulé near Meile (Poitou) having measured 50 ft. in girth in 1804, when it was 538 years old. T. Hartig alludes to another at Saint Bonnet which was 55 ft. in girth. Cf. Kanngieser, in Flora, xcix. 428 (1909). Willkomm, Forstliche Flora, 736, note (1887), mentions also large limes in Germany, one at Staffelstein in Bavaria being 57 ft. in girth at three feet from the ground. We have not been able to confirm these records. 2 f;of.a Vect_ g3 (1856). 3 Flora Herefordshire, 54 (1889). The late Rev. Augustin I.ey, who kindly sent me specimens for examination, in- formed me in a letter, that "it occurs sparingly in aboriginal woodland, through Herefordshire, where there are nine stations in which the tree is native. It extends northwards into Shropshire, westwards into Radnor and Brecon, and southwards along the Wye valley into west Gloucester and Monmouth. In many of its stations it occupies crannies of limestone cliff, where it is physically impossible that it should be planted." * Flora North Yorkshire, 274 (1906). 6 Flora Derbyshire, 91 (1903). c Figured in Card. Chron. xvi. 400, fig. 78 (iSSl). 1664 The Trees of Great Britain and Ireland taller, all running up to a great height with clean straight stems. These are reputed to have been planted in 1690. At Revesby Abbey there is a fine old tree, about loo ft. by 13 ft. 3 in., with the branches descending to the ground. The avenue at Poltimore, which is very fine, is composed of limes J of the species. Mr. Renwick reports a large-leaved lime2 at Ancrum, near Roxburgh, which in 1909 measured 26 ft. in girth at 6 ft. from the ground. (H. J. E.) TILIA VULGARIS, COMMON LIME Tilia vulgaris, Hayne, Arzn. iii. 47 (1813) ; Sargent, in Garden and Forest, ii. 256, fig. no (1889). Tilia intermedia, De Candolle, Prod. i. 513 (1824); Mathieu, Flore Forestière, 32 (1897). Tilia europcea, Linnaeus, Sp. PI. 514 (1753) (in part); Loudon, Are. et Frut. Brit. i. 364 (1838). Tilia cordata xplatyphyllos, Schneider, Laubholzkunde, ii. 374 (1909); V. Engler, Monog. Gatt. Tilia, 144 (1909). A tree, attaining 130 ft. in height and 15 ft. in girth. Bark similar to that of T. platyphyllos. Young branchlets green, glabrous, becoming dark-brown with age. Leaves (Plate 407, fig. 4) larger than those of T. cordata, averaging 4 in. in length and 3 in. in width, slightly wrinkled or uneven, cuspidate or acuminate at the apex, truncate or cordate at the base ; margin slightly ciliate and regularly serrate, the teeth ending in short points ; upper surface dark green, glabrous ; lower surface pale green, with brown axil-tufts at the base and the junctions of the midrib, primary and secondary nerves, and a few scattered long hairs on the nerves, elsewhere usually glabrous ; tertiary nerves on the lower surface prominent, mostly straight and parallel ; petiole green, glabrous, about half the length of the blade. Cymes pendulous, five- to ten-flowered, glabrous ; bract slightly pubescent, sessile or stalked ; sepals, petals, stamens, ovary, and style as in T. platyphyllos. Fruit ovoid or globose, apiculate at the apex, not ribbed when mature, covered with a dense tomentum ; shell thick and tough. The buds are similar to those of T. platyphyllos, showing three external scales, which are glabrous, shining, and ciliate ; but the glabrous branchlets will readily distinguish the common lime in winter, those of T. platyphyllos always being more or less pubescent. This species, though the most common lime in cultivation, both in Britain and on the Continent, is extremely rare in the wild state. Mathieu says that it is occasionally seen in woods3 in France; and Simonkai records reputed wild speci mens from Upsala in Sweden and from Finland. Bolle informed Sargent that he had only once seen an indigenous specimen, a tree growing in the Tyrol. 1 Cf. p. 1667, note I, for other avenues of this species, and p. 1669. 2 Cf. Christison, in Trans. Bot. Soc. Edin. xix. 494 (1893), who states that it was 20 ft. in girth at five feet up, the narrowest point, in 1877. 3 In the Cambridge herbarium there is a specimen gathered by Vincent in 1847 in the wood of ChampigneuiUe, near Nancy ; and another gathered in Switzerland by J. Stuart Mill, labelled " Mountain side, near Altdorf, Canton Uri, completely wild and native." Tilia ,665 It is now universally admitted to be a hybrid between T. cordata and T. platyphyllos. The distinctive marks of these species and the hybrid are :— T. platyphyllos. Branchlets and leaves very pubescent with long hairs. Buds with three external scales. Cymes pendulous, usually three - flowered. Fruit with prominent ribs ; shell woody and hard. T. cordata. Branchlets glabrous or nearly so. Leaves small, glabrous except for axil-tufts, bluish beneath with irregular and not prominent tertiary venation. Buds with two external scales. Cymes erect, five- to seven- flowered. Fruit faintly ridged ; shell thin and fragile. T. vulgaris. Branchlets quite glabrous. Leaves larger than those of T. cordata ; under surface pale green, glabrous except for axil-tufts and a few hairs on the nerves, with parallel straight and prominent tertiary venation (as in T. platyphyllos). Buds with three external scales. Cymesßendzi/ous, five- to ten-flowered. Fruit faintly ribbed ; shell woody and hard. There are at least two distinct forms of the common lime in cultivation in England and elsewhere which require further study, one with leaves light green beneath, longer than broad, the form described above and considered by botanists to be typical T. vulgaris, Hayne ; and the following :— 1. T.pallida,1 Wierzbicki, in Reichenbach, Icon. Fl. Germ. vi. 58, t. 315 (1844). Leaves smaller, often not much larger than those of T. cordata, as broad as or broader than long, yellowish or bluish green beneath. It is readily dis tinguishable from T. cordata by its prominent tertiary venation, and has flowers and fruits like those of typical T. vulgaris. According to V. Engler, it is occasion ally found in the wild state in Hungary. Rarer hybrids also occur :— 2. T. flavescens and T. floribunda, A. Braun, in Doell, Rhein Fl. 672 (1843)- These peculiar trees, possibly hybrids of the same parentage as T. vulgaris, were noticed growing in an avenue at Carlsruhe in 1836. The leaves closely resemble those of T. cordata, but are larger and with paler axil-tufts. The cymes, with numerous flowers, in which staminodes2 are developed, resemble in these respects those of T.japonica, the small-leaved lime of Japan. According to Koch,3 the seed was sown, and produced pure T. cordata seedlings ; but two trees at Kew, labelled T. flavescens, presumably seedlings, have larger leaves than those of the common small-leaved lime, and are peculiar in their yellow branchlets and petioles. These, though young trees,4 bear flowers, few in the cyme, without staminodes, in partly erect and partly pendulous cymes. One obtained from Späth in 1900 is about 20 ft. high ; the other, from Simon-Louis in 1902, is about 15 ft. 1 Identified by Schneider, with T. subparuifolia, Borbas, in Oest. Bot. Zeit, xxxvii. 297 (1887). T. vulgaris, var. pallida, Sargent, Bull. Pop. Inform. No. 30 (1912), and in Card. Chron. Iii. 88 (1912), is the typical form of T. vulgatis, and not the tree described by Wierzbicki. 2 On account of the staminodes, these trees are often supposed to be hybrids of T. cordata with T. aniericana, but they show no resemblance to the latter species in the shape, size, or serrations of the leaves. 3 Dendrologie, i. 481 (1869). 4 Sargent, Bull. Pop. Inform. No. 30, 1912, and in Card. Chron. Iii. 87 (1912), says that plants only a few feet high flower profusely. 1666 The Trees of Great Britain and Ireland 3. T. Beaumontia? which is sold in Späth's and Simon-Louis's nurseries, appears to be a form of the common lime, with pendulous branches. The common lime is not indigenous in Britain, where it is never found except in plantations, avenues, and hedgerows, and rarely2 produces natural seedlings. It comes into flower3 about ten days later than T. platyphyllos, and a fortnight earlier than T. cordata. In the common lime and allied species, the upper surface of the leaves is fre quently found in summer to be sprinkled over with a viscid saccharine fluid, which is popularly known as honey dew. There has been great diversity of opinion as to whether this honey dew is always an exudation from the leaves, or is in some cases voided by aphides on the leaves. Sorauer,4 the latest investigator of this subject, believes that the saccharine excretion originates without the assistance of aphides, and is the result of excessive transpiration, brought about usually by intense sunlight, a common occasion being when a cold night is followed by a hot morning sun. After the honey dew dries and thickens, it becomes the seat of growth of certain fungi, species of Fiimago, which give the leaves a blackened appearance. Paths and garden seats situated under lime trees frequently show a disagreeable coating of this viscid exudation, which has fallen from the leaves. The date of its introduction into England is uncertain, but this tree appears 5 to have been first planted on a large scale by Le Notre,6 in the reign of Charles II., who used it for avenues, as was then the custom in France. The lime trees mentioned by Turner in 1562 as attaining a large size, and the old trees reported by Barrington7 to be growing in 1769 in Moor Park in Hertfordshire and on the river Neath in Glamorganshire, were probably T. cordata, and of indigenous origin. (A. H.) CULTIVATION The lime seems to ripen its seed more often than is generally supposed in warm summers in the south of England, and I have raised seedlings8 from seeds gathered as far north as near Newark in 1904. In the same year Mr. A. C. Forbes sent me some of the common lime from Longleat, saying that very little, if any, of the seed of 1 Cf. Schneider, Laubhohkunde, ii. 374 (1909), who considers it to be a hybrid between T. euchlora and T. platyphyllos. 2 Mr. Anderson has found a few seedlings from trees planted on the edge of Lord Bathurst's deer park, just opposite the Royal Agricultural College at Cirencester.—H. J. E. 3 Cf. Dr. Moss, in Bot. Exchg. Club Rep. 1910, p. 550. 4 Pflanzenkrank, i. 412-414 (1909). The literature about honey dew on the lime is extensive. Boussingaull's article in Comptes Rendus, Ixxiv. 87 (1872), and Riviere's and Roze's articles in Bull. Soc. Bot. France, xiv. 12, 15 (1867), are abstracted in Gard. Chron. 1872, pp. 509, 609. See also various letters in Gard. Chron. 1873, pp. 920, 952, 1308, 1340, 1372, 1404, 1501, 1602. Buckton, British Aphides, \. 39-47 (1876), may also be consulted. 6 Cf. London, Arb. et Frut. Brit. i. 23, 24 (1838), who states, quoting Hasted, Kent, 562 (1769), that Sir John Speilman, in the reign of Elizabeth, brought over two lime trees from Germany, which were planted at Portbridge, near Dartford. These trees were cut down some time previous to 1769, and there is no means of determining what species of lime they belonged to. 8 According to Chalmers, Biog. Diet, xxiii. 251 (1815), Andrew Le Notre, who was born in 1613 and died in 1700, laid out St. James's and Greenwich Parks in the reign of Charles II. 7 In Phil. Trans, lix. 35 (1769). 8 In the west court of the University Library, Cambridge, which is laid out in grass, secluded, and surrounded by high buildings, there were in 1912 several natural seedlings, arising from seed brought by winds or birds. These included Betula pubescens and B. verrucosa, elder, Salix Caprca, Cralagus monogyna, sycamore, and a solitary seedling, three or four years old, which was apparently T. vulgaris. It differed slightly in having slight pubescence on the branchlets and under surface of the leaves, thus showing a reversion to T. platyphyllos.—A. II. Tilia 1667 the small-leaved lime was fertile, possibly because it flowers later than the other. A small proportion of this seed germinated in the first spring, but most of it lay dormant till the following year, and this has been my experience with many sowings of seed from abroad of the various European forms, as well as of the American lime. After germination they grow slowly for the first two years, and the young wood is liable to be killed back in winter. For this reason layering is the method adopted by nurserymen, though the varieties are usually grafted. I cannot say that my experience is as yet long enough to justify me in preferring seedlings.1 With regard to soil, the lime is not particular, but requires a good deep loam to bring it to perfection. It transplants very well, and may, if properly prepared by cutting round the roots two years previously, be safely moved when 20 to 30 ft. high. Owing to the depreciated value of the timber, the lime cannot now be recom mended except as an ornamental tree, the principal objection to it for this purpose being the early period at which its leaves wither and fall in autumn. LIME AVENUES The common lime is one of the most valuable avenue trees that we have ; the fashion for planting them is, however, not very ancient, having apparently been introduced by Le Notre and other French landscape gardeners in the latter half of the seventeenth century, from which period most of our best avenues date. Of these one of the finest is the avenue at Burghley Park, Stamford, the seat of the Marquess of Exeter. This is about 3000 yards long, with four rows of trees planted 6 yards apart in the row, 10 yards between the two outer rows and 20 yards between the inner ones. The trees are 100 ft. to 110 ft. in height on an average, and all appear to have been pollarded when young, though they have the upright habit which distinguishes most of the older lime avenues. I was informed by Mr. C. Richardson, of Stamford, that about fifty-five years ago—when lime wood was much more valuable than it now is, and made 55. to 6s. per foot, single trees being sometimes sold at ^40 to ^50—an offer was made by a syndicate of London timber merchants to buy the whole of this avenue for .£100,000. This story appears hardly credible, and I could obtain no verification ; but, if made, the offer was refused, and there is no chance of such a price being paid for lime trees now. Another beautiful avenue of fine tall limes is at Stratton Park, Hants, the seat of the Earl of Northbrook, whose late father informed me that it probably dates from about 1715. This avenue shows a common defect, which consists in the mass of spray that springs from some point usually near the root, though sometimes at 10 to 20 ft. up the trees, or even higher. I have searched in vain the works of Evelyn, Duhamel, Miller, Boutcher, and Loudon, for any reference to these abnormal growths, 1 T. platyphyllos is usually imported from France as seedlings ; and these appear to thrive in some cases better than plants of the common lime, which have been raised from layers. A young avenue of T. platyphyllos at Terling, Essex, with trees about 30 ft. high, is very thriving. The lime avenue at the back of Trinity College, Cambridge, which was celebrated by Tennyson, consists of two parts. That on the west side of the Cam now consists of 38 common limes, one half of which are very burry and much decayed, gaps showing where a few have died. These trees, which now average 6 ft. 9 in. in girth, were planted in 1671, at a cost of £10 : 6s., plus carriage from London amounting to £i : 4s. On the east side of the Cam there arc 20 trees, all but one of which are T. platyphyllos. These, which now average 6 ft. in girth, were planted in 1717, and look much healthier than the others. Cf. Willis and Ctarke, Archil. Hist. Univ. Camb. ii. 641, 646 (1886).—A. II. 1668 The Trees of Great Britain and Ireland which, so far as I can learn, seldom appear on wild trees, or on any species but T. vulgaris. It is scarcely due to soil, since in some of the finest old lime avenues, as those at Cassiobury Park, Waldershare Park, and Newhouse Park, these growths appear in some trees only. They extend to a considerable height up some of the trees, which are much stunted, as it seems, from this cause. This may arise from the affected trees having been propagated by layers from inferior shoots, or having been planted later to fill gaps, and thus having to contend with trees already established. It is, however, a point which deserves careful attention on the part of nursery men ; as, though these growths may be pruned off annually, they constantly reappear at the same spot, and not only take a great deal of trouble and time to remove, but eventually disfigure the trees ; and limes which produce them rarely attain the same height or beauty as those which are free from them. The trees now sold by nursery men, which are always propagated from layers, seem to be more subject to these growths than the limes planted two centuries ago ; and I believe that limes raised from seed are rarely if ever affected. The longest avenue of limes which I have seen is comparatively modern, and as I am informed by Mr. A. H. Elliott, agent for the Clumber estate, was planted by Henry, fourth Duke of Newcastle, about the year 1840. It is i mile and 1590 yards long, and consists of 1315 trees planted in a double row on each side of the drive at Clumber. The trees are 31 ft. apart each way, and the total width is 143 ft. The trees are fairly uniform in habit, but have spreading bushy tops, and when I saw them in 1906 did not exceed about 60 ft. in height by about 4^ ft. in girth. The soil is rather sandy, and the trees when planted were only 5 ft. high, and were not pruned sufficiently after planting to develop a good trunk, so that this avenue is never likely to rival those at Burghley, Ashridge, or Cassiobury. Mr. Elliott tells me that the trees have suffered considerably from the attacks of the following geometrid moths : Clteimatobia brumata, C. boreata, Hybernia aurantiaria, H. defoliana, H. progemmaria, Anisopteryx œscularia ; but this damage has been checked, if not entirely prevented, by putting grease bands on the trees, which arrest the female moths when they try to ascend the stems in the winter months, and by killing the pupae in the soil in July with gas-lime. Besides this avenue there are at Clumber two much older ones, over 150 years old, running north and south on either side of the elm avenue leading to West Drayton. One of these is 385 yards long and 30 yards wide, the other 330 yards long and 55 ft. wide. The trees are planted 24 ft. apart. They were pollarded in 1888 in order to save their lives. At Newhouse Park, near Mamhead, Devonshire, Sir Robert Newman showed me a fine avenue which seems to have been planted about 200 years ago in anticipation of a mansion which was never built. It is only 20 ft. wide and the trees 10 ft. apart ; but favoured by a fine soil and climate, the trees, which seem to have been pollarded at 10 ft., have shot up to an immense height, averaging at least 120 ft., and several exceeding 130 ft. Two which I measured were 115 ft. by 5 ft. 9 in., and 135 ft. by 7 ft. 6 in. I have little doubt that they were seedlings. In a chestnut avenue at the same place the trees were much shorter and thicker, about 70 ft. to 80 ft. by 15 ft. to 18 ft. in girth. Tilia 1669 At Cassiobury Park, Herts, there is a lime avenue supposed to have been planted by Le Notre, but some of the trees, which are very inferior to the rest in height and symmetry, are smothered in dense masses of small spray at 20 ft. to 30 ft. from the ground, and seem to have been planted later, possibly to replace dead trees in the original avenue. This avenue is 24 yards wide and the trees 8 yards apart. The best of the trees are 120 ft. to 130 ft. high, and one measures 13^ ft. in girth. There is a fine avenue, about half a mile long, at Denham Court, near Uxbridge. At Braxted Park, Essex, the property of C. H. Du Cane, Esq., there is a lime avenue composed of three rows of trees on each side, which shows extraordinary variation in the growth of the trees, giving it a very irregular appearance. The tallest at the bottom of the hill are about 120 ft. high, and covered to an extent I have never seen elsewhere, with mistletoe growing in large bushes nearly up to their tops. At the top of the hill near the entrance gate, many of the trees are poor and stunted, with masses of spray at their root and higher up, and with gouty swellings on their branches which may be due to the mistletoe. At Betchford Park, Surrey, there is a remarkable avenue of very old trees, some of which, when seen by Henry in 1906, were 130 ft. in height and 12 ft. to 13^ ft. in girth. This avenue was described as a very fine one by Dr. Aikin1 in 1798. At Doneraile Court, in Ireland, there is a fine avenue, one tree measuring 98 ft. by IG ft. in 1907 ; most of the trees were covered with masses of spray. An excellent article by Alfred Rehder of the Arnold Arboretum, on the lime as an avenue tree, in Möllers Deutsche Gärtner Zeitung, 1904, p. 188, should be consulted by those who think of planting. After giving the distinctive charac ters of the species, and describing their peculiarities of growth, he says that the choice must depend on the character of the soil and climate, and considers that the large-leaved lime, T. platyphyllos, is the best where the ground is deep and moist and where rapid growth and heavy shade are required. For drier soil he prefers T. cordata, which does not, however, make such a large or fine tree. T. vulgaris is in most of its characters intermediate, and this is the lime which is most generally used in England, though according to Rehder most, if not all, of the limes celebrated for their size and age in Germany are T. platyphyllos. He thinks that T. euchlora, Koch, is the best for town planting, because its smooth leaves do not hold the dust so much as those of other species, and because its leaves do not fall so early. He prefers T. petiolaris, the pendulous silver lime, for park avenues, and for single specimens where its branches can show their full beauty. Rehder does not give any observations as to the relative advantage of trees propagated from seed, from layers, or from grafts ; but he rightly says that it is important that all the trees should be propagated from the same variety. REMARKABLE TREES Among the most remarkable limes that I have seen is a walk at Ashridge Park. These trees are individually much larger than those at Burghley or Stratton, and 1 In Monthly Magazine for 1798, quoted in Gard. Chron. 1841, p. 4. VU D 1670 The Trees of Great Britain and Ireland though planted only 4 to 5 yards apart, average 120 ft. high and about 10 ft. in girth. They have the amount of variation in their leaves that one would expect to find in seedlings, and though 250 years old1 only one out of forty is decayed. The bark of these trees is more like that of an elm than the usual bark of a lime, but this is perhaps owing to some peculiarity in the soil. A great lime at the end of the pinetum at Ashridge is of a totally different character, having very drooping branches and an immense spread. It has a trunk about 30 ft. high, and does not exceed 80 ft. to 90 ft. in total height, but the branches cover a circumference of no paces. In Windsor Park, near Cranbourne Tower, there are some extremely tall and graceful limes growing with beech in a circle which were planted2 in 1697. The best of these that I measured was 130 ft. by 14 ft. in girth, a beech close by it being 125 ft. by 10 ft. A remarkable case of the tendency of the lime to layer which occurs at Rotham- sted, is figured in the Gardeners' Chronicle, June 5, 1875. Here a row of fine old limes have dropped their branches on each side to the ground, and these have grown up in a thick mass, forming a shady corridor on each side of the trunks. There is another good example at Enville Hall, Stourbridge, where hundreds of young stems have arisen from layers, the whole mass measuring 140 paces round in 1904. A curious instance of natural inarching of the lime is described and figured in Gard. Chron. xi. 277 (1879). The branches of the lime sometimes spread laterally to a great distance. One of the best instances I have seen was shown me by Mr. Tudway at the Coombe, near Wells in Somersetshire. The tree is a large-leaved lime growing in a sheltered dell, about loo ft. by 14 ft., and has three immense horizontal limbs 8 to 9 ft. in girth, one of which extends for 64 ft. from the trunk. In Stoneleigh Park, Warwickshire, there are some fine limes remarkable for their wide-spreading branches. In the Thames valley there are many fine lime trees, one at Osterley Park being about 120 ft. in height ; while at Crowsley Park, near Henley, there are several in a clump, one of which was 118 ft. by 12^ ft. in 1908. E. Lees, in his account3 of the Forest and Chace of Malvern, speaking of the common lime, says :—" Some very fine trees now stand in a field about half a mile south of Bromsberrow Church, and by the side of the road leading from Ledbury towards Gloucester. Two of these, growing near each other, have become conjoined, both by the amalgamation of their arms, and by a lateral junction at the root.4 The largest of these trees is 27 ft. in circumference at three feet from the ground, and is 36 ft. round the base ; the other is 11 ft. 3 in. in girth at a yard from the ground, and 19 ft. in circumference at the base. The whole mass, if measured as one tree (and the interval between the boles where the connecting root joins them is only 19 in.) is full 48 ft. in circumference. In a field on the Priory Farm, Little Malvern, are several large 1 In a. book called Chiltern and Vale Farming, p. 153, published in 1745, it is stated by the anonymous author, who lived close to Ashridge, that they were planted in 1660, and in 1745 or thereabouts were near 3 ft- in diameter at the bottom. 2 W. Menzies, History of Windsor Great Park, 44 (1864). 3 Abstracted in Gard. Chron. 1870, p. 1536, figs. 264, 265, 266. 4 I visited these trees in 1905, and found the largest now standing to be about 80 ft. high and 20 ft. 9 in. in girth. The fruit was fully formed on 18th July. Tilia 1671 trees of Tilia platyphyllos, but these do not belong exactly to forest trees, having certainly been planted either by one of the Priors of Little Malvern or some of his lay successors to the Priory lands." A very large lime formerly grew in Hagley Park, Worcestershire, the seat of Lord Cobham, which, according to Lees,1 in 1874 measured 27 ft. in girth at 3 ft. from the ground, but this I am told was blown down about twenty-five years ago. At Arley Castle there is a fine tree, about 120 ft. high by 12 ft. 8 in. in girth. The most curious instance of artificial layering I have seen anywhere is at Knole Park, where a lime described by Loudon as having covered nearly a quarter of an acre in 1820 still grows. The central stem no doubt originally dropped its branches on the ground in a circle of about 8 yards in diameter. These have grown up into trees 80 ft. to 90 ft. high, some of which are thicker than their parent. These again have layered themselves in a second concentric circle 20 yards in diameter, the trees in which are 20 ft. to 40 ft. high, and these are now rooting their outer branches in a third circle more irregular than the others, and 8 yards distant from it, so that the total diameter of the group is 36 yards. All the stems are more or less covered with spray, and the central one seems to have long ceased to grow. Strutt paid less attention to the lime than it deserves. He figures only two, one of which, at Cobham, is in the same plate with a sycamore ; the other at Moor Park, near Ricksmansworth, Herts. This stood at the end of a line of large limes, and was a very wide-spreading tree, with a trunk 17^ ft. at three ft. from the ground, and branches 120 yards in circumference. I am informed by Mr. Haynes, gardener at Moor Park, that this tree was blown down in 1860; but it is still alive, and some of the branches have taken root in the ground and have sent up stems about forty-five feet high. Within fifty yards of it another tree in the same row is now 21 ft. in girth. A similar case of a lime having been blown down and the branches taking root occurs at Stratton Strawless. This tree, as Mr. Birkbeck tells me, was mentioned by Sir T. Browne in the reign of Charles I. It was blown down in 1895 and lay till 1900, when the roots were covered with a mound of sand. When I saw it in 1909, many branches were throwing up vigorous shoots, and the tree looked as if it might live for centuries. Its trunk was about 12 ft. in girth. A very fine tall red-twigged lime by the water at Gatton Park, in 1904, was 131 ft. by 12^ ft. A large spreading tree, at Osberton Grange, Notts, is about 80 ft. high by 19 ft. in girth. At Dallam Tower, Westmoreland, there is an old lime in an exposed situation in the park which measures no less than 22 ft. 3 in. in girth, though not over 65 ft. in height ; and, as showing the influence of situation on trees, I may say that, in a sheltered hollow close to the house at the same place, I measured a lime 128 ft. high, double the height of the first, but only 7 ft. 8 in. in girth. There is a row of very large and apparently old limes at Hawsted, near Bury St. Edmunds, in the same field, and probably planted at the same time as the Oriental planes which we have described.2 The largest is about 105 ft. by 20 ft. Another, whose trunk is covered with large burrs, is 16 ft. in girth. The leaves on the shoots from the base of these trees vary considerably in size and shape. 1 Card. Chron. i. 49 (1874). * Cf. Vol. III. pp. 621, 622. 1672 The Trees of Great Britain and Ireland In the west garden at Hatfield House, Herts, there is a remarkable pergola of the common lime, about 280 yards long, and 10 ft. wide. The trees, which are pruned every year, stand 12 ft. apart in the row on each side, and are 7^ ft. high. In some parts of England, especially in Essex and Herts, the lime is infested by mistletoe, which often kills the branches and causes irregular excrescences, which sometimes have an elongated gourd-like shape. A remarkable specimen of this, taken from a tree in front of Spains Hall, the seat of A. W. Ruggles-Brise, Esq.," in Essex, is now in the Forestry Museum at Cambridge. It was stated in the Gardeners Chronicle? in answer to an inquiry by me, that such swellings are abundant on the limes at Hampton Court and at Dropmore, where Mr. Page states that a large tree was cut down on account of it being in a dying condition. In Scotland 2 the lime is at many places almost as fine as in England. An immense tree growing at Kinloch, Meigle, is, as I am told by Sir John Kinloch, about 90 ft. by 21 ft., and spreads over an area of a hundred yards in circumference. A wide-spreading lime at Gordon Castle, which is known as the Duchess' tree, measures about 89 ft. high and 17 ft. 4 in. in girth ; its layered branches form a dense mass of shoots and have not been trained into trees like the lime at Knole. Their total circumference is not less than 126 paces, so that it covers as large an area of ground as the Newbattle beech.8 This tree is mentioned in Old and Remarkable Trees as having been, in 1867, 70 ft. by \£>\ ft. at 3 ft., the circumference of the branches being 310 ft. There is a wide-spreading common lime at Pitfirrane, near Dunfermline, of which the gardener, Mr. Percy Brown, has sent us a photograph. It was, in 1912, 74 ft. high, and 13 ft. 3 in. in girth, the circumference around the branches being 298 ft. At Leny there are some fine limes, one of which I found to be 105 ft. by 12 ft. 3 in., which, at such a high elevation above the sea, is remarkable. At Ancrum, near Roxburgh, Mr. Renwick saw a common lime 17 ft. 3 in. in girth in 1909. At Roseneath, near the great silver firs (see p. 729), is an old avenue of large- leaved limes, covered with such a mass of small spray that it was impossible to see the bark near the ground, and one of these, measured over the spray with the tape as tight as I could make it, was no less than 24 ft. There are some fine lime trees of great size in the park at Taymouth Castle. In Ireland we have seen no limes of remarkable size, and the tree never seems to have been so generally planted as in England. There is an avenue of fair-sized trees at Muckross Abbey ; and Loudon mentions a tree in the park at Charleville forest, Co. Meath, which was reported at that time to be 110 ft. high and 5^ ft. in diameter at I ft. from the ground. At Rossanagh a remarkable tree was growing in 1908 which, as I was informed by Mr. W. T. Tighe, was blown down about 1825. His grandfather had it pulled up into a leaning position, and placed a large boulder over the roots to keep it firm. It now leans at an angle of about 40°, and has grown into a flat trunk 6 ft. wide on the side but only 14 ft. 10 in. in girth. It is about loo ft. in height and twice as large in girth as any other of the trees in the same line, which appear to have been planted at the same time. 1 Card. Chron. xli. 240, 257 (1907). * Cf. Christison, in Trans. Bot. Soc. Edin. xix. 494 (1893). 3 Vol. I. p. 23, Plates 8, 9. Tilia 1673 TIMBER The wood of the lime is pale yellow or white, light, soft, and close-grained, and is not liable to become worm-eaten. It was formerly valued by pianoforte-makers for sounding-boards ; and cutting boards used by shoemakers, glovers, and harness- makers, were made of it. I am informed by Mr. Anderson that twenty-five years ago he was sometimes able to sell the best part of large trees at as much as 55. or 6s. per foot ; and at Longleat a large lime tree blown down in the park realised 45. 6d. per foot ; but its use has now been superseded by foreign imports from America and elsewhere, and from one to two shillings per foot is its usual value. Owing to its softness, consistency, and non-liability to split, it was preferred for wood- carving ; and all the finest carvings by Grinling Gibbons are said to be done in lime wood.1 In northern Europe, especially in Russia, the inner bark of the small-leaved and large-leaved limes is largely used for making the bast mats which are used as dunnage in grain cargoes, and also imported for covering garden frames. The shoes worn by the Russian peasants are made from plaited lime bark, and Loudon says that ropes were made from it in his time in Devonshire and Cornwall ; but this, like so many other rural industries, has now, I believe, quite died out. I am indebted to Mr. J. Rose, of Messrs. Broadwood & Sons, for the following :—" Fine lime-tree was at one time very eagerly sought after in this country for the manufacture of pianoforte keys. When large and freely grown it is a beauti fully straight and silky-grained wood, easily worked, not given to warping, very light in weight, and yet very tough. These qualities made it an admirable material for the purpose. But it became more and more difficult to obtain lime-tree of fine quality, and it was replaced by the importation of American basswood, a wood of similar character, easier to obtain in good sizes, free from knots, and straight in the grain, which is imported in the form of boards, or of glued-up and planed keyboards ready for the ivories. It has also been replaced, in part, for key-making by continental- grown pine, which has distinct advantages for the purpose." A marked feature in the timber trade in recent years has been the importation of sawn timber, which has greatly affected the sale of home-grown timber. The manufacturer is now supplied with foreign timber ready sawn, seasoned to some extent, and often carefully graded ; whereas home-grown timber has to be collected in comparatively small parcels, and its selection and handling require a great amount of knowledge and experience possessed by a very few persons. (H. J. E.) 1 Evelyn (Si/va, Hunter's ed. i. 205 (1801)), says : " Because of its colour and easy working, and that it is not subject to split, architects make with it models for their designed buildings ; and the carvers in wood use it not only for small figures, but for large statues and entire histories in bass and high relieve ; witness beside several more the festoons, fruitages, and other sculptures of admirable invention and performance, to be seen about the choir of St. Paul's and other churches, royal palaces, and noble houses in city and country ; all of them the works and invention of our Lysippus, Mr. Gibbons, comparable, and for ought appears equal to anything of the antients. Having had the honour (for so I account it) to be the first who recommended this great artist to his Majesty Charles II., I mention it on this occasion with much satisfaction." 1674 The Trees of Great Britain and Ireland TILIA EUCHLORA Tilia euchlora, Koch, in Wochenschr. Gärtn. u. Pßanzenk.K. 284(1866), and Dendrologie, i. 473 (1869); Schneider, Lautholzkunde, ii. 374 (1909); V. Engler, Monog. Gatt. Tilia, 149 (1909). Tilia multiflora, Simonkai, in Math. Term. Közl. xxii. 328 (1888) (not Ledebour). Tilia rubra, var. euchlora, Dippel, Laubholzkunde, iii. 63 (1893). Tilia dasystyla, Jack, in Garden and Forest, i. 332 (1888); Nicholson, in Kew Hand-list Trees, 45 (1894) (not Steven).1 A tree, attaining 50 ft. in height and 6 ft. in girth, but possibly larger in its native country. Bark grey and scaly. Young branchlets green, glabrous, the short shoots, however, being slightly pubescent. Leaves (Plate 407, Fig. 10) intermediate in size between those of T. cordata and T. vulgaris, averaging z\ in. in width and length, orbicular-ovate, coriaceous, cuspidate at the apex, oblique and cordate at the base ; upper surface dark shining green, glabrous ; lower surface paler, almost glaucous, glabrous, except for brownish axil-tufts at the base and at the junctions of the midrib, primary and secondary nerves ; margin regularly serrate, the teeth ending in long slender points ; petiole glabrous, slender, more than half the length of the blade. Buds, with three external green glabrous ciliate scales. Cymes glabrous, pendulous, exceeding the leaves in length, three- to seven- flowered ; bract glabrous, tapering at both ends, shortly stalked ; flowers similar to those of T. platyphyllos, but the ovary is tomentose with long hairs, which are continued on the base of the style, the upper f of which is glabrous. Fruit ovoid, indistinctly five-ribbed, covered with dense short brownish grey tomentum, the base of the style persistent at the apex ; shell thick and woody. This species is remarkably distinct in appearance, owing to the dark green and remarkably glossy upper surface of the leaves. It comes into flower about the end of July at Kew, later in the season than most of the limes, and the flowers have a peculiar colour, owing to the distinctly yellow tinge of the petals and filaments. It has been much confused with T. dasystyla? Steven, a native of the Crimea, which has leaves quite different in colour and shape, and, as its name indicates, a densely tomentose long style. Tilia euchlora is represented in the Kew herbarium by a wild specimen from Karabagh in Russian Armenia ; but it is supposed by Schneider and V. Engler to be a hybrid ; and, if this is the case, its parents are possibly T. caucasica? Ruprecht, the common large-leaved lime in the Caucasus, and T. cordata. It is always propagated in nurseries by budding on the common lime ; and seedlings of it appear to be unknown. 1 Loudon, Arb. et Frut. Brit. i. 366 (1838), refers to Steven's species from the Crimea, and not to the tree called by us T. euchlora, Koch. 2 T. dasystyla is represented in the Kew herbarium by Steven's type specimen and two other specimens, all from the Crimea. 3 The Caucasian lime can scarcely be identified with T. rubra, De Candolle, in Cat. Plant. Hort. Monsf. 150 (1813), which is described as having leaves pubescent beneath as in T. platyphyllos, and evidently refers to the southern form of the latter species in Europe. Tilia 1675 This species is perhaps the handsomest of all the limes, on account of its shining foliage, which is very late in falling in autumn, and seems to be free from insect and fungoid attacks and from honey dew. It is apparently quite as hardy as the common lime, and young trees at Kew are remarkably thriving and healthy. It was unknown in Loudon's time, and seems to have been introduced a short time before 1866, when it was first accurately distinguished by Koch. It is rather rare in cultivation in England, though it is planted in Berlin and other German cities,1 and thrives2 in the Arnold Arboretum in Massachusetts. (A. H.) TILIA TOMENTOSA, WHITE LIME Tilia tomentosa, Moench, Verz. Ausl. Bäume Weissenst. 136 (1785); V. Engler, Monog. Gatt. Tilia, 116 (1909); Schneider, Laubholzkunde, ii. 386 (1909). Tilia alba, Aiton,3 Hort. Kew. ii. 230 (1789), and iii. 300 (1811); VValdstein and Kitaibel, Icon. PL Hung. i. 2, t. 3 (1802) ; Loudon, Arb. et Frut. Brit. i. 372 (1838). Tilia pallida, Salisbury, Prod. 367 (1796). Tilia rotundifolia, Ventenat, Mem. Inst. Paris, iv. 12, t. 4 (1803). Tilia argentea, Desfontaines, in De Candolle, Cat. PL Hort. Monsf. 150 (1813). Tilia petiolaris, De Candolle, Prod. i. 514 (1824) (not J. D. Hooker). A tree, attaining 100 ft. in height and 15 ft. in girth, usually with markedly ascending branches. Young branchlets covered with white stellate tomentum, more or less retained in the second year. Leaves (Plate 407, Fig. 3), about 3 to 5 in. across, nearly orbicular, cuspidate at the apex, cordate or truncate at the base ; margin often lobulate, serrate or biserrate, the serrations ending in short blunt cartilaginous points ; upper surface green, with scattered stellate pubescence ; lower surface covered with a dense whitish tomentum, without axil-tufts ; petiole stout or slender, less than half the length of the blade, stellate-pubescent. Buds, with three external grey tomentose scales. Flowers, in seven- to ten-flowered pendulous tomentose cymes, which are shorter than the leaves ; bract tomentose, sub-sessile ; sepals tomentose, clothed with long hairs at the base within ; petals glabrous, longer than the sepals; staminodes slender, spatulate, shorter than the petals ; stamens, shorter than the staminodes, numerous, with the halves of each anther on a distinct short stalk ; ovary ovoid, tomentose ; style glabrous. Fruit ovoid, elongated, apiculate, slightly five-angled, grey tomentose, smooth or only indistinctly warty ; shell woody. This species shows in the wild state considerable variation in the shape of the 1 Mr. Bean, in Kew Ball. 1908, p. 390, says that this species, which is so promising a tree for street planting, is abundant in the Boskoop nurseries, near Gouda, in Holland. 2 Garden and Forest, i. 332 (1888). 3 In Aiton, Hort. Kein. ii. 230 (1789), it is erroneously stated that the common white lime is a native of North America, an error which was rectified in the second edition of this work, iii. 300 (1811), where Hungary is correctly given. Some writers have supposed that T. heterophylla was the species referred to ; but the type specimen in the British Museum, inscribed Tilia alba in Solander's handwriting, though bearing neither flowers nor fruit, is without doubt a branch of the common European lime, identical with var. argentea. 1676 The Trees of Great Britain and Ireland leaves—some specimens, distinguished by V. Engler, as var. typica, having suborbicular leaves, usually cordate at the base ; whilst others, var. petiolaris? V. Engler, have leaves usually broader than long, and more or less truncate or subcordate at the base. These are connected by numerous intermediate forms, and can scarcely be maintained as distinct varieties. All the wild specimens which I have seen, are characterised by leaves, thin in texture, usually pale green above, greyish and not snowy-white beneath, and with slender petioles. Trees similar to the wild form are occasionally seen in cultivation ; but the tree which is more commonly cultivated under the name T. argentea, and which possibly was the one described by Moench as T. tomentosa, and in that case technically the typical form of the species, differs considerably in having larger leaves, thick in texture, more or less orbicular, uneven on the surface, dark shining green above, snowy white beneath, margin often lobulate, petioles short and stout. It is convenient to distinguish this cultivated form, the origin of which is unknown to me, as var. argentea. It has always ascending branches, and is possibly a sport. This species is a native of south-eastern Europe, and Asia Minor ; but does not extend as far eastward as the Caucasus. Its northern limit is southern Hungary, where it is found in some parts of Banat, Slavonia, and Croatia. It is widely distributed throughout the Balkan peninsula, extending southwards as far as Laconia in Greece ; and spreads eastward through Roumania and Moldavia to Bessarabia, Podolia, and the Crimea. In Asia Minor, it is limited to Bithynia and the island of Chios. It is usually a component of mixed deciduous woods, growing in valleys and mountain slopes at a low elevation, but occasionally forms pure woods of small extent. (A. H.) CULTIVATION The white lime was introduced into England in 1767, and has been planted as an ornamental tree at many places ; but I know no avenues of it in England. This species should be multiplied either by seed, which here only ripens in hot seasons such as 1911, or by layering, as when grafted, as is often done, on the com mon lime, the scions grow thicker than the stock and produce an unsightly swelling at the point of junction. As an ornamental tree, and for use in towns, it is much superior to the common lime, on account of the freedom of the leaves from honey dew. The only place where I have seen the silver lime growing wild in Europe is in the forests of Bosnia, near Maglai in the valley of the Bosna, at about 1000 ft. elevation. Here it was scattered in forests of oak, and other deciduous trees, and was so conspicuous when the silvery white undersides of the leaves were upturned by the wind, that at a distance of a mile or so I at first supposed it to be a tree covered with white flowers.2 Probably some such experience must have induced a former owner of Highclere to plant it largely in that beautiful park, where, as Loudon 1 To be carefully distinguished from T. feticlaris, Hooker (see p. 1677). 2 In sunny weather, the leaves on the sunny side, especially at the ends of the branchlets, are reversed, turning their white sides to the light. This is a provision against excessive transpiration of water ; and has been observed in the other silver limes and in Quercus conferta. See Kerner's remarks in his Nat. Hist. Plants, Eng. Trans, i. 338 (1898).—A. H. Tilia 1677 remarks, its presence could be detected at some miles distance through the apparently dense forest by the white tops appearing at intervals among other trees. The finest tree of this species that we have seen grows at Albury.Park, Sussex, in front of the Duke of Northumberland's house, and measures 100 ft., or perhaps a little more, in height, by 13^ ft. in girth (Plate 373). The ascending branches seem characteristic of this species, in cultivation at least, but this tree has an unusually regular and perfect head. I was informed by the late Mr. Leach that at Albury the flowers of this tree are poisonous to bees,1 whose dead or stupefied bodies are found lying on the grass below it in August, and this observation is confirmed by Mr. Comber at The Hendre and by other observers. At Henham Hall, Suffolk, a tree measured 76 ft. by 8 ft. 10 in. in 1909. It is grafted on the common lime, and is of the typical upright habit. At Hewell Grange a large tree, with the bark decaying on one side, was 92 ft. by 9 ft. 8 in. in 1909. At Dropmore there are two trees in the avenue to the Taplow gate, the larger of which is about 70 ft. by n ft. 3 in. At Harpsden Rectory, Oxon., Henry saw two trees, the larger of which was 80 ft. by 6 ft. 8 in. in 1907. At Arley Castle, a round- headed tree of upright habit measured2 62 ft. by 6 ft. 9 in. in 1903. At Beauport, Sussex, there is a tall but slender tree in a rather crowded situation, which has been grafted at about 10 ft. from the ground. Dr. Masters reported3 a tree of fine propor tions and symmetry at Strathfieldsaye, 70 ft. by 6 ft. 2 in., in 1899. Mr. Bean4 saw a tree 80 ft. high in the Royal Gardens at Sans Souci, Berlin. (H. J. E.) TILIA PETIOLARIS, WEEPING WHITE LIME Tilia petiolaris, J. D. Hooker, Bot. Mag. t. 6737 (1884) (not De Candolle) ; Boissier, Ft. Orientalis, Suppl. 136 (1888). Tilia alba, Koch, Dendrologie, i. 478 (excl. syn.) (1869) (not Aiton). Tilia tomentosa, Moench, var. petiolaris, Kirchner, in Petzold and Kirchner, Arb. Muse. 162 (1864); V. Engler, in Mitt. Deut. Dend. Ges., 1907, pp. 218-221. Tilia tomentosa, Moench, var. sphœrobalana, Borbas, \x\Bot. Centrait, xxxvii. 168 (1889); V. Engler, Monog. Gatt. Tilia, 121 (1909). A tree, attaining 80 ft. in height, differing from the wild form of T. tomentosa, as follows : Branches and branchlets pendulous ; leaves with long slender petioles, exceeding half the length of the blade ; fruit globose, depressed at the summit, from which arises a short stout style, very warty on the greyish surface, and divided by five vertical furrows into as many lobes ; seeds6 often imperfect. The leaves (Plate 407, Fig. 2) average about 3 in. across, and are obliquely orbicular, cordate or truncate at the base, cuspidate at the apex, flat on the surface and not wrinkled ; margin finely and regularly serrate, the teeth ending in short points ; under surface covered with a dense white tomentum ; upper surface dark 1 The flowers of the large tree of T.petiolaiis at Kew are equally poisonous to bees. Cf. p. 1679, note I. 2 R. Woodward, Hortus Arleyensis, 25 (1907). 3 In Card. Chron. xxvi. 162 (1899). « Kew Bull. 1908, p. 395. 6 Engler, out of fifty fruits which he examined, found only three with good seed, I to 2 in each fruit. VII E 1678 The Trees of Great Britain and Ireland green, with scattered stellate pubescence. The branchlets, buds, and flowers are identical with those of T. tomentosa. This tree appears to be a sport of T. tomentosa of unknown origin. Schneider considers it to be a native of southern Hungary and the Balkan States ; and no doubt, specimens of silver limes with longer petioles than usual occur in that region, but no one, so far as I am aware, has ever seen a tree in the \vild state with the pendulous habit and the peculiar fruits of T. petiolaris, Hooker. V. Engler admits that it is only known in cultivation. The history of this tree is obscure ; but it seems to have been first accurately distinguished by Kirchner, who, in 1864, described it as T. tomentosa, var. petiolaris, and erroneously identified it with T. petiolaris, De Candolle.1 Hooker, in Bot. Mag. t. 6737 (1884), following Kirchner, adopted De Candolle's name; but, though this was incorrect, it is convenient to retain the name as T. petiolaris, Hooker. This peculiar tree was not known in Loudon's time ; but must have been introduced from the Continent soon afterwards, as the fine specimen in the Cambridge Botanic Garden was probably planted2 in 1842. Before 1864 it was known in cultivation as T. americana pendula3 ; and Koch in 1869 considered it to be of American origin. i. Tilia orbicularis, Carrière,4 which originated in Simon-Louis' nursery at Plantières, is evidently a seedling of T. petiolaris, from which it differs mainly in being less pubescent. The leaves, which are dark glossy green above and dull grey beneath, also differ from those of T. petiolaris in the shorter petioles, which are, however, slender as in that species. The serrations are also slightly sharper and occasionally more irregular. Flowers as in T. petiolaris, but with the bract larger and nearly glabrescent, and the sepals and pedicels covered with a less dense and greyish tomentum. Fruit strictly globose, not depressed at the summit, and showing no furrows, but having the same warty surface as that of T. petiolaris. This tree, which is not nearly so pendulous in habit as T. petiolaris, is reputed, on account of the dark glossy green of the upper surface of the leaves, to be a hybrid between T. petiolaris and T. ettchlora ; but, in the present state of our knowledge of hybrids, it is judicious to say nothing about its parentage until experimental sowings have been made on a large scale of the seed of T. petiolaris.5 Two small trees of T. orbicularis, obtained from Plantières in 1900, are thriving at Kew, one of which has already produced flowers and fruit. They retain their foliage late in the season. (A. H.) 1 T. petiolaris, De Candolle, was founded in 1826 on a branch without flowers or fruit, now preserved in the Geneva Herbarium, which was taken from a tree cultivated in the Imperial Botanic Garden at Odessa ; but Lange, in Flora, i. 233 (1827), who had seen this tree, states that it was identical in every respect with the ordinary form of T. argentca cultivated at Paris ; and a drawing at Kew of De Candolle's specimen confirms Lange's opinion. The name T. petiolaris, De Candolle, thus disappears, being a mere synonym of T. tomentosa. 2 The oldest herbarium specimen, which I have seen, is one in fruit from a tree growing in a street at Nancy, collected by Billot in 1861. According to Bunbury, Arb. Notes, 67 (1889), the fiue tree at Barton was planted by his father; but no exact date can be now ascertained. 3 Cf. Rehder, iu Mitt. Deiit. Dead. Ges. 1904, p. 209. * Ex Beissner, in Mitt. Deut. Dend. Ges., 1898, pp. 86 and 88. It appears, however, to have been first accurately described by Jouin in La Semaine Horticole, 1899, p. 335. 6 A branch from a seedling of T. petiolaris, raised in the Arnold Arboretum and sent to me by Prof. Sargent in 1910, bears foliage identical with that of the parent. Sargent, Bull. Pop. Inform. No. 30 (1912), and in Gard. Chron. Hi. 88 (1912), states, however, that plants raised in tlie Arnold Arboretum from the seeds of a tree of T. petiolaris, which was growing near T, americana, the two flowering at the same time, are identical with trees of T. spectabilis. See p. 1686. Tilia 1679 Tilia petiolaris is a beautiful weeping tree, which has not been nearly so generally planted as it deserves to be. There are good examples in the Botanic Gardens at Kew, Cambridge, and Glasnevin. At Stowe, near Buckingham, there is a hand some tree, 74 ft. by 6 ft. 8 in. There are two fine trees, girthing 7 ft. 4 in. and 5 ft. 7 in., and about 80 ft. high, growing on the bank of the Thames near Cliveden on the Wharfe Estate, belonging to Lord Boston. A very similar one on the lawn at Barton, Suffolk, was measured by Henry as 83 ft. by 8 ft. 2 in. in 1908. Another at Chiswick House, which has the trunk decayed on one side, measured 77 ft. by 10^ ft. in 1903. At Bicton a handsome tree, near the house, grafted at seven feet from the ground, measured 80 ft. by 5^ ft. in 1906. At Hatherop Castle there is a beautiful specimen of moderate size on the lawn (Plate 374), which has layered naturally ; and many plants have been propagated from it also by artificial layers. At Gunnersbury House there is a good tree, which in 1912 measured 56 feet high by 6 ft. in girth at 4 ft. from the ground. At Aldenham there is a tree on the lawn which ripened fruit in 1911 from which I raised seedlings. I noticed many dead bees1 under it on August 20. In Scotland it appears to be perfectly hardy at Durris; and Henry found, in 1905, a tree at Bargaly, 41 ft. by 4 ft. 8 in. Mr. Bean2 saw a fine specimen at Herrenhausen, in Hanover, which was 9 ft. 2 in. in girth in 1908. (H. J. E.) TILIA MONGOLICA Tilia »wngolica, Maximowicz, in Mil. Biol. x. 585 (1880), and Enum. PI. Mongol. 118, t. n (1889); L. Henry, in Rev. Hort. 1902, p. 476, figs. 214, 215, 217 ; Rehder, in Sargent, Trees and Shrubs, i. 121, t. 61 (1903). A small tree, scarcely exceeding 30 ft. in height, and flowering when only a few feet high. Young branchlets glabrous, reddish, becoming grey in the second year. Leaves (Plate 407, Fig. 7) about z\ in. wide, acuminate at the apex, with one or two sharp-pointed lateral lobes ; base truncate or cordate ; coarsely serrate, with a few large triangular teeth, tipped with long callous points ; upper surface dark green, shining, glabrous ; lower surface glaucous, with pubescent tufts in the axils at the base and at the junctions of the primary and secondary nerves, elsewhere glabrous ; petiole glabrous. Flowers, six to twelve in a cyme ; bract stalked ; sepals erect, villous within, glabrous without ; petals erect, longer than the sepals ; staminodes five, obtuse ; stamens, as long as the sepals, thirty-five to forty, in five bundles ; style glabrous. Fruit ovoid, mucronulate, without ribs or only slightly ribbed, thick-walled, shortly tomentose. This species is very distinct in appearance, the small coarsely serrate leaves resembling those of a birch, and opening with a reddish tint in spring. 1 In 1908, the bodies of innumerable bees, poisoned by the flowers of a tree of T. petiolaris at Tort worth, had so much manured the ground under its outer branches, that a very green ring of turf was visible in the autumn following, and was noticed by the Earl of Ducie to be even more conspicuous in 1909. 2 Kew Bull. 1908, p. 392. i68o The Trees of Great Britain and Ireland This tree occurs in northern China, at Jehol, and on the Po-hua mountain, west of Peking ; and has also been found in the Moni Ula range, north of Ordos, in Mongolia. It was introduced into cultivation by Dr. Bretschneider,1 who sent seeds from Peking to the Museum at Paris in 1880, and to the Arnold Arboretum in 1882. The specimen in the Jardin des Plantes at Paris is about 20 ft. high. A small tree at Kew flowered, when only 5 ft. high, at the end of July 1907. This species has lately been introduced into the Coombe Wood Nursery by Mr. Purdom, who has been collecting for Messrs. Veitch in northern China. (A. H.) TILIA PAUCICOSTATA Tilia pauticostata, Maximowicz, in Act. Hort. Petrop. xi. 82 (1890); Schneider, Laubhokkmide, ii. 371 (1909) ; V. Engler, Monog. Gatt. Tilia, 87 (1909). Ttlia Miqueliana, var. chinensis, Diels, in Engler, Bot. Jahrb. xxxvi., Beibl. No. 82, p. 75 (1905) (not Szyszylowicz). A small tree. Young branchlets glabrous, green. Leaves about 2\ in. long and 2 in. wide, ovate, usually truncate and rarely cordate at the base, ending at the apex in a long non-serrate acuminate cusp ; green and glabrous on both surfaces, except for minute axil-tufts of pubescence beneath, which are, however, absent at the base of the blade ; tertiary veins on the lower surface few, irregular, not parallel, but anastomosing, more or less prominent ; margin with regular fine serrations ending in long points ; petiole about an inch long, green, glabrous. Cymes erect, each with seven to fifteen flowers ; bracts glabrous, stalked ; staminodes present ; style pilose at the base. Fruit globose, tomentose, faintly five- ribbed. This species differs from T. cordata and its allies, in the prominent tertiary venation and the green and not glaucous under surface of the leaf, which is usually truncate at the base. It is a native of the provinces of Kansu, Shansi, and Szechwan, in western China, where it was collected by Potanin, Giraldi, and Wilson. The latter sent a living plant2 in 1901 to Coombe Wood, which I consider to be probably of this species. From it many grafts have been taken, and it now produces coppice shoots with large leaves, which show at the base of the blade and on the adjoining end of the petiole a trace of scattered stellate pubescence. This is probably a juvenile character, disappearing on adult plants ; and a young tree at Kew, one of the grafts, about 8 ft. high, bears leaves similar to those of the adult wild tree, though slightly larger, and in these the stellate pubescence has almost disappeared. (A. H.) 1 Hist. Europ. Bot. Disc. China, ii. 1050 (1898). The seedlings mentioned here as being alive at Kew in 1893 cannot be traced. 3 This is probably the plant referred to T. Miqueliana, var. chinensis, in Hortus Veitchii, 381 (1906). Tilia 1681 TILIA OLIVERI Tilia Oliveri, Szyszylowicz, in Hooker, Icon. Plant, ad t. 1927 (1890); Schneider, Laubholzkunde, ii. 387 (1909); V. Engler, Monog. Gatt. Tilia, 114 (1909). Tilia penditla, V. Engler, ex Schneider, Laubholzkunde, ii. 387 (1909), and Monog. Gait. Tilia, 113(1909). Tilia mandshurica, Szyszylowicz, in Hooker, Icon. Plant, ad t. 1927 (1890) (not Ruprecht and Maximowicz). A tree, attaining in western China about 50 ft. in height. Young branchlets glabrous. Leaves variable in size, usually longer than broad, averaging 3 to 4 in. in length and z\ to 3^ in. in breadth, orbicular-ovate, cordate at the base, cuspidate at the apex ; margin regularly serrate, with shallow sinuses between the crenate teeth, which are very short and end in cartilaginous points ; upper surface dark green, glabrous ; lower surface covered with a dense white thin tomentum, without axil-tufts ; petiole glabrous, one-half to three-fourths the length of the blade. Cymes usually much longer than the leaves, each with about twenty flowers, which are similar to those of T. tonientosa, but are smaller in size and on short thickened pedicels. Fruit globose, grey tomentose and tuberculate on the surface, thick-shelled, apiculate, ^ in. in diameter. This species, which promises to be a beautiful ornamental tree, is readily distinguished from the other limes with a pure white under surface to the leaves, by the glabrous branchlets and petioles, and the crenately serrate orbicular leaves. This species is a native of central China, where it was discovered by me in 1888, in the mountains north of the Yangtze, in the Fang and Wushan districts of Hupeh. T. Oliveri was founded by Szyszylowicz on a branch (Henry, No. 7089) from a small shrubby tree, growing in a sunny exposure on high cliffs, and bearing in consequence small leaves, averaging 2 in. in length. Another specimen (Henry, No. 7452 B), gathered by me at no great distance, but in a shaded valley, bore leaves averaging 3^ in. in length, and was identified by Szyszylowicz with T. mandshurica, which is a native of northern China. This specimen has been made the type of a new species, T. pendula, V. Engler. A third specimen, collected since by Wilson (No. 2274), with leaves intermediate in size, is considered by Schneider to be T. pendula, but by Engler to be T. Oliveri. A careful examination of the whole material shows that all the specimens belong to one species. T. Oliveri was introduced by Wilson, who sent seed from central China in 1900. It is now growing vigorously at Coombe Wood, and a small specimen is thriving at Kew. (A. H.) 1682 The Trees of Great Britain and Ireland TILIA MANDSHURICA Tilia mandshurica, Ruprecht et Maximowicz, in Bull. Acad. St. Pétersb. xv. 124 (1856); Maxi- mowicz, Prim. Fl. Amur. 62 (1859), and in Mél. Biol. x. 586 (1880); Baker and Moore, in Journ. Linn. Soc. (Bot.) xvii. 380 (1879); Franchet, PI. David, i. 60 (1884); Forbes and Hemsley, in Journ. Linn. Soc. (Bot.) xxiii. 94 (1886); Komarov, in Act. Hort. Petrop. xxv. 28 (1907). Tilia pekinensis, Ruprecht, in Bull. Acad. St. Pétersb. xv. 125 (1856), and in Maximowicz, Prim. Fl. Amur. 469 (1859); Bayer, in Verh. Zool. Bot. Ges. Wien, xii. 49 (1862). A tree, attaining in Manchuria about 60 ft. in height and 10 ft. in girth. Young branchlets and buds covered with brownish tomentum. Leaves, 4 to 5 in. in breadth and length, orbicular-ovate, usually cordate at the base, cuspidate at the apex ; margin, often with one or two lobes, coarsely serrate, the teeth ending in long awn- like points ; upper surface with a scattered stellate pubescence, which forms dense tufts at the base of the blade ; lower surface densely covered with whitish stellate tomentum, but without axil-tufts ; petiole half the length of the blade, stout, brown tomentose. Flowers similar to those of T. tomentosa, but with bracts, pedicels, and sepals more densely covered with a brownish tomentum. Fruit globose, tomentose, and slightly warty, either without ribs or with five indistinct ribs towards the base. This species is closely allied to T. tomentosa ; but has larger leaves, with long- pointed serrations and different fruit. Young trees have usually lobed leaves, as is often the case in other species; and T.pekinensis? founded on this character, cannot be retained even as a distinct variety. Tilia mandshurica is widely spread throughout the whole of Manchuria, and also occurs in Korea, where it was found on the Diamond Mountains by Père Faurie. It is not uncommon in the mountains west and north of Peking. It occurs scattered or in groups throughout the broad-leaved forest of these regions. It is known to the Chinese, like all the other species of lime, as the tuan tree ; and the bark is used for making ropes and sandals. It was probably introduced by Maximowicz into the St. Petersburg botanic garden ; but is extremely rare in cultivation, the only specimen which I have seen being a small tree at Kew, which was procured from Booth of Hamburg in 1871. As it comes into leaf very early in the spring, it is often cut by frost, and is not in a thriving condition. (A. H.) 1 Var. pekinensis, Engler, ex Schneider, Laubhokkunde, ii. 384 (1909). Tilia 1683 TILIA MAXIMOWICZIANA Tilia Maximmt>icziana, Shirasawa, in Bull. Coll. Agric. Univ. Tokyo, iv. 158, t. xviii (1900), and Icon. Ess. Forest. Japon, ii. t. 50 (1908); Schneider, LaubJwlzkunde, ii. 385 (1909). Tilia Miqueliana, Sargent, in Garden and Forest, vi. in, fig. 19 (1893), and Forest Flora of Japan, 19, t. 8 (1894) (not Maximowicz). Tilia Miyabei, Jack,1 in Mitt. Deut. Detid. Ges., 1909, p. 285. A tree, attaining in Japan 100 ft. in height and 10 ft. in girth. Young branchlets densely covered with a greyish brown tomentum. Leaves usually large,2 about 5 in. in breadth and length, cordate at the base, cuspidate at the apex ; margin ciliate, coarsely and regularly serrate, the serrations ending in blunt cartilaginous points ; upper surface dark green, with scattered stellate tomentum on the surface between the nerves, and dense tomentum on the nerves, especially at the base of the blade ; under surface greyish, densely covered with stellate tomentum, and with conspicuous brownish axil-tufts at the junctions of the nerves ; petiole stout, less than half the length of the blade, covered with greyish brown tomentum. Buds covered with greyish or rusty brown tomentum. Flowers, ten to eighteen in each pendulous tomentose cyme ; bract shortly stalked, strongly veined and tomentose on both surfaces, the tomentum brown and very dense on the midrib ; sepals lanceolate, acuminate, pubescent on both surfaces ; petals keeled, glabrous ; staminodes five, keeled, toothed at the apex, as long as but narrower than the petals ; stamens sixty-five to seventy-five, united in five bundles, shorter than the petals, each half of the anther on a short stalk ; ovary pubescent, ovoid ; style glabrous. Fruit globose, about ^ in. in diameter, with a thick woody grey pubescent five-ribbed shell. This species is a native of Japan, ranging from central Hondo (lat. 36°) north wards to Hokkaido. In Hondo, according to Shirasawa,3 who has observed it in the provinces of Kotsuke, Rikuchu, and Mutsu, it grows at altitudes of 800 to 1600 ft., in deep valleys, in mixed woods with Almis tinctoria, Populus Sieboldii, and Quercus grosseserrata. It appears to be more common in Hokkaido, where it was seen by Elwes at Sapporo and Asahigawa, in virgin forest at 500 to 750 ft. above sea-level, where it attains about 100 ft. in height and forms a wide-spreading tree, with a stem 10 ft. in girth.4 Shirasawa states that its wood is of little use except for firewood ; but the bark, after steeping, is plaited by the Ainos into coarse cloth and mats. There is a small tree of this species at Grayswood, Haslemere, which the late Mr. Chambers procured from the Yokohama Nursery Co. in 1894. It has not suffered from frost, but is rather slow in growth. Elwes has raised plants at Colesborne from 1 Jack proposes this name on account of the earlier use of T. Maximcnuiceii, Baker, in Journ. Bet. xxxvi. 319 (1898), an untenable name for another species. Cf. p. 1657, note 3, B. 3 Those on coppice shoots, gathered by Elwes in Hokkaido, are 8 to 10 in. long. 3 Specimens collected by Shirasawa in different localities are now in the Kew Herharium. 4 Sargent says it attains loo ft. in height in Hokkaido, where it frequently grows in company with T. japonica, which is much smaller in size. Cf. p. 1657, note 3, C. 1684 The Trees of Great Britain and Ireland seed sent by Shirasawa in the spring of 1905. These are 3 to 5 ft. in height, and differ from adult trees in having only traces of stellate pubescence on the leaves and branchlets. (A. H.) TILIA MIQUELIANA Tilia Miqueliana? Maximowicz, in Bull. Acad. Imp. Sc. St. Petersb. xxvi. 434 (1877), and Mél. Biol. x. 587 (1880); Shirasawa, Icon. Ess. Forest. Japon, i. text 116, t. 72, figs. 11-24 (1900), and in Bull. Coll. Agric. Univ. Tokyo, iv. 160 (1900)5 Schneider, Laubholzkunde, ii. 385 (1909); V. Engler, Monog. Gatt. Tilia, m (1909). Tilia'mandshurica, Miquel, Prol. 206 (1867) (not Ruprecht and Maximowicz); Franchet et Savatier, Enutn. PL Jap. i. 67 (1875). Tilia Kinashii, Léveillé and Vaniot, in Bull. Soc. Bot. France, li. 422 (1904). Tilia Franchetiana, Schneider, in Fedde, Repert. vii. 201 (1909), and Laubholzktmde, ii. 386 (1909). A tree, attaining in Japan about 40 ft. in height. Bark on young trees grey, smooth ; on old trunks fissured longitudinally. Young branchlets with a minute grey stellate pubescence, very variable in amount. Leaves (Plate 407, Fig. 9) extremely variable in shape, but usually much longer than broad, averaging 3 to 4 in. in length and 2 to 2\ in. broad, ovate or deltoid ; cordate at the base ; acute, slightly acuminate or rounded, but rarely cuspidate at the apex ; often lobulate and coarsely, irregularly, and sharply serrate, the teeth ending in incurved short callous points ; upper surface shining green, with scattered slight pubescence, mostly on the nerves and at the base of the midrib ; lower surface covered with a grey stellate tomentum, without axil-tufts ; petiole short, less than half the length of the blade, with minute stellate pubescence. Buds grey tomentose. Flowers, ten to twenty-two in a cyme ; bract, peduncle, pedicels, and bracteoles tomentose ; sepals ovate, acute, tomentose on both surfaces, shorter than the narrow obovate petals ; stamens sixty to seventy-five, united in five clusters ; staminodes more slender and shorter than the petals ; ovary and base of the style covered with pale hairs. Fruit nearly globose, five-ribbed at the base, grey tomentose and warty. This species is unknown in the wild state, and only occurs in Japan as a planted tree, most often seen in the courts of Buddhist temples, where it is regarded as sacred. It is reported by tradition to have been introduced from China by a Buddhist priest about the year 1190 A.D. ; but it has not yet been found anywhere in that country.2 Its extreme variability points to a hybrid origin. The species was founded by Maximowicz on two specimens, which he regarded as two forms : one with ovate-orbicular oblique leaves, long slender cymes, and broad bracts about 4 in. in length ; the other with remarkably deltoid nearly symmetrical leaves, short cymes, and narrow bracts about 2^ in. long. It is impossible, however, following the opinions of Schneider and V. Engler, to retain those two forms as two distinct 1 The tree described and figured under this name hy Sargent, in Garden and Forest, vi. in, fig. 19 (1893), and Forest Flora Japan, 19, t. 8 (1894), is T. Maximowicziana. 2 Tilia Miqrieliana, var. chinensis, Szyszylowicz, in Hooker, Icon. Plant, ad t. 1927 (1890), collected by me in Hupeh in central China, is referred by V. Engler, Monog. Gatt. Tilia, 130 (1909), to T. chinensis, Maximowicz, in Act. Hort. Petrop. xi. 83 (1890), a plant collected in Kansu. The Kansu and Hupeh plants do not seem to be quite identical ; but neither can be considered the same as the Japanese T. Miqtteliana. They have not heen introduced into cultivation. Tilia 1685 varieties or species ; for the trees cultivated in Europe show great variation, no two specimens being alike, and both deltoid and ovate leaves are occasionally present on the same branch. Recent specimens from Japan, collected by Shirasawa, cannot be exactly matched with either of Maximowicz's specimens. This species has lately been introduced into England, and the only specimens which we have seen are young trees at Kew, Aldenham, Casewick, and Glasnevin. The tree at Kew, obtained from Hesse's nursery in 1900, is now about 12 ft. high and appears to be thriving. It is also cultivated by Simon-Louis at Plantières, Metz. (A. H.) TILIA AMERICANA, AMERICAN LIME, BASS-WOOD Tilia americana, Linnaeus, Sp. PI. 514 (1753) ; Loudon, Art. et Frut. Brit. i. 373 (1838); Sargent, Suva N. Amer. \. 52, tt. 24, 25 (1891), and Trees N. Amer. 671 (1905). Tilia earoliniana, Miller, Diet. ed. 8, No. 4 (1768). Tilia latifolia, Salisbury, Prod. 367 (1796). Tilia nigra, Borkhausen, Handb. Forstbot, ii. 1219 (1800). Tilia glabra, Ventenat, Mem. Acad. Sc. Paris, iv. 9 (1802). Tilia canadensis, Michaux, FI. Bor. Am. i. 306 (1803). Tiua stenopetala, Rafinesque, FI. Ludovic. 92 (1817). A tree, occasionally attaining in America 130 ft. in height and 12 ft. in girth. Bark deeply fissured, with ridges broken on the surface into small thin scales. Young branchlets green, glabrous. Leaves (Plate 407, Fig. i) broadly oval, averaging 5 in. to 6 in. long and 3^ in. to 4^ in. broad, turning yellow in autumn, cordate or truncate at the base, cuspidate at the apex ; margin ciliate, with coarse triangular serrations, ending in long callous points ; upper surface dull dark green, glabrous ; lower surface1 light green, glabrous, except for minute pubescent tufts in the axils at the junctions of the midrib, primary, and secondary nerves, but absent at the base of the blade ; petiole glabrous, i^ in. to 2 in. long. Cymes pendulous, many-flowered ; bract stalked, broad and rounded at the apex, glabrous except for slight stellate pubescence on the midrib beneath ; peduncle glabrous ; pedicels slightly pubescent ; sepals ovate, acuminate, densely hairy within, slightly pubescent without, shorter than the lanceolate petals ; staminodes present. Fruit globose or ovoid, without ribs, covered with thick rufous tomentum ; shell thick. The American lime is readily distinguished from the other species with glabrous branchlets and leaves, by the minuteness of the axil-tufts, which are, moreover, not present at the base of the leaf. In winter, the branchlets are shining, glabrous ; buds with three external scales, glabrous except for tufted cilia at their tips. The leaves are remarkable for their variation in size, an interesting account of which is given by Penhallow.2 In Canada, some trees have tolerably uniform leaves, about 3§ in. in diameter. Other trees have in addition many leaves 4^ in. across. On vigorous shoots, especially on epicormic branches, the leaves are often 8 in. 1 In T. americana the midrib and principal nerves are remarkably yellow, and this character is also seen in its reputed hybrids, T. Moltkei, T. spectabilis, and T. Michauxii. * Canadian Record of Science, ix. 291 (I9°5)- VII F 1686 The Trees of Great Britain and Ireland wide. This lime coppices freely, and on shoots so produced the foliage is always enormous, reaching a maximum of 10 in. in diameter. This variation is due in part to the amount of nutrition available, which is greatest on coppice shoots, owing to their extensive root system ; but the size of the leaves is also much influenced by shade. Trees with large leaves are often known in gardens as var. mississippiensis, but this name is erroneous, as there is no reason to suppose that such trees are limited to the Mississippi basin. Tilia americana, since its introduction into Europe, has given rise to hybrids1 with other species, especially with T. tomentosa, which is probably the other parent in the following :— i. Tilia Moltkei, Schneider, Laiibholzkunde, ii. 381 (1909). A tree with leaves similar in shape and serrations to those of T. americana, but larger, 6 in. to 7 in. long and 5 in. to 6 in. wide, and differing in the lower surface, which is pale or greyish green, more or less covered with scattered stellate pubescence, without axil-tufts, but with occasional long hairs on the midrib. Buds, petioles, and branchlets glabrous, and identical with those of T. americana. This tree, which is remarkable for its vigorous growth and handsome foliage, originated2 in Späth's nursery at Berlin, and is named after the famous general, who planted a young tree in front of Herr Späth's house in 1888. Specimens are growing at Kew and Aldenham. •2.. Tilia spectabilis, Dippel, Laubholzkunde, iii. 73 (1893). A tree resembling T. Moltkei in the shape, colour, and serration of the leaves, but differing as follows :—Young branchlets apparently glabrous, but showing traces of stellate pubescence. Buds pubescent in their upper half. Leaves (Plate 407, Fig. n), variable in size, 4 in. to 6 in. long and 3 in. to 5 in. wide; under surface with scattered stellate pubescence, very variable in amount,8 and with long hairs on the principal nerves as well as on the midrib, but without axil-tufts ; petiole glabrous or with a few scattered hairs. The flowers are intermediate in character between those of T. americana and T. tomentosa, and open at Kew three weeks earlier than those of the latter species. The stellate pubescence on the bract, peduncle and pedicels is like that of T. tomentosa, but less in amount. The shape of the bract and the size of the flowers are similar to T. americana. This tree has probably been long in cultivation, but I have not been able to trace its origin.4 It is often planted in botanic gardens, as at Kew, where there are several specimens, one nearly 40 ft. in height.5 At Cambridge a tree, about 35 ft. by 3^ ft. in 1907, has long been labelled T. heterophylla ; and reputed trees of the latter species usually turn out to be either T. spectabilis or T. Michaiixii. 1 In America it probably forms hybrids with T. heterophylla. See T. Michauxii, p. 1689. 2 It is mentioned as a new plant, Tilia americana Moltkei, in Späth's Catalogue, No. 57, p. 71 (1883). 3 In cases where the stellate pubescence on the under surface of the leaves becomes dense and greyish, trees of T. spectabilis are often confused with T. tomentosa and T. Michauxii. They are readily distinguishable from the latter by the absence of axil-tufts, and from the former by the glabrous branchlels. The long hairs on the midrib and nerves seem to be peculiar to T. spectabilis and T. Mollkei. 4 Dippel found it in 1893 in Froebel's nursery under the name T. alba spectabilis, and in the Zoeschen arboretum as T. Blechiana ; hut it must have originated much earlier. Cf. ante, p. 1678, note 5. 6 This tree has been labelled T. argcntca, and resembles T. tomentosa in habit. Tilia 1687 3. Tilia viridis, Simonkai, in Math. Term. Koezl. xxii. 320 (1888). In addition to the usual form of T. spectabilis described above, there appears to be another hybrid of the same parentage, which is represented at Kew by a tree about 25 ft. high, obtained under the name T. spectabilis from Späth in 1899. It is closer to T. tomentosa, as the buds, branchlets, and upper surface of the leaves are covered with a scattered stellate pubescence, which is denser on the under surface of the leaves than in ordinary T. spectabilis. This tree closely resembles, if it is not identical with, a specimen * taken from a tree cultivated at Baden in 1835, which was identified by A. Braun with T. argentea, var. virescens, Spach, in Ann. Sei. Nat. ii. 344 (1834). Spach's description was based on a tree in the Trianon, which was said to have been raised from seeds of T. tomentosa ; and if this account is correct, this tree would appear to have been the first cross observed between T. tomentosa and T. americana. DISTRIBUTION Tilia americana is widely spread, occurring in Canada from the valley of the Assiniboine River and the southern shores of Lake Winnipeg eastward to northern New Brunswick, and extending in the United States southwards to Virginia and along the Appalachian mountains to Georgia and Alabama, and ranging westward to Dakota, Nebraska, Kansas, Indian Territory, and eastern Texas. It grows in rich often moist soil, and formerly occurred as pure forest.2 It attains its largest size in the alluvial lands of the lower Ohio river, where Ridgway records a tree 135 ft. in height and 17^ ft. in girth.8 According to Loudon, the American lime was cultivated4 by Miller in 1752, but had not been extensively distributed. It is very rare at the present time in cultiva tion, and the tree at White Knights, mentioned by Loudon as 60 ft. in height, is no longer living. The only specimens which we have seen in England are small trees at Kew, Eastnor Castle, and Liphook. (A. H.) TIMBER The wood of the American limes is very similar in character to that of the European species, and according to Sargent is largely used in the United States under the name of whitewood for the manufacture of cheap furniture, carriage panels, and woodenware. He states that, though one of the woods most largely used for making pulp, the quick decomposition of the sap makes it unfit for white paper. It is imported to some extent into Europe under the name of basswood, and has 1 This specimen is considered by Bayer, in Vcrhand. Zool. Bot. Ges. Wien, xii. 50 (1862), to be a hybrid between T. tomentosa and T. americana, which he calls T. argentea-nigra. V. Engler (op. cit. 152) considers it to be T. cordata x T. tomentosa, but it shows no trace of T. cordata parentage. 2 In U.S. Forest Service, Circular 63 (1907), which gives an account of this species, it is said to do well when planted in pure stands, and to be the most prolific of American trees in shoots from the stumps. 3 Sargent, Bull. Pop. Inform. No. 30 (1912), and in Gard. Chron. Hi. 87 (1912), says that it shows its greatest beauty ill the forests of New Brunswick, northern New England, and the valley of the St. Lawrence. The leaves of planted trees in eastern Massachusetts are, especially in dry summers, made brown by the red spider, which, however, is easily contiollecl by spraying. 4 In the London Catalogue of Trees, 1730, p. 81, the Carolina lime tree is mentioned as " Tilia, with leaf more longly mucronale. Seeds sent from Carolina by Catesby in 1726, hardy, and may be propagated as other limes." 1688 The Trees of Great Britain and Ireland partially displaced native limewood in the pianoforte trade on account of its cheapness. Owing to the facility with which large thin sheets can be turned off the log by rotating it against a knife-edge, it is coming into use for three-ply boards, a manu facture which, though only recently invented, is likely to grow rapidly for many purposes. (H. J. E.) TILIA HETEROPHYLLA Tilia heterophylla, Ventenat, in Anal. Hist. Nat. Madrid, ii. 68 (1800), and Mém. Acad. Se. Paris, iv. 16, pi. 5 (1803) ; Sargent, Silva N. Amer, i- 57, t 27 (1891), and Trees N. Amer. 674 (1905). TiUa americana, Linnseus, var. /leterophylla, Loudon, Arb. et Frut. Brit. i. 375 (1838). A tree, attaining in America 60 ft. in height and 12 ft. in girth. Young branchlets glabrous. Leaves similar to those of T. Michauxii in size and shape, but differing in being covered beneath with a silvery white tomentum, without axil- tufts, and having finer serrations with shorter straighter points. The flowers appear to differ in the bract1 pubescent on both surfaces, and the stellate-pubescent peduncle and pedicels. There appears to be no constant difference in the fruit. Both this species and T. Michauxii are readily distinguished from the other white limes by the glabrous branchlets, and the different shape of the leaves, which are usually very oblique at the base, and always longer than broad. T. pubescens, Aiton, the white lime of the Gulf States, is not in cultivation, and probably would not live in our climate. It has pubescent branchlets. T. heterophylla, according to Sargent, is found on rich wooded slopes or near the banks of streams; and ranges from Ithaca, New York, southwards along the Alleghany mountains to northern Alabama, and westward to middle Tennessee, Kentucky, southern Indiana, and Illinois. It is most abundant and of its largest size on the high mountains of North Carolina and Tennessee. Typical T. heterophylla is described by most authors as having the leaves covered beneath with a silvery white tomentum ; but Sargent informs me in a letter that Rehder found trees on the mountains of West Virginia, with leaves nearly glabrous beneath,2 which he considered to be undoubtedly T. heterophylla. Such trees are probably hybrids with T. americana ; and a further study of the American limes is necessary, as the relationship of T. Michauxii and the glabrous forms of West Virginia with T. heterophylla in its typical form, is very obscure. T. heterophylla was introduced, according to Loudon, in 1811 ; but he had seen no specimens except small trees in the Chiswick Garden ; and these may have been T. Michauxii, which he did not distinguish as a separate species. Most of the trees which occur in cultivation under this name are either T. Michauxii or T. spectabilis ; and we have seen no living specimens in England which can be identified with T. heterophylla. (A. H.) 1 The bract is said by Schneider to be stalked in T. heterophylla, and sessile in T. Michauxii ; but native specimens show it to be variable in both species. 2 T. eburnea, Ashe, in Bot. Gaz. xxxiii. 231 (1902), usually considered to be a peculiar variety of T. heterophylla in the Carolinas and Georgia, is said to lose its pubescence in autumn occasionally. Tilia 1689 TILIA MICHAUXII Tilia MicJiauxii, Nuttall, N. Amer. Sylva, i. 108 (1865)5 Sargent, Trees N. Amer. 673 (1905); Schneider, Laubholzkutide, ii. 387 (1909). Tilia alba, Michaux, Hist. Arb. Amer. Sept. iii. 315, t. 2 (1813), and Nuttall, N. Amer. Sylva, iii. 84, t. 132 (1865). (Not Aiton, Hort. Kew?) A tree, attaining in America 80 ft. in height and 9 ft. in girth. Young branchlets and buds glabrous. Leaves (Plate 407, Fig. 5) very variable in size, averaging1 4 to 5 in. in length, and 3 to 4 in. in breadth, always much longer than broad, ovate, very oblique and cordate or truncate at the base, cuspidate at the apex ; margin with large triangular serrations ending in long incurved points ; upper surface dark green, glabrescent ; lower surface covered with a scattered or dense greyish tomentum, very variable in amount, and with conspicuous axil-tufts at the junctions of the midrib, primary and secondary nerves ; petiole stout, short, glabrescent. Cymes, ten- to twenty-flowered ; bract tomentose above, glabrous below-; peduncle glabrous ; pedicels slightly pubescent ; sepals lanceolate, pubescent on both surfaces ; petals lanceolate, acute, twice as long as the sepals ; staminodes present. Fruit globose, marked with five furrows and minutely tuberculate on the grey slightly tomentose surface. This species was first described and figured by Michaux, who, however, considered it to be the same as the European white lime. Nuttall2 based his name T. Michauxii on Michaux's description and figure. It appears to have much the same distribution as T. heterophylla, but is imperfectly known,8 and in all probability is a hybrid of that species with T. americana, resembling the latter in the serrations, axil-tufts, and yellow midribs and nerves of the leaves. The characters of the flowers are intermediate between the two species mentioned. The extreme variability in the amount of the tomentum on the leaves points to a hybrid origin. It is not very common in cultivation in England; but there are two fine trees at Beauport, which are both grafted ; one of these measures 84 ft. by 9 ft. i in., and the other is nearly as large. There are also small trees at Kew and Tortworth. There is an ill-shaped tree at Castlemartyr in Ireland, which Elwes found in 1908 to be about 40 ft. by 3| ft.; some of its leaves were no less than 12 in. long and 9 in. wide. It is cultivated under the name T. americana rubra in the Heatherside nursery ; and appears to be known in continental nurseries under various names, as T. americana pubescens, T. gigantea, T. macrophylla, T. hybrida superba, etc. (A. H.) 1 In cultivated trees in Europe the leaves are usually larger, averaging 7 in. long and 5 to 6 in. broad ; but this may be a juvenile character. 2 Nuttall says : " The Tilia alba of Michaux, not being the T. alba of Kitaibel and Aiton, it is necessary to change the name, and we propose to call it T. Michauxii." 3 I am unable to follow V. Engler's view, that all the trees in the wild state in America, supposed to be T. Michauxii by Sargent and others, are simply T. heterophylla. Engler restricts the hybrid T. americana X T. heterophylla to trees found in cultivation in Europe. Wild T. Michauxii in America, judging from a considerable number of specimens which I have seen, is readily distinguishable from the true T. heterophylla ; and I see no reason for supposing that the trees referred by me to T. Michauxii, which are in cultivation in Europe, are not originally from America. A small tree at Kew, received from Biltmore, U.S.A., is undoubtedly T. Michauxii. TRACHYCARPUS Trachycarpus, H. Wendland, ex Gay, in Bull. Soc. Bot. France, viii. 429 (1861); Bentham et Hooker, Gen. PI. iii. 929 (1883); Beccari, in Martelli, Webbia, i. 41-73 (1905). Chamœrops, Martius, Hist. Nat. Palm. iii. 247 (1836-1850) (in part). TALL unarmed trees, belonging to the order Palmaceae, with flabellate leaves, without a rachis, deeply divided into numerous plicate segments, which are narrow, entire in margin, and bifid at the apex. Flowers monoecious ; spadices numerous, panicled, sessile between the leaves, and embraced by sheathing coriaceous spathes ; individual flowers small ; sepals three, ovate ; petals three, broadly ovate, valvate ; stamens six, with free filaments and short dorsifixed anthers, rudimentary in female flowers ; ovary of three carpels, connate at the base, represented in the male flowers by three small subulate pro cesses ; stigmas three, recurved ; ovules basilar, erect. Fruits, one to three, globose or oblong, with a terminal style ; containing one erect seed, which is grooved on the ventral side, and has a small hilum at the base. Three species1 of Trachycarpus are known, natives of the Himalayas, Assam, Burma, and China. TRACHYCARPUS FORTUNEI, CIIUSAN PALM Trachycarpus fbrtunei, H. Wendland, in Bull. Soc. Bot. France, viii. 429 (1861); Masters, in Gard. Chron. xxiv. 304, fig. 65 (1885). Trachycarpus excelsa, H. Wendland, in Bull. Soc. Bot. France, viii. 429 (1861); Planchon, in Flore des Serres, xxii. 207, t. 2368 (1877); Beccari and J. D. Hooker, in J. D. Hooker, Flora Brit. India, vi. 436 (1892); Diels, in Engler, Bot. Jahrb. xxix. 233 (1900); Wright, in Journ. Linn. Soc. (Bot.) xxxvi. 168 (1903); Masters, in Gard. Chron. xxxv. 312, Supply. Illust. (1904); Beccari, in Martelli, Webbia, \. 41 (1905). Chamœrops excelsa, Martius, Hist. Nat. Palm. iii. 251, t. 125, figs. 2, 3 (1836-1850) (not Thunberg2); Miquel, Anal. Bot. Ind. ü. 21 (1851)5 Gay, in Bull. Soc. Bot. France, viii. 410 (1861); Houllet, in Rev. Hort. xl. 370, coloured plate (1868). Chamœrops Fortunei, W. J. Hooker, in Bot. Mag. t. 5221 (excl. figs. 6, 7) (1860). A palm, attaining 30 to 50 ft. in height, with the trunk annulately scarred and clothed throughout with old leaf-sheaths. Leaves : lamina sub-orbicular or fan- 1 The two species, not described by us, as they are scarcely hardy in cultivation, are : T. Martiana, Wendland, widely spread in the central and western Himalayas, Assam, and Burma ; and T. Takil, Beccari, in Webbia, 52 (1905), which is a peculiar species occurring in Kumaon. The latter species is cultivated by Prof. Beccari in his garden at Florence. Bean, in Kew Bull. 1912, p. 291, quotes Beccari's description and gives a figure of this tree, which is readily distinguishable from T. Fortunei by its different habit and by the closer and neater fibre of the trunk. z Chamœrops excelsa, Thunberg, Fl. Jap. 130 (1784), is plainly from the description another palm, Rhapisßabelüformis, Alton, Hort. Kew. iii. 473 (1789). Cf. Sims, Bol. Mag. t. 1371 (i8il). 1690 Trachycarpus 1691 shaped, averaging i| ft. in length and breadth, deeply cut into numerous linear plicate segments, which are \ to i in. broad, entire in margin, and bifid at the apex ; petiole about i^ ft. long, trigonous or nearly semicylindric in section, convex above, and flat beneath, with the two margins irregularly serrate or dentate towards the base ; lower sheathing part furfuraceous, and separating into a network of brown fibres. Staminate panicles1 broad, thick, and obtuse, enveloped in a strong stout brownish spathe, at first erect, soon pendent, with arched ramifications ; flowers small, numerous, close together, deep yellow-orange. Pistillate inflorescence smaller, less thick than the staminate one, clothed in an acuminate erect spathe, with distant spreading slender ramifications ; flowers numerous, not close together, small, pale yellow or greenish. Drupe reniform, hollowed on one side. This species is variable in habit, the leaves in some individuals being straight and stiff, whilst in others they droop at the tips ; and the petioles show a varying amount of serration. There are no grounds, however, for supposing that there are two forms, corresponding to a supposed Chinese and Japanese origin. Plants which are sold as distinct, under the names Chamœrops excelsa and C. Fortunei, are usually imported seedlings, which have been raised in the Riviera from the same batch of seed. 1. Var. gracilis, Carrière, in Rev. Hort, xlvii. 220 (1875). Petioles, longer and more slender than in the type, abruptly inflexed ; laminae more deeply divided with narrower segments. The variety was described from a plant in the Jardin des Plantes at Paris, which was obtained from Thibaut and Keteleer's nursery in 1862. 2. Var. surculosa, Henry (var. nova). Leaves smaller and stiffer than in the type. Dwarf in habit, and freely reproducing itself by suckers like Chamœrops humilis of the Mediterranean region. This peculiar variety, which was obtained a few years ago from Japan by Sir E. G. Loder, is possibly a distinct species. It thrives at Leonardslee, where it has produced flowers ; but no fruit has yet been noticed. T. Fortunei is a native of the central provinces of China, where it is certainly wild on the hills and lower slopes of the mountains in Szechwan, Hupeh, Kiangsu, and Chekiang. It is everywhere much cultivated, so that it is impossible to define its original distribution in the wild state. It is commonly planted around Shanghai, where it lives unprotected through the severest winters.2 It is also cultivated on the island of Chusan, whence the name given to it of Chusan palm by Fortune,3 who saw it growing wild in great perfection near Yen Chow and on the Sung Lo mountain in Chekiang. It is known to the Chinese as Tsung, and is of great economic importance, owing to the large quantity of strong and useful fibre found on the stem at the bases of the leaf-stalks. This fibre, which is similar to the coir 1 Inflorescences bearing both staminate and pistillate flowers have been noticed by Beccari. The germination of this species is described by Gatin, in Ann. Sc. Nat. (Bel.) iii. figs. 28, 29 (1906). 2 It is unknown in North China, being killed when planted out at Chefoo and Tientsin. Cf. Gay, in Ball. Soc. Bot. France, viii. 412 (1861). 3 Cf. Fortune, Wanderings in China, 53, 54 (1847); Tea Counliies, 58, 88, 117, 318 (1852); Residaice among the Chinese, 5, 145, 189 (1857); and in Card. Chron. 1850, p. 757, and 1860, p. 170. 1692 The Trees of Great Britain and Ireland obtained from the cocoa-nut, is used for making ropes and cables, which are very durable under water ; in the manufacture of hats and rain-cloaks* (called so-i), which are used by the peasants in wet weather ; and in making brushes and mattresses. This palm is largely cultivated in southern and central Japan, where it is known as Shuro, but it does not seem to be a true native,2 though it is naturalised in some places.3 It is recorded from Tamsui and Bankinsing in Formosa; but is probably only planted.4 This palm was first introduced by Siebold,5 who sent seeds from Nagasaki in Japan to Leyden in 1830. Of the plants which were raised, one was sent to Kew in 1837, where it was long an inmate of the Palm House ; but it died some years ago. The Chinese plant was introduced by Fortune,6 who sent six living plants to Kew in 1849, two of which are still flourishing in the Temperate House, and are about 40 ft. and 50 ft. in height respectively ; while two, which were planted out-of-doors, have remained stunted in growth and are only about 10 ft. high. Another of these plants, which was sent to Osborne in 1849, was planted out in front of the royal residence, and had attained7 14^ ft. high in 1881. It is now about 2i ft. high and 2 ft. 4 in. in girth at a foot from the ground. (A. H.) The Chusan palm is perfectly hardy8 in our climate, but grows slowly in the neighbourhood of London and to the northwards. It flourishes best in the south west of England and in the south of Ireland, where it attains a height of 25 to 30 ft., and ripens seeds regularly. In the grounds of Tregothnan, I found in 1905 self- sown seedlings in the grass, and transferred some of them to my own garden at Colesborne, one of which has borne a temperature close to zero without injury. On the lawn at Williamstrip Park, Gloucestershire, an old specimen has flowered on several occasions. At Duncan House, Torquay, there is a group of about six trees, 14 to 25 ft. high in 1903, around which there are hundreds of young plants, which have come from their seed.9 Plate 375 shows a tree planted by the Rev. Hon. T. Boscawen at Lamorran in Cornwall, which was about 25 ft. high in 1905. At Riverhill, near Sevenoaks, the seat of J. T. Rogers, Esq., I saw one of the original seedlings introduced by Fortune, which was 27 ft. by 2 ft. 6 in. in 1911. At Powis Castle, I measured one in 1906, 25 ft. high, girthing i ft. 10 in. at the ground, but 3 ft. 4 in. at five feet up. At Heckfield Place, near Winchfield, Hants, there are two good specimens10 of different sexes, which were planted in 1869. They measured in January 1912, \*j\ feet and 15^ feet respectively. 1 A rain-cloak and hat, brushes, cordage, and other articles made from this fibre, are displayed in the museum at Kew. 2 Makino, in Tokyo Bot. Mag. xviii. 20 (1904), divides the species into two forms, var. typica and var. Fortunei; but in this he simply follows Wendland, and gives no distinguishing characters ; moreover, he states of both these supposed forms, that they are planted and not native in Japan. 3 Cf. Franchet and Savatier, Enum. Plant. Jap. ii. 2 (1879). 4 Cf. Matsumura, Index Plant. Jap. ii. 168 (1905). Elwes saw it planted in many places in Formosa up to about 4000 ft. elevation ; and the fibre is used for the same purposes as in China. 6 Cat. Kais. PI. Jap. 1856, p. 7, note. 6 Fortune also obtained seed near Tsee-Kee, north-west of Ningpo, in 1853, from which plants were raised in England. T Cf. Smith, Records of Kew Gardens, 116 (1885). 8 Cf. Gard. C/iroii. 1860, pp. 170, 362, where it is stated that this palm suffered much but did not die in Northampton shire in the severe winter of 1859-1860 ; while at Swansea, where the temperature fell to 10° Fahr., they were quite unmjured. 9 Gard. Chron. xxxiv. 329 (1903) ; one of these palms was figured in Card. Chron. xxiv. 420, Supply Illust. (1898). «> Cf. Card. Chron. xv. 12, fig. 4 (1881), and xxiv. 216 (1885). Trachycarpus 1693 In Surrey, this palm can be grown very successfully in light dry sandy soil, provided the plants are placed in a sheltered position. In these conditions, they grow slowly, attaining about 10 ft. high in thirty years ; but they flower regularly, and are very ornamental.1 There is a fine specimen, 25 ft. high and 4 ft. in girth, at Joldwynds, near Dorking, the seat of Sir P. Bowman, Bart. At Wisbech, Lord Peckover has a very old specimen, probably one of the original seedlings, which was 25 ft. high in 1912. A younger tree, planted about thirty-five years ago, is nearly 20 ft. high, and flowers more profusely than the older tree. In Scotland2 there is a good specimen at Ardchapel, Shandon, on Gare Loch in Dumbartonshire, which was planted in 1866 by the late Prof. Swan of St. Andrews. In 1905, it was 11 ft. high, with a girth of 3 ft. 8 in. at five feet from the ground. It has produced flowers regularly since 1881. There are small but thriving specimens in Arran, at Craigard, Lamlash ; Whitefarland ; and Cromla, Corrie. In Ireland, it is slow in growth and of no great height in the open at Glasnevin. Two trees at Narrow Water, Co. Down, were 12 ft. high in 1903. It grows well at Mount Usher in Wicklow, where it ripens fruit ; and is represented at Fota by several fine specimens, one of which is 25 ft. high. The Chusan palm was introduced3 into France by seed sent from central China by M. de Montigny in 1851 ; and was first planted in Algeria in 1853. It exists in the climate of Paris, where it bore 16° Cent, below zero in 1870-1871 ; and it is hardier4 than Chamœrops humilis at Montpellier, where the temperature occa sionally falls to the same point. There are fine specimens6 in Brittany, notably one in the park of Porzantrez near Morlaix, which was planted in 1856, and had attained 27 ft. high in 1907. In Spain and Portugal there are fine specimens,6 one at Lisbon being 40 ft. high. (H. J. E.) 1 Cf. The Garden, 15th November 1902, p. 341. 2 Cf. Rev. Dr. Landsborough, in Trans. Bot. Soc. Edin. xx. 510 (1896), and xxiii. 140, 141 (1908). 3 Gay, in Bull. Soc. Bot. France, viii. 422 (1861). 1 Planchon, in Flore des Serres, t. 2368 (1877). There is a small plant of Chamarops humilis in the bamboo garden at Kew; but it would not survive without protection in winter. Landsborough, in Traits. Bot. Soc. Edin. xxiii. 141 (1908), mentions a thriving plant of this species at Cromla in Arran. 5 Pardé, in Bull. Soc. Dend. France, 1908, p. 20. 6 Ibid. 1910, p. 106. VII ACANTHOPANAX Acanthopanax, Miquel, in Ann. Mus. Lugd. Bat. i. 10 (1863); Bentham et Hooker, Gen. PI. \. 938 (1863); Viguier, in Ann. Sei. Nat. (Bot.) iv. 33 (1906). Kalopanax, Miquel, in Ann. Mus. Lugd. Bat. i. 16 (1863); Harms, in Engler and Prantl, Pflanz- Fam. iii. 8, p. 50 (1894). Brassaiopsis, Seemann, vn. Journ. Bot. ii. 290 (1864) (not Decaisnc and Planchen). Panax, sub-genus Acanthopanax, Decaisne and Planchon, in Rev. Hort. 1854, p. 105. DECIDUOUS trees or shrubs, belonging to the order Araliacese. Branchlets with or without spines. Stipules inconspicuous or absent. Leaves either compound and three- to five-digitate or simple and palmately lobed. Flowers small, in panicled or solitary umbels, polygamous or perfect, on pedicels that are not articulated ; calyx-tube coherent with the ovary, limb dentate ; petals five, rarely four, valvate in the bud ; stamens as numerous as the petals and inserted with them on the margin of the disc, filaments filiform, anthers ovate or oblong ; ovary usually two-celled, rarely five-celled ; styles, two or five, more or less connate at the base, with the apex recurved and stigmatose internally. Fruit baccate, later ally compressed or globose, with a fleshy exocarp ; nutlets two to five, crustaceous or cartilaginous, each containing one seed. About eighteen species are known, natives of Eastern Asia, extending from Amurland and Manchuria, through Japan, Formosa, and China to the Himalayas. The following species is the only one attaining timber size. ACANTHOPANAX RICINIFOLIUM Acanthopanax ricinifolium, Seemann, in Journ. Bot. vi. 140 (1868), and Revis. Heder. 86 (1868) ; Fraiichet and Savatier, Enum. PL Jap. i. 193 (1875); Hance, in Journ. Bot. 1885, p. 323; Hemsley, in Journ. Linn. Soc. (Bot.) xxiii. 340 (1888); Sargent, in Garden and Forest, vi. 234i "g- 36 (l893). and Forest Flora Japan, 45, fig. 16 (1894); Shirasawa, Icon. Ess. Forest. Japon, ii. t. 56, figs. 11-24 (!9°8); Beissner, in Mitt. Deut. Dend. Ges. 1909, p. 289; Schneider, Laubholzkimde, ii. 429 (1909). Acanthopanax acerifolium, Schelle, in Mitt. Deut. Dend. Ges. 1908, p. 217, and 1909, p. 289. Panax ricinifolium, Siebold and Zuccarini, Ft. Jap. Fam. Nat. i. 91 (1843), and in Abh. Ac. Munch. iv. 2, p. 199 (1845). Kalopanax ricinifolium, Miquel, in Ann. Mus. Lugd. Bat. i. 16 (1863); Harms, in Engler, Bot. Jahrb. xxix. 488 (1900). Kalopanax ? sp., Zabel, in Mitt. Deut. Dend. Ges. 1904, p. 63. Brassaiopsis ricinifolia, Seemann, in Journ. Bot. ii. 291 (1864). 1694 Acanthopanax A tree, attaining in Yezo 80 to 90 ft. in height, and 10 ft. or more in girth. Bark very thick,1 dark in colour, and deeply fissured. Branchlets stout, reddibh brown, glabrous, armed with short stout spines, which are enlarged at the base and are often glaucous. Leaves sub-orbicular, but broader than long, averaging 6 to ID in. in diameter, sub-cordate at the base, palmately five- or seven-nerved; on adult trees, divided to one-third of the length of the blade by acute sinuses into five ovate-triangular acuminate lobes, serrate in margin, with callous tips to the teeth ; upper surface dark green, shining, glabrous ; lower surface with slight whitish tomentum on the nerves, forming axil-tufts at their base, where they unite ; petioles long and slender, more or less covered with scattered tomentum, especially near the base. Flowers on long slender glabrous pedicels in many-flowered umbels, arranged in terminal compound panicles, which are sometimes 2 ft. in diameter ; small, white in colour, appearing in August and September. Fruit black, globose. i. Var. Maximowiczii, Schneider, Laiibholzkunde, ii. 429 (1909). Aralia Maximmviczii, Van Houtte, in Flore des Serres, xx. 39, t. 2067-2068 (1874). This varietal name is applied for the sake of convenience to young trees in cultivation, which only show juvenile foliage. The leaves differ from those characteristic of adult trees, in being deeply lobed, with the sinuses extending more than two-thirds the length of the blade. It is impossible to say whether, as a result of being perhaps propagated by grafts from young seedlings, these trees will preserve the seedling foliage indefinitely, or will, when older, develop normal leaves. Seedlings occasionally show both kinds of foliage ; and Siebold's original specimens also bear both shallow- and deeply-lobed leaves, with all intermediate stages. This species is a native of China, Korea, Japan, Saghalien, and the Liu Kiu Islands. In China, it is widely spread in the central provinces, from Szechwan and Hupeh to Chekiang and Fokien, usually forming a tree 40 to 50 ft. high in the mixed forests on the mountains at no great elevation. It is known to the Chinese as tz'e ch'iu ;2 but its timber is little valued. (A. H.) In Japan, where the tree is called hari-giri, it attains its largest size in the forests of Hokkaido, where, Sargent states, it is often 80 feet in height and 4 or 5 ft. in diameter, with a tall straight stem, covered with furrowed bark, and giving off great limbs, which spread horizontally. In central Hondo, it is smaller in size. Mayr3 who measured a tree in Hokkaido, which was 90 ft. high, states that this species is fast in growth and remarkable for its capacity of enduring shade in the forest. In Kiusiu I saw it in the forest, which covers the lower slopes of the volcano of Kireshima at 2000 to 3000 ft. altitude. Here it is not so fine a tree as in Hokkaido, the largest that I measured being about 60 ft. by 5 ft. 10 in. A. ricinifolium was introduced into Europe by Van Houtte, who figured in 1874 a young plant, which had been raised from a single seed received from the 1 The bark is well shown in a photograph of a tree growing in Japan, reproduced by Jack in Mitt. Detit. Dend. Ges. 1909, p. 285. 2 Cf. Bretschneider, Bot. Sinic. ii. 344 (1892) and iii. 480 (1895). 3 Fremdländ. Wald- a. Paikliäume, 437 (1906). 1696 The Trees of Great Britain and Ireland St. Petersburg botanic garden. A tree* in the forest garden at Münden, probably derived from Van Houtte's original specimen, was 21 ft. high in 1895. This species is best raised from seed imported from Japan ; but it can be propagated by cuttings of the half-ripened wood, which strike root if placed in gentle heat.1 Mayr says that it bears transplanting badly, often dying to the ground, but afterwards producing an abundance of suckers from the collar of the root. It ripens its young shoots well, and is perfectly hardy at Colesborne, where seedlings, that I raised in 1906, are now 4 to 6 feet high. The best specimen that I know in England is in the beautiful grounds of South Lodge, Horsham, the seat of F. D. Godman, Esq., which was, in 1905, 28 ft. high, fifteen years after planting. A solitary specimen at Kew has done badly, and appears to suffer from spring frost. At Westonbirt, there is a vigorous young tree, about 20 ft. high.2 At Grafrath, near Munich, this species has been planted in plots, and has thriven well for so far ; but the trees are too young to give as yet any evidence as to their value in forestry.3 M. Hickel states that there is a fine specimen in Messrs. Barbier's nursery at Beuvronne.4 TIMBER Mayr says that this species yields a light timber, which, though disagreeable in odour, is used in Japan for building generally, and for making shafts of spears. Sargent states that the wood is rather hard, straight-grained, and light brown, with a fine satiny surface. The timber has been largely imported into England during the last few years under the name sen, by which it is known in Hokkaido. It has been sold in some cases as Japanese ash, and mixed with the timber of Fraximis mandshurica, which it somewhat resembles ; but it is better fitted for making furni ture than for the purposes for which ash is generally used. (H. J. E.) 1 Rehder, in Mailer's Deut. Gart. Zeit, ist July 1897. 2 In Gard. Mag. 1888, p. 526, mention is made of a plant then 12 ft. high in Messrs. Veitch's nursery at Coombe Wood. This was disposed of some years ago. 3 Weiss, in Milt. Deut. Dend. Ges. 1912, p. 3, states that it has been tried in groups in the forest at Augsburg, where the trees are about 20 ft. high, but show no sign of being useful. 4 Cf. Hickel, in Bull. Sac. Dend. France, 1907, p. 153, and 1909, p. 230. ACACIA Acacia, Adanson, Fam. des fiantes, ii. 319 (1763); Willdenovv, Si>. PI. iv. 1049 (I8os); Bentham et Hooker, Gen. PI. i. 594 (1865). TREES, shrubs, climbers, and rarely undershrubs, belonging to the division Mimoseae of the order Leguminosae. Branchlets with or without stipular or infra-stipular spines. Leaves alternate, either compound and equally bipinnate with minute leaflets, or reduced to simple phyllodes,1 which are equivalent to dilated and flattened petioles, with their surfaces placed vertically. Flowers, yellow or white ; in globose heads, cylindrical spikes, or panicles, which are solitary or fascicled in the axils oi the leaves ; or panicled and ending the branchlets ; perfect or polygamous, small, regular : sepals five, four, or three, rarely absent, free or united ; petals as many as the sepals, free or united ; stamens numerous, free or slightly connate at the base ; ovary sessile or stalked, with usually numerous (rarely only two) ovules ; style long and slender, ending in a minute stigma. Pod linear or oblong ; flat or nearly cylindrical ; straight, falcate, or twisted ; opening by two valves or indéhiscent. Seeds, more or less flattened, usually marked on each face with an oval or horseshoe-shaped depression, oblique spot or ring ; funicle filiform or thickened into a flat aril under or around the seed. About 450 species of Acacia are known, natives of the warmer parts of the world, and occurring in Africa, Asia, America, Australia, and Polynesia. A considerable number of species are cultivated under glass in this country ; but in the south-west of England, and in the south of Ireland, four or five species, some of which are shrubs, have succeeded out-of-doors. Two of these species, representative of the two kinds of foliage which occur in the genus, will now be described. ACACIA DEALBATA, SILVER WATTLE Acacia dealbata, Link, Enu?n. Hort. Berol. ii. 445 (1822); Loddiges, Bot. Cab. t. 1928 (1833); Loudon, Art. etFrut. Brit. ii. 666 (1838); J. D. Hooker, Fl. Tastn.i. in (1860); Bentham, Fl. Austral, ii. 415 (1864); Gamble, Indian Timbers, 301 (1902); Rodvvay, Tasm. Fl. 43 (1903). Acacia irrorata, Sieber, ex Sprengel, Syst. iii. 141 (1826). Acacia decurrens, Willdenow, var. dealbata, Von Mueller ex Maiden, Forest Flora N. S>. Wales, iii. 56 (1908). An evergreen tree, attaining occasionally in Tasmania 100 ft. in height and 1 These phyllodes are not leaves turned edgeways, as is shown by the fact that they are not twisted at the base. Moreover, phyllodes are occasionally produced, which bear bipinnate leaves at their ends. On seedlings of thï species of Acacia with phyllode foliage, the first leaves are bipinnate ; succeeding leaves have flattened stalks with fewer leaflets ; ultimately only phyllodes are produced. 1697 1698 The Trees of Great Britain and Ireland 11 ft. in girth ; but usually much smaller. Young branchlets angled, hoary, covered with a minute whitish pubescence. Stipules reduced to inconspicuous scales. Leaves equally bipinnate ; rachis 3 to 4 in. long, hoary and pubescent like the branchlets, often glandular ; pinnae ten to twenty pairs, hoary and pubescent, each bearing thirty to forty pairs of crowded linear leaflets, which are about ^ in. long, pubescent, glandular at the sessile base. Flower heads, in axillary and terminal panicled racemes, globose, yellow, about ^ in. in diameter ; flowers twenty to thirty in a head, mostly pentamerous. Pods, straight or curved, flattened, 2 to 3 in. long, \ to \ in. wide, not or slightly constricted between the seeds, glaucous on both surfaces. A. dealbata, which is known in Australia as the silver wattle, occurs in New South Wales, Victoria, and Tasmania. It is widely distributed in Tasmania, where it usually attains 50 ft. in height and 3 to 6 ft. in girth ; and yields a timber of little value, which is used for making staves of cheap casks. The bark is not so rich in tannin as that of typical A. decurrens? of which species A. dealbata is considered to be a variety by Maiden. A. dealbata was introduced from Tasmania about 1820; and is now much cultivated on the Riviera for its flowers, which are sent in large quantities to Paris and London, under the popular name of " Mimosa." In France this species flourishes on the west coast as far north as Nantes, where, however, it is killed to the ground in severe winters, but nevertheless sends up shoots afterwards with increased vigour.2 (A. H.) In England,3 it can be grown in the open air in the south-west; and has attained 50 ft. in height after seventeen years growth from seed at Abbotsbury, where it produces flowers annually in great abundance and good seed, from which plants have been raised. At Trebah in Cornwall there is a fine tree of about the same age.4 In Ireland, the finest we know is at Derreen (Plate 376), and is believed by the Marquess of Lansdowne to have been planted about thirty-two years. When I measured it in 1910, it was 48 ft. high, with four stems measuring 2 ft. i in. to 2 ft. 7 in. in girth. This species is now completely naturalized in the Nilgiris,8 where it is useful for firewood. Gamble says that it is readily reproduced by suckers and coppice shoots. In France it does not thrive on soils containing lime.6 (H. J. E.) 1 Typical A. decurrens has branchlets and foliage, which are nearly glabrous and not hoary ; and is known in Australia as Green or Black Wattle. A plant, imported from Johannesburg, has been growing since 1909 in the open air at Blackmoor, Liss, Hants, and is now about 4 ft. high. A. decutrens, var. mollis, Lindley, has tomentose foliage, but the pubescence assumes a golden yellow tinge on the branchlets. 2 Cf. Maiden, Forest Flora N.S. Wales, iii. 60 (1908). 3 A tree in the Temperate House at Kew is about 50 ft. high, but with a slender stem, about 6 in. in diameter ; the bark is broken on the surface into small scales. In Gard. Chron. lui. 45 (1913) mention is made of a tree, 70 ft. high and 2 ft. 2 in. in girth, growing in the conservatory at Branksome Hall, Darlington, which was raised from a root-cutting twenty-five years ago. 4 Card. Chron. Iii. 44 (1912). 6 Brandis, in Indian Forester, viii. 26 (1882), quotes General Morgan's account of the remarkable change in the period of flowering of this species in the Nilgiris, where it was introduced by seed in 1845. The trees here flowered at first in October, which is the month in which the parents flower in Australia. In 1860, they began to flower in September; in 1870, they flowered in August ; in 1878, in July ; and ill 1882, in June, which is the spring month in the Nilgiris corresponding with October in Australia. 0 Cf. Mottet, in Rev. Hort. 1896, p. 503. Acacia 1699 ACACIA MELANOXYLON, BLACKWOOD Acacia melanoxylon, R. Brown, in Alton, Hort. Kew. v. 462 (1813); Sims, Bot. Mag. t. 1659 (1814); Loudon, Art. et Frut. Brit. ii. 663 (1838); J. D. Hooker, fl. Tasm. i. 109 (1860); Bcntham, Fl. Austral, ii. 388 (1864), and in Trans. Linn. Soc. xxx. 481 (1875); Gamble, Indian Timbers, 301 (1902); Rodway, Tasm. Flora, 42 (1903); Maiden, Forest Flora N.S. Wales, ii. 103, plate 57 (190?)- t Acacia latifolia, Desfontaines, Table École Bot. Paris, 207 (1815). Acacia arcuata, Sieber, ex Sprengel, Syst. iii. 135 (1826). Acacia brevipes,1 Cunningham, in Bot. Mag. t. 3358 (1834). An evergreen tree, attaining occasionally, in Australia, 120 ft. in height, and 6 to IG ft. in girth ; but usually only about 80 ft. high. Young branchlets angled, minutely grey tomentose, or rarely glabrous. Stipules absent. Phyllodes cori aceous, glabrous, usually falcate, very variable in size, averaging z\ to 4 in. long and ^ to f in. wide ; lanceolate or oblong, gradually tapering to an obtuse or acute apex, which is tipped with a cartilaginous point ; much narrowed towards the base ; with three or four conspicuous longitudinal veins, connected by anastomosing veinlets ; margin entire, cartilaginous. True leaves often present erratically on young trees, bipinnate, with a tomentose rachis and secondary axes, the latter bearing oblong apiculate leaflets, about § to \ in. long. Flower heads, three to four in an axillary raceme, globose, yellow, about \ in. in diameter ; peduncles \ to \ in. long, glabrous ; flowers minute, thirty to fifty in a head, pentamerous, with coherent calyces and denticulate sepals ; petals free. Pods linear, flat, often curved in a circle, 2 to 4 in. long, about £ in. broad, with thickened cartilaginous margins, and glaucous coriaceous valves. Seed small, nearly orbicular, with a long dilated pale red funicle, encircling it in double folds. A. melanoxylon is known in Australia as blackwood, and is widely distributed in Tasmania, Victoria, and New South Wales, extending into South Australia and Queensland. It is most common on rich soil in valleys or grassy ranges; and ascends in the mountains to considerable elevations. It yields a timber of the highest class, which Gamble compares to light-coloured walnut. It is hard and close-grained, taking a fine polish; and is very beautiful when figured. It is used in Sydney and Melbourne for making billiard tables, furniture, gunstocks, coaches, and railway carriages ; and is imported by English pianoforte manufacturers.5 (A.H.) A. melanoxylon was introduced into England about 1808, and is occasionally grown out-of-doors in the south-west. At Abbotsbury, raised ten years ago from seed, it is about 35 ft. high ; and seedlings have been raised from it. At Tregothnan, I saw several trees flowering well in 1911, one of which was 35 ft. high by i ft. 8 in. in girth. In the south of France,3 it is the best of the acacias for avenues, as it forms a tall tree regular in habit ; and at Hyères there are numerous natural seedlings. It was introduced in 1840 into the Nilgiris in India, where it is completely naturalised. (H. J. E.) 1 This is a variety, which appeared in cultivation at Kew, with longer and more falcate phyllodes, attaining 5 to 7 inches in length. 2 Cf. Penny, Tasm. Forestry, 9 (1905). 3 Pettier, in Le Jardin, xxii. 75 (1908). LAURELIA Laurelia, Jussieu, in Ann. Jfus. Paris, xiv. 134 (1809) ; Tulasne, in Archiv. Mus. Hist. Nat. Pans, via. 414 (1855); A. de Candolle, Prod. xvi. pt. 2, p. 674 (1868); Bentham et Hooker, Gen. PI. iii. 145 (1880) ; Perkins and Gilg, in Engler, Pßanzenreich, iv. loi, Monimiaceœ, 76 (1901); Perkins, in Engler, op. cit., SuppL 46 (1911). Pavonia, Ruiz and Pavon, ft. Peruv. et Chil. Prod. 127, t. 28 (1794) (not Cavanilles). Theyga, Molina, Sag. Chile, 163 (i8ro). Thiga, Molina, Sag. Chile, 297 (i8ro). EVERGREEN trees, belonging to the order Monimiacese ; with opposite coriaceous fragrant serrate penninerved leaves, which are without stipules. Flowers polygamous or dioecious ; in simple or panicled axillary cymes ; perianth with six to twelve spreading lobes in two or three series. Male flowers, with a flat receptacle, and five to twenty stamens ; filaments short, each with two lateral glands at the base ; anthers with two oblong lateral cells, which dehisce by valves opening upwards. Female and perfect flowers, with a receptacle at first cup-shaped, bearing stamens, which are often reduced wholly or in part to stamin- odes, and numerous fusiform villous ovaries, which are tipped with long hirsute styles, and are one-celled, containing an erect anatropous ovule. Fruit, consisting of the enlarged and almost closed receptacle, which has become globose, ovoid, or tubular, and ultimately opens by splitting irregularly into three or four valves ; on these are placed the ripened ovaries or achenes, which are ovoid, pilose, and end in a long plumose unbranched style. The valves of the receptacle remain firmly closed in moist air, but spread widely when dry. Three species of Laurelia are known, two of which, not being in cultivation, need only be briefly mentioned :— 1. Laurelia Novœ-Zelandiœ, Cunningham, in Ann. Nat. Hist. i. 381 (1838). A tree attaining 150 ft. in height in New Zealand. 2. Laurelia sempervirens, Tulasne, in Arch. Mus. Hist. Nat. viii. 416 (1855). Laurelia aromatica, Poiret, in Lamarck, Encycl. Meth. Suppl. iii. 313 (1813). Pawnia sempervirens, Ruiz and Pavon, Syst. Veg. PI. Peruv. et Chil. 253 (1798). This species, with which L. serrata has been much confused, has a more northerly distribution, occurring in Peru, as well as throughout Chile, where it is called laurel by the inhabitants. It differs from L. serrata1 in having the leaves un dulate serrate, with rather distant appressed obtuse teeth ; flowers in loose panicles with long pedicels ; fruit receptacles fusiform or ovoid ; seed plumose throughout, even to the tip of the style. It is a tall tree, very abundant in the evergreen forests, and yielding wood that is easily worked and much used.2 (A. H.) 1 L. serrata has sharply serrate leaves ; flowers in short crowded panicles, with short pedicels ; fruit receptacles globose ; seed with the style not plumose at the tip. y Cf. Castello and Dey, Jeog. Vej. Rio Valdivia, 51 (1908). I7OO Laurelia 1701 LAURELIA SERRATA Laurelia serrata?* Philippi, in Bot. Zeit. xv. 401 (1857); Castillo and Dey, Jeog. Vej. Rio Valdivia, 52, fig. 27 (1908); Stapf, in Bat. Mag. t. 8279 (1909); Perkins, in Engler, Pflanzenreich, iv. ioi, Monimiaceœ, Suppl. 47 (1911). Laurelia aromatica, Masters, in Gard. Chron. xxxvi. 401, fig. 172 (1904) (not Poiret). A tree, attaining in Chile about 70 ft. in height. Bark greyish, smooth, with persistent brown lenticels, and resembling that of Zelkova crenata. Young branch- lets, with two pubescent furrows, which have slightly projecting margins. Leaves coriaceous, in opposite pairs, narrowly elliptic or broadly lanceolate, averaging 4 to 5 in. long, and \\ to if in. wide ; tapering to an acuminate apex, usually tipped with a cartilaginous point ; cuneate at the base ; lateral nerves, ten to twelve pairs, dividing and looping before reaching the margin, which is entire towards the base, but is elsewhere serrate, with sharp teeth, directed towards the apex of the leaf or slightly spreading and tipped with a glandular thickening ; both surfaces glabrous, except on the slightly pubescent prominent midrib above ; petiole ^ to \ in. long, blackish, densely pubescent on the upper surface. Flowers, three to nine, in simple or panicled silky cymes ; pedicels about ^ in. long. Receptacle in the fruiting stage, globose but constricted at the apex, ultimately splitting irregularly, ashy-grey externally ; achenes ovoid, nearly \ in. long, densely hirsute with spreading hairs, except at the tip of the long and per sistent style. L. serrata is a native of the evergreen forests of southern Chile and northern Patagonia, occurring from Valdivia to the valley of the river Aysen. It is called huan-huan by the inhabitants, and seems to be a smaller and rarer tree than L. sempervirens ; and is said by Castillo and Dey to have an inferior wood, which has an unsupportable odour when freshly cut, yet is much used. L. serrata was probably introduced about 1860 by Pearce, who collected in Chile for Messrs. Veitch: It is one of the rarest trees in cultivation, and can only be propagated by seed or by layers, cuttings having always failed to take root at Kew. (A. H.) The only specimens which we have seen are the following :—At Penjerrick, near Falmouth, the seat of R. Fox, Esq., there is a splendid tree in perfect health and looking as if it would become considerably larger, which was in flower in April, 1911, when it measured 47 ft. in height and 3 ft. 4 in. in girth. Another at Kilmacurragh, Co. Wicklow, is about 30 ft. by 3 ft. ; but is not so handsome. This tree was procured by the late Mr. Acton from Messrs. Rollison of Tooting about 1868; and from it were drawn the figures cited above in the Botanical Magazine and in the Gardeners Chronicle. (H. J. E.) 1 Laurelia serrata, Bertero, in Mercur. Chil. 1829, manip. 15, p. 685, translated in Amer. Journ. Sei. xxiii. 89 (1833), is a name without any description. VII H ILEX Ilex, Linnseus, Sp. PI. 125 (1753); Bentham et Hooker, Gen. PI. i. 356 (1862); Loesener, in Nova Acta Ac. Leap. Carol. Ixxviii. 8 (1901); Schneider, in Gartenflora, Iii. 452 (1903), and Laubhohkunde, ii. 157 (1907). Prinos, Linnseus, Gen. PI. 153 (1754). Aquifolium, Haller, Enum. Stirp. Heh<. i. 296 (1758). TREES or shrubs belonging to the order Aquifoliacese, mostly evergreen, rarely deciduous. Leaves alternate, simple, usually short-stalked ; margin entire, crenate, serrulate, or with spiny teeth ; stipules minute, deltoid, often deciduous. Flowers axillary, solitary or cymose, normally dioecious, regular, usually tetra- merous, rarely pentamerous or hexamerous ; calyx gamosepalous, hypogynous, with a short tube and four to six lobes, imbricate in the bud ; corolla rotate, with four to six petals, free or connate at the base ; disc absent. Male flowers, with four to six stamens, alternate with the petals and adhering to them at the base ; anthers introrse, with two cells opening longitudinally ; ovary rudimentary, either without cells or with empty cells containing no ovules, stigma absent. Female flowers, with four to six stamin- odes, like the stamens, but sterile and small ; ovary free, superior, with usually two to four (rarely nine to twenty-two) cells, each usually containing one pendulous ovule ; stigma sessile, with as many lobes as there are cells in the ovary. Fruit drupaceous, with the calyx and stigma persistent ; epicarp fleshy, containing two to four (or more) one-seeded nutlets. Seed pendulous, with a minute embryo at the apex of the copious albumen. The genus comprises about 270 species, which are distributed throughout the greater part of the tropical and temperate regions of the world. About twenty-five exotic species are cultivated in Britain, either shrubs or small trees, not coming within the scope of our work ; but of these the following four species will be briefly described on account of their relationship to the common holly. I. Ilex Perado, Aiton, Hort. Kew. i. 169 (1789); Loudon, Arb. et Frut. Brit. ii. 519(1838)- Ilex madertensis, Lamarck, Encyc. iii. 146 (1789). A small tree, attaining 30 ft. in height. Young branchlets glabrous. Leaves obovate-oblong, yellowish green, z\ to 3 in. long, \\ to 2 in. broad, flat on the surface, thick and coriaceous ; apex rounded or minutely cuspidate, and tipped with a slender spine ; base rounded or cuneate, decurrent on the glabrous petiole, which has two longitudinal depressions on the lower side ; margin not undulate, entire or with a few minute serrations, mostly towards the apex, each tipped with a slender 1702 Ilex 1703 spine ; under surface reticulate. Flowers reddish ; fruit ellipsoid, reddish, with nutlets similar to those of /. Aquifolium. I. Perado, in its typical form, is confined * to the Madeira Islands, whence it was introduced into England in 1760. It has usually been cultivated since in green houses ; but Loudon states2 that in the Chiswick Garden and several other places near London it bore uninjured the winter of 1837-1838 without protection. It has, however, when tried out-of-doors at Coombe Wood, been always killed in ordinary winters. It is perfectly hardy in Wicklow, where at Kilmacurragh a healthy shrub was bearing fruit in July 1908; and at Powerscourt, where a fine specimen is about 20 ft. high, with a diameter of spreading branches of 24 ft. It is doubtful if this species has been one of the parents of any of the hybrid hollies.8 II. Ilex platyphylla, Webb and Berthelot, Phyt. Canar. ii. 135.1. 68 (1836); W. J. Hooker, Bot. Mag. t. 4079 (1844). A tree, attaining 40 ft. in height. Young branchlets minutely pubescent above the insertions of the leaves. Leaves thick and coriaceous, yellowish green, much larger as a rule than in I. Perado, 4 to 6 in. long, 2 to 4 in. broad, ovate-oblong, flat on the surface ; base rounded or cuneate, decurrent on the glabrous petiole, which shows two longitudinal furrows on the lower side ; apex shortly acuminate and tipped with a slender spine ; margin not undulate, entire or with few or numerous minute serrations ending in slender spines ; reticulate beneath. Flowers white ; fruit sub- globose, reddish or blackish, with larger nutlets than in /. Aquifolium. This is common in the Canary Islands, and also occurs in the Madeiras, in the latter case associated with 7. Perado. From the latter species it differs mainly in the larger acuminate leaves. It does not seem ever to have been in cultivation4 in England except at Kew6 under glass ; and there is no evidence that it was concerned in the origin of any of the hybrid hollies. III. Ilex balearica, Desfontaines, Hist. Arb. ii. 362 (1809). Ilex maderensis, Willdenow, Enum. PL Hort. Berol. Suppl. 8 (1813) (not Lamarck); Moore, in Gard. Chron. ii. 751 (1874). Ilex Aquifolium, Linnseus, var. lalearica, Lamarck, Encycl. iii. 145 (1789); Loudon, Arb. et Frut. Brit. ii. 516 (1838). A small tree, attaining 30 or 40 ft. in height. Young branchlets stout, greenish, densely covered with a minute pubescence. Leaves thick and coriaceous, concave on the upper surface, 3 in. long, 2 in. broad, ovate, shortly acuminate at the apex, which ends in a slender spine ; margin not undulate, either entire or with a few (three to ten) irregularly placed serrations, which end in slender spines ; base of the blade rounded or cuneate, scarcely decurrent on the minutely pubescent petiole, 1 /. Perade, var. azorica, Loesener, in Nova Acta Acad. Leap. Carol. Ixxviii. 247 (1901), which is the holly of the Azores, differs from the typical form of the Madeiras in having smaller leaves. » Trees and Shrubs, 161 (1842). Loudon, Gard. Mag. xiv. 226 (1838), records a shrub of /. Perado at Ilendon Rectory which was 6\ ft. high. 3 Cf., however, p. 1712, note I. 4 Plants commonly cultivated under this name are, in my opinion, either /. balearica or one of its hybrids with /. Aquifolium. 6 A small specimen now growing in the Temperate House at Kew was sent from the Canaries by Dr. Ferez. 1704 The Trees of Great Britain and Ireland which is convex on the lower side. Fruit and nutlets slightly larger than in /. Aqui- folium. This species occurs in southern Spain and in the Balearic Islands.1 It is closely allied to /. Perado, but differs in having dense pubescence on the branchlets, and the base of the leaf is not decurrent on the petiole. It is very distinct from the common holly, which has leaves with very undulate margins and large sinuate teeth. /. balearica is said to have been introduced into England in 1744, and was certainly cultivated in the Royal Garden at Versailles in 1789. It is usually pro pagated by budding or grafting upon the common holly, and is perfectly hardy at Kew, Cambridge, and Paris, but requires protection during winter in Germany. This species regularly produces flowers2 and fruit in England; and appears to have given rise, in conjunction with the common holly, to a series of hybrids, which began to be noticed about 1800, though they were considered at the time to be simply varieties of the latter species (see p. 1712). IV. Ilex opaca, Aiton, Hort. Kew. i. 169 (1789). A tree, attaining in America 50 ft. in height and 12 ft. in girth. Young branchlets minutely pubescent, becoming pale brown in their first year. Leaves elliptic, about 3 in. long and i^ in. broad, convex and dull green above, concave and conspicuously reticulate beneath ; margin with a few irregular spreading spiny teeth. Fruit dull red, £ in. in diameter ; nutlets four. This species, which somewhat resembles the common holly in foliage, is readily distinguishable by the brown branchlets ; and is a native of the United States from Massachusetts to Florida, and westward to Indiana and Texas. It was introduced in 1744 into England, and forms a small tree in our climate, which produces flowers and fruit regularly. Though often grown with the common holly in nurseries and botanic gardens, there is no evidence that this species has taken part in the origin of the hybrid hollies. (A. H.) ILEX AQUIFOLIUM, COMMON HOLLY Ilex Aquifolium, Linnaeus, Sp. PI. 125 (1753); Loudon, Arb. et Frut. Brit. ii. 505 (1838); Willkomm, Forstl. Flora, 786 (1887); Mathieu, FloreForestûre, 58 (1897) ; Loesener, in Nova Ada Ac. Leap. Carol. Ixxviii. 248 (1901); Schneider, Laubholzkunde, ii. 163 (1907). A tree, occasionally attaining 50 to 70 ft. in height, and 9 to 12 feet in girth, often shrubby and then rarely over 30 ft. high. Bark smooth, greyish. Young branchlets green or purplish, minutely pubescent. Buds minute, with two acuminate outer scales. Stipules two, usually persistent as withered minute scales at the base of the petiole. Leaves persistent about fourteen months, coriaceous, thick, ovate, 1 It is probably much more widely spread in the Mediterranean region, but I have been unable to study its distribution. /. Aquifolium, var. platyphylloides, Christ, in Bar. Schweiz. Bot. Ges. xiii. (1903), a tree 30 ft. high occurring on rocky cliffs on the west side of Lake Maggiore in Italy, is probably a form of I. balearica. 8 In nurseries the male plant of /. balearica is sometimes called /. maderensis, following the erroneous view of Moore, in Gard. Chron. ii. 751 (1874), that it came from Madeira, while the name /. balearica is wrongly restricted to the female plant. /. Aquifolium, var. platyphylla of most nurseries, appears to differ in no respect from /. balearica ; but see p. 1714. Ilex 1705 oblong, or elliptic, 2^ to 4 in. long, i to 2 in. broad ; acute or acuminate at the apex, usually broad and rounded at the base ; dark green and shining above, paler and duller beneath ; margin waved, encircled by a cartilaginous rim, usually with sinuate teeth, each ending in a sharp spine, which spread in different planes ; but on the upper branches1 of old trees entire, or with one or two spinescent teeth towards the apex ; primary veins pinnate, six to eight pairs, dividing and looping before reaching the margin ; secondary and tertiary veins indistinct ; petiole glabrous, short, rounded and not sulcate on the lower side, the base of the blade not being decurrent upon it. Flowers small, in short axillary cymose fascicles, normally dioecious,2 rarely polygamous; sepals four, greenish; petals four, white, placed crossways, slightly connate at the base ; stamens four ; ovary four-celled, one ovule in each cell. Fruit globose, red, about ^ in. in diameter, crowned by the four-lobed stigma, usually containing four nutlets. The holly is usually bisexual, all the flowers on a tree being either exclusively staminate or pistillate; but, in rare cases, a few perfect flowers, containing both good pollen and well-formed ovules, are produced in addition.3 Dallimore states that female trees, which are isolated, often bear large quantities of berries ; but in such cases most of the seed is infertile, and there appears to be no doubt that pollination is always effected by the pollen being brought from another tree by either insects or the wind. The seeds when sown do not germinate for a long time, two or even three years elapsing before the appearance of the seedlings.4 These have two ovate entire obtuse shortly stalked cotyledons, about f in. long, which are raised above ground on a glabrous caulicle, about an inch in length. The glabrous angled stem gives off in the first year three to five simple alternate leaves, similar to the adult leaves in form, but much smaller, and with small sinuate spiny teeth. These leaves have minute ovate acute black stipules, which are soon deciduous. VARIETIES The common holly is variable in the wild state, there being two forms,5 one with green and the other with purple branchlets ; moreover, differences in habit, in the spinescence of the foliage, and in the colour of the fruit, are not uncommon. The peculiar geographical forms,6 which occur wild in western Asia and in China, are not in cultivation. The horticultural varieties, that are usually given in nursery 1 Kerner, Not. Hist. Plants, Eng. Trans. 433 (1898), believes lhat the spiny leaves of holly are an adaptation against browsing by ruminants ; and points to the fact that on adult trees leaves without spines are only produced on branches beyond the reach of these animals. Withering, Arrange. Biit. Plants, ii. 211, note (1796), long ago noticed the same fact, which was also referred to by the poet Southey. Mr. R. A. Phillips tells me that spineless leaves always commence about half-way up wild holly trees in Ireland ; and Sir Herbert Maxwell mentions the same fact about wild holly in the wood between Murthly and Dunkeld. v Darwin, Forms of Flowers, 297 (1892) states that the stamens in the female flowers, though quite destitute of pollen, are but slighlly shorter than the perfect stamens in the male flowers. The male trees produce a greater number of flowers, and these have smaller corollas than occur in the other sex. 3 Masters, in Gard. Chron. xxiii. 27, fig. 8 (1885) and iv- 35^ (1888) describes polygamous flowers; and states that occasionally a tree which has hitherto only borne staminate flowers becomes covered with berries. « Cf. Lubbock, Seedlings, i. 337, fig. 240 (1892). 6 Cf. Bromfield, in Phytologist, iii. 536 (1849). 0 Vars. caspia and chinensis, Loesener, in Nova Ada Ac. Leap. Carol. Ixxviii. 262, 263 (1901). 1706 The Trees of Great Britain and Ireland catalogues, comprise both the hybrids, which we keep distinct as being of mixed origin ; and the true varieties, due only to the common holly, which in many cases have arisen as branch sports, and been subsequently propagated by grafting. The latter usually revert, when old, in isolated branches on the tree, to the type of the common holly. I. DIFFERING FROM THE TYPE IN HABIT. 1. Var. pendula, Loudon, Trees and Shrubs, 1113 (1842). Branches pendulous ; leaves as in the common holly. The original tree, from which this variety was propagated by Barren at Elvaston Castle, formerly existed in a private garden in Derby. Loudon also mentions* another weeping form, which was discovered about 1842 in Dalkeith Park. There are several good specimens of the weeping holly at Kew, which are clothed to the ground, and somewhat resemble the pendulous form of Sophora japonica. Dallimore states that as no leader is formed, it is necessary to keep a shoot tied up, in order to obtain a tall plant. There are two variegated pendulous forms—var. aurea pendula, Waterer's weeping holly, and var. argentea marginata pendula, Perry's weeping holly. 2. Var. fastigiata, Loudon, Gard. Mag. xix. 442 (1843). Branches erect. Loudon mentions two specimens, one in a garden in Edin burgh, and another in a garden in Derby. The fastigiate holly is rare, but is represented at Kew by a small shrub. II. DIFFERING FROM THE TYPE IN THE COLOUR OF THE FRUIT. 3. Vzx.fmctu luteo, Loudon, Arb. et Frut. Brit. ii. 509 (1838). Vars. chrysocarpa and xanthocarpa, Koch, Dendrologie, ii. pt. i. 210, 216 (1872). Berries yellow. This is one of the oldest known varieties.2 Seedlings have been raised in Waterer's and Paul's nurseries, which have produced orange-coloured fruit, and are supposed to have originated from crossing between the yellow and red-fruited forms.8 Loudon mentions a variety4 with white berries, and another variety6 with black berries, neither of which we have seen. III. DIFFERING FROM THE TYPE IN FOLIAGE. A. Leaves, with both marginal and superficial spines. 4. Var. echinata, De Candolle, Prod. ii. 14 (1825). Var. ferox, Alton, Hort. Keiv. i. 169 (1789); Loudon, Arb. et Frut. Brit. ii. 507 (1838). Ilex echinata, Miller, Gard. Did. ed. 8, No. 2 (1768). Leaves ovate, about 2 in. long, with their edges rolled backwards ; upper surface covered more or less with sharp prickles ; margin with irregular large spines. This, which is called the Hedgehog Holly, is the oldest known variety, as it was mentioned by Parkinson6 in 1640, and was cultivated at Fulham about 1700. It occurs now only as a male plant ; but Parkinson and Martyn refer to it as bearing berries, so that a female form was in all probability formerly in cultivation. Dalli- 1 Loudon, Card. Mag. xix. 442 (1843). 2 In Cole's History of Plants, published in 1657, the yellow-berried holly is mentioned as having been found wild near Wardour Castle. 3 Moore, in Gard. Chron. ii. 520 (1874). 4 Koch, Dendrologie, ii. part i. 212 (1872) calls this var. letaocarfa, and speaks of it as common. 6 Cf. Arb. et Frut. Brit. iv. 2545 (1838). « rheat Bet I4g6 (1640). Ilex 1707 more* considers this variety to have originated from the typical form directly as a branch sport, which was subsequently propagated by grafting, and which by further sports has given rise to other varieties, like crispa. Two variegated forms of this variety are known :— Var. ferox argentea, Loudon, op. cit. 509. Spines and margin cream-coloured ; known as the Silver-Striped Hedgehog Holly. Var. ferox aurea, Loudon, op. cit. 509. Margin green ; centre of the blade with a triangular yellow blotch ; known as the Gold-blotched Hedgehog Holly. B. Leaves deformed ; marginal spines abnormal. 5. Var. latispina, Goeppert, in Gartenflora, Hi. 318 (1854) ; Moore, in Gard. Chron. ii. 812, fig. 164 (1874). Leaves thick in texture, broadly oval or quadrangular, with an acuminate apex ending in a long decurved spine; averaging 2 in. long and \\ in. broad; margin with a few coarse spines, variable in number and direction. Dallimore considers that this variety originated probably as a branch sport from var. crispa. 6. Var. trapeziformis, Moore, in Gard. Chron. ii. 812, fig. 164 (1874). This variety mainly differs from the last-mentioned in the slightly smaller leaves, which have the margin entire or with only one or two erratically placed spines. 7. Var. monstrosa, Goeppert, in Gartenflora, iii. 318 (1854); Moore, in Gard. Chron. ii. 750, fig. 147 (1874). Leaves ovate-oblong, 3^ in. long, i to i^ in. broad (exclusive of the spines), resembling var. latispina in the long acuminate apex with a deflexed spine, but with a longer blade and more numerous marginal spines, which project in all directions. Dallimore states that this originated in the Handsworth Nurseries. 8. Var. albo-picta, Loudon, Arb. et Frut. Brit. ii. 508 (1838). Var. argentea medio-picta, Moore, in Gard. Chron. iv. 688 (1875). Leaves similar to var. monstrosa, but slightly smaller and variegated; the centre of the blade having an irregular blotch of creamy white, the margin remaining green. This variegated holly is usually called " milkmaid " or " silver milkmaid," and has been known from an early period. 9. Var. hastata, Moore, in Gard. Chron. ii. 687, fig. 138 (1874). Var. kewensis, Loesener, in Nova Ada Ac. Leap. Carol. Ixxviii. 266 (i901)- Leaves dark green, coriaceous, f to i^ in. long, with two or three pairs of long marginal spines towards the base ; the upper half of the blade forming a large entire acuminate or emarginate triangular lobe. This is a dwarf shrub, with purple branches, which originated in the Handsworth Nurseries. 10. Var. Beetii, Moore, in Gard. Chron. ii. 520, fig. 107 (1874). Leaves nearly orbicular, about \\ in. in diameter, dark green, with a thickened margin, and a few (usually five or six) large spines, pointing in various directions. This originated in the Handsworth Nurseries. 1 Holly, Yew, and Box, 68, 69 (1908). 1708 The Trees of Great Britain and Ireland 11. Var. crassifolia, Aiton, Hort. Kew. i. 169 (1789) ; Loudon, Arb. et Frut. Brit. ii. 508 (1838); Moore, in Gard. Chron. ii. 752, fig. 152 (1874). Leaves very thick in texture, lanceolate-oblong, if to 2 in. long, with the apex recurved ; margin with triangular teeth, ending in coarse points. This, which is called the Leather-leaf or Saw-leaved Holly, is one of the oldest known varieties,1 and is remarkably constant in character ; no case of a branch reverting to the common holly having been noticed.2 It is shrubby with purple branchlets ; and always bears pistillate flowers, which ripen into peculiarly flattened fruit. C. Leaves small, less than 2 in. long, with regular marginal spines. 12. Var. lineata, Moore, in Gard. Chron. ii. 752 (1874). The form of the common holly with the smallest leaves, which are narrowly lanceolate, f in. long, flat, with four or five minute spines on each side. This is a small compact bush with green branchlets. Var. microphylla, Moore, loc. cit., appears to differ only in having purple branchlets. 13. Var. recurva, Aiton, Hort. Kew. i. 169 (1789) : Loudon, Arb. et Frut. Brit. ii. 507 (1838): Moore in Gard. Chron. ii. 687, fig. 138 (1874). Leaves ovate, acuminate, ending in a long spine, about i^ in long, curved from base to apex, and usually twisted to one side above the middle ; spines divaricate and resembling those of the typical form. This is a dense shrub, and bears staminate flowers. 14. Var. serratifolia, Loudon, Arb. et Frut. Brit. ii. 507 (1838); Moore, in Gard. Chron. ii. 687, fig. 138 (1874). Var. myrtifolia, Goeppert, in Gartenflora, iii. 320 (1854); Moore, in Gard. Chron. ii. 687, fig. 138 (1874). Leaves ovate-lanceolate, \\ in. long, ^ to f in. broad, with divaricate spines similar to those of the type. This forms a dense pyramidal shrub. Several variegated small-leaved forms are known, which may be mentioned here :— Var. myrtifolia aureo-maculata ; centre of the leaf with deep yellow blotches. Var. myrtifolia aureo-marginata ; centre of the leaf mottled, margin pale yellow. Var. myrtifolia elegans ; leaf with a green centre and a narrow golden edge ; and Var. Ingrami ; leaf mottled with irregular white streaks, margin pinkish. 15. Var. angustifolia? Loudon, Arb. et Frut. Brit. ii. 507 (1838); Moore, in Gard. Chron. ii. 752, fig. 154 (1874). Leaves lanceolate, about i^ in. long and \ in. broad, often ending in a long acuminate entire apex ; spines numerous, regular, slender. This is a female tree, and is very distinct in its narrow erect almost fastigiate habit, and its elegant small leaves. At Kew, Tortworth, and elsewhere, there are specimens 30 to 40 ft. high, 1 First described by Hanbury, Complete Book of Gardening (1770), as the saw-leaved holly. 2 Cf. Baltimore, Holly, Yew, and Box, 67, 90(1908). 3 Var. angustifolia, Hohenacker, in Bull. Soc. Nat. Mosc. iii. 319 (1838), was found wild in woods near Lenkoran, on the south-west coast of the Caspian Sea ; and is a form of var. caspia, Loesener, differing from the horticultural variety here described. Ilex 1709 as straight as a lance, and always very pointed at the top. It bears small, but not numerous berries. Variegated forms of this are known :— Var. angustifolia albo-marginata ; a small shrub ; leaves as in the green variety, but creamy white on the margin ; and Var. angustifolia aureo-macidata, centre of the leaf unevenly marked with yellow. 16. Var. handsworthensis, Moore, in Gard. Chron. ii. 520, fig. 108 (1874). Leaves ovate, acuminate, i|- in. long, f in. broad, with numerous regular strong spines, pointing towards the apex, the margin only slightly undulate. This has neat and elegant foliage, and originated in the Handsworth Nurseries. 17. Var. ciliata, Loudon, Arb. et Frut. Brit. ii. 507 (1838); Moore, in Gard. Chron. ii. 752, fig. 153 (1874). Leaves ovate-lanceolate or ovate, \\ to 2 in. long, \ to i in. wide ; margin scarcely undulate, with regular slender spines. This is a male tree at Kew, with purple branchlets and dark foliage. 18. Var. ouata, Moore, in Gard. Chron. ii. 752, fig. 149 (1874). Ilex ovata, Goeppert, in Gartenflora, iii. 321 (1854) ; Koch, Dendrologia, ii. pt. i. 216 (1872). Leaves ovate, with a broad truncate or rounded base, i^ to 2 in. long, i^- in. broad ; margin slightly undulate with small sinuate teeth, tipped with minute spines. This remarkably distinct-looking holly is only known in the staminate form, and has purple branchlets. It is slow in growth. It is undoubtedly a sport of the common holly, as branches with ordinary leaves of the latter are often present on old specimens. D. Leaves 2 to 4 in. long, with regitlar marginal spines. 19. Var. Foxii, Moore, in Gard. Chron. ii. 752, fig. 150 (1874). Leaves broadly ovate, 2 to 2^ in. long, i^ to i\ in. broad, with distant sinuate spinous teeth of moderate size on the slightly undulate margin. This variety, which is only known in the staminate form, is very distinct in appearance. 20. Var. whittingtonensis, Moore, in Gard. Chron. ii. 687, fig. 138 (1874). Leaves lanceolate or narrowly elliptic, 2^ to 3 in. long, i in. broad, dense shining green on the upper surface ; margin undulate with numerous stiff divaricate spines. This is a distinct variety, bearing staminate flowers. 21. Var. Fisheri, Moore, in Gard. Chron. ii. 520, fig. 105 (1874). Leaves ovate, with a broad base and an entire triangular apex, 3 in. long, i^ to 2 in. broad ; dark shining green above ; margin slightly undulate, with large sinuate spiny teeth, variable in number, one to eight on each side. This is a vigorous handsome tree, which originated in the Handsworth Nurseries. It has green branchlets and is a male. 22. Var. aurea regina, Moore, in Gard. Chron. v. 44 (1876). Leaves similar in shape to var. Fisheri, but variegated ; margin broad, well- defined, deep golden yellow ; centre of the blade mottled with different shades of green. This beautiful holly is usually called " Golden Queen," and bears staminate flowers. vli 1 17TO The Trees of Great Britain and Ireland 23. Var. Lichtenthalii, Simon-Louis, Cat. 1880-1881, p. 50. Leaves oblong or narrowly elliptic, 4 in. long, 2 in. broad ; dark dull green above ; midrib beneath yellow ; margin slightly undulate, and with large regular sinuate spiny teeth. This is a distinct large-leaved form, with purplish branches, and is possibly of hybrid origin. 24. Var. alcicornis, Moore, in Gard. Chron. ii. 433, fig. 90 (1874). Var. Robinsoniana, Dallimore, Holly, Yew, and Box, 76 (1908). Leaves narrowly elliptic, mostly with an entire cuneate base, about 3 in. long and i in. broad (exclusive of the spines) ; margin undulate, with numerous large spiny teeth, \ in. or more long, variously directed. This variety, which is distinct in appearance, has green branchlets, and is a free grower. It was originally sent out by Lawson. E. Leaves variable in margin, some qidte entire, others with spiny teeth. 25. Var. donningtonensis, Moore, in Gard. Chron. ii. 687, fig. 138 (1874). Leaves lanceolate, i^ to 2 in. long, ^ in. broad, very variable, either entire in margin or with a few irregular divaricate spines ; usually recognisable by some of the non-spiny leaves bearing one or two peculiar curved lobes at the base. This variety, which has dark purple branchlets, always bears staminate flowers; and originated in the Handsworth Nurseries. 26. Var. Smithiana, Moore, in Gard. Chron. ii. 520, fig. 106 (1874). Leaves lanceolate or elliptic, 2 to 2\ in. long, f in. wide, similar to the last in having leaves both entire and with a few irregular spines, but considerably larger and not showing the peculiar lobes at the base. This always has staminate flowers. 27. Var. heterophylla, Alton, Hort. Kew. i. 169 (1789); Loudon, Arb. et Frut. Brit. ii. 506 (1838). This name was originally given to the wild form of the common holly, in which spiny leaves occur near the ground and entire leaves in the upper branches of the tree ; but it is now often applied to the next variety, and to intermediate forms. Two variegated varieties occur :— Var. heterophylla aiireo-marginata ; leaves of two kinds, entire and spiny, with an irregular golden margin ; and var. heterophylla aiireo-picta, leaves mostly entire and blotched with yellow in the centre. 28. Var. integrifolia, Goeppert, in Garteitflora, iii. 320 (1854); Moore, in Gard. Chron. ii. 812, fig. 164 (1874). Leaves mostly ovate, entire; flat, but slightly twisted at the apex; about 2 in. long and i in. broad ; acute, acuminate, or rounded at the apex. Leaves bearing a few spines are usually present. This variety occurs in both sexes. 29. Var. Watereriana, Moore, in Gard. Chron. vi. 232 (1876). Leaves oval, i to if in. long, either entire or with a few irregular small spines ; edged with a broad irregular band of golden yellow. Waterer's Golden Holly is a staminate form with green branchlets, which are striped with yellow. It is a neat, dense, slow-growing variety. Ilex 1711 F. Leaves all (or nearly all) entire in margin ; spines absent, except at the apex, which usually terminates in a spiny point. 30. Var. laurifolia, Loudon, Arb. et Fmt. Brit. ii. 507 (1838) ; Moore, in Gard. Chron. ii. 812 (1874). Leaves narrowly elliptic or oblong-lanceolate, entire, flat, about 2 to 3 in. long, and i in. broad ; rarely a few leaves are present bearing one or two spines. This is known only as a male plant. Several variegated forms are known :— Var. laiirifolia sidphiirea ; mottled green in the centre, with a broad band of sulphur yellow along the margin. Var. laîirifolia aîirea, with a narrow bright golden yellow margin ; and Var. laiirifolia aureo-picta, blotched in the centre with deep golden yellow. 31. Var. marginata, Loudon, Arb. et Frut. Brit. ii. 507 (not Moore).1 Var. scotica, Moore, in Gard. Chron. ii. 812, fig. 164 (1874). Leaves obovate or oblong, rounded and spineless, or with a slight cuspidate spine at the apex, 2 to 2\ in. long, i to i^ in. broad ; margin entire, undulate, strongly thickened ; upper surface often marked near the apex with a cup-like depression. This peculiar holly, which is occasionally cultivated under the erroneous name /. Dahoon? is now generally named var. scotica, but seems to be without doubt the plant appropriately called var. marginata by Loudon. It is considered by Dallimore to have arisen as a branch sport from var. crispa, and often shows leaves approaching in character those of the latter variety; but in opposition to this view may be mentioned the fact that var. marginata is always female, while var. crispa is a staminate form. Two variegated forms occur, which are known as :— Var. scotica aureo-picta, a handsome variety which originated in the Cheshunt Nursery ; it has leaves variegated with golden yellow in the centre ; and Var. scotica aurea, a dwarf form, having leaves with a broad golden edge. 32. Var. crispa, Loudon, Arb. et Frut. Brit. ii. 507 (1838). Var. calamistrata, revoluta, and conforta, Goeppert, in Gartenflora, iii. 319 (1854). Var. tortuosa, Waterer, ex Moore, in Gard. Chron. ii. 812, fig. 164 (1874). Var. marginata, Moore, in Gard. Chron. ii. 812, fig. 164 (1874) (not Loudon). Leaves spirally twisted and variously folded, about 2 in. long and i in. broad, shining green ; margin much thickened, undulate, entire, or with one or two errati cally placed spines ; apex rounded or prolonged into a long stout decurved spine. This variety, which is called the Screw-leaved Holly, is always a male tree, and is supposed by Dallimore to be a branch sport of var. echinata, as it occasionally bears leaves like those of the latter in having spines on the upper surface. Var. crispa aureo-picta has leaves variegated with gold blotches. It often produces branches with green leaves, and occasionally has leaves with superficial spines. 1 Var. marginata, Moore, k var. crispa, Loudon. 2 /. Dalioon, Walter, is a native of the southern United States, and is apparently now not in cultivation. It was killed at Kew in the great frost of 1895. Cf. Keui Bull. 1896, p. 10. 1712, The Trees of Great Britain and Ireland G. Variegated Hollies. Many of the variegated hollies have been mentioned above, where they are placed with the green varieties, which they resemble. Of these the most useful for ornament are var. Watereriana (see No. 29), var. aurea regina (see No. 22), and var. argentea marginata pendula (see No. i). A few remain to be noticed. 33. Var. aureo-marginata, Loudon, Arb. et Frut. Brit. ii. 508 (1838). Leaves like the common holly, but yellow in margin, with the centre of the blade showing various shades of green. This includes a considerable number of sub-varieties, the most noteworthy of which are var. aureo-marginata angustifolia and var. aureo- marginata bromeliœfolia. 34. Var. albo-marginata, Loudon, Arb. et Frut. Brit. ii. 508 (1838). Leaves like the common holly, but with a silvery white or cream-coloured margin. Of the numerous sub-varieties, the best are var. argentea regina, " Silver Queen," and var. handsworthensis argenteo-variegata, " Handsworth New Silver Holly." 35. Var. aureo-picta, Loudon, Arb. et Frut. Brit. ii. 509 (1838). This variety, which is usually known as var. aurea medio-picta, or " Gold Milkmaid," has leaves like the common holly in shape, but.with their centre irregularly marked with a large golden yellow blotch. 36. Var. lutescens, Petzold and Kirchner, Arb. Muse. 350 (1864). Var. flavescens, Moore, in Gard. Chron. vi. 616 (1876). Leaves like the common holly in size and shape, but differing in having a soft yellow tinge when young, which usually lasts throughout the season, but is best marked on the side of the tree which is most exposed to the light. This beautiful variety is known as Moonlight Holly, the dark central mass of green foliage with light yellow terminal shoots giving the effect of a shrub seen by moonlight. HYBRIDS The following, which are usually considered to be varieties of the common holly, are of hybrid origin, the parents being probably /. balearica and /. Aquifolium? These hybrids are not known to produce branches which revert to the type of the common holly. They are vigorous trees, characterised by large leaves, with the margins flat or much less undulate than in the case of /. Aquifolitim, the spinous teeth being also less sinuate, in these characters approaching /. balearica. I. Leaves conspicuously reticulate beneath. i. Ilex Wilsoni, Fisher, ex Proc. Hort. Soc. 1899, p. cxix. ; Baltimore, Holly, Yew, and Box, 143 (1908). Ilex Aquifolium, var. latifolia, Loudon, Arb. et Frut. Brit. ii. 507 (1838). Ilex Aquifolium, var. princtps, Moore, in Gard. Chron. xiii. 45, fig. 8 (1880). 1 Chambers, Vestiges of Creation, 310(1851), quotes from the Gardener's and Farmer's Journal, 1848, p. 164:—"The following was related to us by Mr. M'JSfab (of the Edinburgh Botanic Garden) : he had sown the seeds of/, balearica, from which he had produced the common holly. He had also raised from the seeds of the tender Madeira holly (/. Perado) a variety identical with that known as Hodgins's holly ; and although the offspring of a tender parent, yet like Ilodgins's variety, it was also quite hardy." From this it would appear that /. Perado may have taken part in the origin of some of the hybrid hollies. Ilex ï 7 ! 3 Leaves broadly ovate or oval, 3 to 5 in. long, 2\ in. wide, flat or slightly concave and shining above ; conspicuously reticulate beneath ; margin with numerous regular spiny teeth directed towards the apex, \ in. long, and lying in the plane of the blade. This is one of the most ornamental hollies, and is a pistillate form, producing abundance of large berries. /. Wilsoni was exhibited in 1899 by the Handsworth Nurseries (where there is now a specimen 13 ft. high) as a new kind of seedling origin ; but is apparently identical with var. princeps, from the same firm, described by Moore* a few years earlier. It seems also to be the same as var. latifolia, sold by Lawson in Loudon's time. 2. Ilex Mundyi, Dallimore, Holly, Yew, and Box, 142 (1908). (f)Ilex Aquifolium, var. nigra, Moore, in Gard. Chron. ii. 433 (1874). This differs mainly from /. Wilsoni in the dull and not shining upper surface of the leaf, the marginal spines of which are directed outwards rather than towards the apex. This is a male plant, which was sent out about twenty years ago by the Hands- worth Nurseries, where the tallest specimen is now about 10 ft. 3. Ilex Lawsoniana, Dallimore, Holly, Yew, and Box, 141 (1908). Ilex Aquifolium, var. Lawsoniana, Moore, in Gard. Chron. v. 624, fig. no (1875). A variegated form, like /. Mundyi in the shape and texture of the leaf, which is, however, often sub-entire in margin ; the centre of the blade is marked with broad irregular blotches of yellow. It is one of the handsomest of the golden hollies,2 but often reverts to the green state. II. Leaves not conspicuously reticulate beneath. 4. Ilex altaclerensis, Dallimore, Holly, Yew, and Box, 139 (1908). Ilex Aquifolium, var. altaclerensis, Loudon, Arb. et Frut. Brit. ii. 507 (1838); Moore, in Gard. Chron. ii. 752 (1874). Leaves broadly ovate, 3 to 3^ in. long, and 2 to 2^ in. wide, sub-entire, or with short spiny sinuate teeth variable in number and mostly near the apex of the blade . margin slightly undulate ; petiole purplish, often f in. long, with the base of the blade decurrent on it for a short distance. This is a very fine male tree with purplish branchlets which originated at Highclere in the early part of the nineteenth century. 5. Ilex Hodginsii, Dallimore, Holly, Yew, and Box, 140 (1908). Ilex Aquifolium, var. Hodginsii, Moore, in Gard. Chron. ii. 433 (1874). (?) Ilex Aquifolium, var. Shepherdii, Goeppert, in Gartenflora, iii. 318 (1854) (not Moore3). Leaves ovate or elliptic, 3 to 4 in. long, z\ to 3 in. broad, dark green and very shining above ; margin undulate and with rather distant regular large triangular 1 Moore states that var. princeps was raised from " the seed of /. Aquifolium nigrescens crossed with a male seedling from /. balearica." 2 Gumbleton, in Card. Chron. iii. 595 (1888), states that /. Lcnpsoi.iana originated in " Ilodgen's nursery at Clough- jordan in Tipperary," whence grafts were sent to Messrs. Lawson at Edinburgh. This is an error, and should read " Ilodgins's nursery at Dunganstown. " 3 Ilex Aquifolium, var. Shepherdii, Moore, in Card. Chron. ii. 751 (1874), judging from the description and Moore's specimen in the Kew Herbarium, was identical with /. Hendersoni, or only differed from it in the yellowish green leaves being shining and not dull above. i7J4 The Trees of Great Britain and Ireland divaricate spiny teeth. This has purplish branchlets, and is very vigorous in growth. It is a staminate form, and is saidl to be the best holly for planting in towns, being less affected by smoke than the other kinds. Vars. nobilis and belgica, Moore, in Gard. Chron. ii. 433 (1874), are scarcely distinct from /. Hodginsii. A form of the latter, called /. Hodginsii, " King Edward VII.," with leaves mottled in the centre, and having a broad yellow margin, was put in commerce in 1898 by Messrs. Little and Ballantyne. Var. nobilis variegata has smaller leaves than /. Hodginsii, with a yellow blotch in the centre and a broad green margin. /. Hodginsiiis sold under the name 1. Shepherdii, by the Handsworth Nurseries, who inform us that Mr. Shepherd, past curator of the Liverpool Botanic Garden, received two varieties of seedling hollies from Hodgins, nurseryman in the early part of the nineteenth century at Dunganstown near Wicklow.2 The late Mr. Holmes obtained for the Handsworth Nurseries a stock of these varieties, one of which he named /. Shepherdii and the other /. Hendersoni, the latter name after a friend of the curator. Fine specimens of the original grafted plants still exist at Handsworth. Mr. Holmes always asserted that it was to the variety /. Shepherdii that other nurserymen subsequently applied the name /. Hodginsii? 6. Ilex nigricans, Henry. Ilex Aquifolium, var. nigricans, Goeppert, in Gartenflora, iii. 319 (1854). Ilex Aquifolium, vars. atrovirem and nigrescetis, Moore, in Gard. Chron. ii. 751, 752 (1874). Leaves ovate, 2\ to 3^ in. long, i^ to 2 in. broad, dark shining green above ; differing from /. Hodginsii in the margin being almost flat or only slightly undulate, with smaller sinuate teeth, ending in long slender spines, often very regular and numerous. Under this name may be included perhaps two or three distinct forms, sold as maderensis, maderensis atrovirens, platyphylla, etc., the nomenclature of which is at present confused in different nurseries. 7. Ilex Hendersoni, Dallimore, Holly, Yew, and Box, 140 (1908). Ikx Aquifolium, var. Hendersoni, Moore,* in Gard. Chron. ii. 752, fig. 148 (1874). Leaves elliptic or ovate, 2\ to 3 in. long, \\ to 2 in. broad, dull yellowish green above ; mostly entire or sub-entire, or with a few short sinuate spinous teeth, the margin being only slightly undulate. This is a female tree, which produces large red fruit, but in no great abundance. 8. Ilex camelliœfolia, Henry. Ilex Aquifolium, var. camelliœfolia, Koch, Dendrologie, ii. pt. i. 210 (1872); Moore, in Gard. Chron. ii. 812, fig. 164 (1874). Leaves ovate-oblong, about 4 in. long and 2 in. broad, entire or with a few small sinuate spiny teeth, the margin being slightly undulate ; upper surface deep green and very shining ; petioles and branchlets purplish. This is a very vigorous tree, with dense foliage, and is a female, bearing larger and darker coloured fruits than the common holly. 1 Ilibberd, in Robinson, Eng. Flower Garden, 468 (1893). 2 Cf. London, Arb. et Frut. Brit. i. 116 (1838). 3 It is preferable to use /. Hodginsii, as it is now more commonly known, appearing as an alternative name in the Handsworth Catalogue. /. Sheflierdii is ambiguous, having been applied to both varieties. 4 Cf. p. 1713, note 3. Ilex Ï7Ï5 /. Aquifolium, var. Marnocki, Dallimore, Holly, Yew, and Box. 84(1908), differs mainly from the preceding in the leaves being peculiarly twisted about the middle. Messrs. Fisher, Son, and Sibray tell us that it arose at Ilandsworth as a chance seedling about forty years ago, and state in their catalogue that it " bears immense berries on the two- and three-year-old wood, forming sprays of vermilion red colour from i to 2 ft. long." Their largest specimen is about 16 ft. high. DISTRIBUTION The common holly is a native of western and southern Europe, and extends in two or three geographical forms into Asia Minor, the Caucasus, northern Persia, and China. It cannot live in regions1 where the winter temperature is low, its distribution in Europe being limited to the north and eastward by the January isothermal line of 32° F., so that it is unknown in Russia and Sweden and in the eastern parts of Germany and Austria. It occurs in Norway only on the islands2 and fjords along the west coast from Christianssund (lat. 63° 10') southwards, and is also met with in west Jutland and on the island of Rügen. It exists here and there in Germany, west of a line drawn from Mecklenburg to Bonn, and also in the Black Forest. It is more abundant in the Alps, where it ascends to 4000 feet elevation in Switzerland and in the Tyrol. In France3 it is most common in the central and southern departments, growing, according to Fliehe, on all soils that are not marshy, but preferring those deficient in lime ; it is usually shrubby, but attains large dimensions in Vendée and Corsica.4 Throughout its area it is most often seen as underwood or small trees in the shade of the broad-leaved forest ; but it also grows in many places amidst scrub or in rocky situations in full sunlight. (A. H.) CULTIVATION Though the holly is usually looked upon and treated as a shrub, yet in many parts of England it attains the dimensions of a forest tree, which on account of its beautiful foliage and berries has always been one of the greatest ornaments of our natural woodlands and hedgerows. As it is a most useful nurse to oaks, beeches, and other valuable timber trees, and forms excellent shelter for game, it should be encouraged and planted in all places where the soil suits it. Holly is hardy in all parts of Great Britain ; and though it will grow on almost any soil, thrives especially on deep sandy loam and on soils with cool subsoil, grows well on chalky and limy soils, and in very moist climates on thin rocky hillsides. It sows itself freely where rabbits are kept down, but grows slowly at first. The berries should be gathered in winter, and mixed with sand or light soil in which they decay slowly ; but as the seeds rarely germinate until a year has elapsed, 1 Cf. A. de Candolle, Geog. Bot. Rais. i. 147, 162, pi. i (1855), who gives much information concerning the distribution of the holly. 2 Holly grows on some of the islands near 15ergen, in Hardanger and Sogn fjords ; and attains on Amuglen over 30 ft. in height and 5 ft. in girth. 3 Cf. vol. iii. p. 560, note 4. 4 I have two specimens living at Colesbome, which I brought home in 1903 from the Sila mountains in Calabria, where holly grows at 4000 to 5000 ft. altitude.—H. J. E. 1716 The Trees of Great Britain and Ireland it is better to keep them in sand until the following spring, as is done with haws. When sown the beds should be covered with leaves, fern, or branches, to keep out the frost and drought until they begin to germinate. A rich light soil is best to encourage rapid growth when young, and after two years in the seed-bed the seedlings should be transplanted in the month of May with as little damage as possible to the tender roots. I have found that the seedlings grow faster under a wall with a north aspect than in the full sun. At four years old the seedlings should again be trans planted, and the strongest of them, which may then be 2 to 3 ft. high, can be planted out, either in early autumn if the soil is cool and moist enough, or just before growth commences in May. In the latter case a mulch will be desirable, as they suffer from drought after transplanting, and in my experience it is never wise to transplant1 hollies between November and the end of March. In forming holly hedges care must be taken to have the ground thoroughly clean and deeply dug beforehand, and to keep the young plants free from grass and weeds, which often choke the young trees. For want of this precaution, and even in spite of it, deaths occur on dry or poor ground which may permanently ruin the regularity of the hedge ; and a great deal of money and time are wasted by planting holly hedges and not attending to them afterwards. Holly bears pruning well, and requires attention when young to make a regular and even hedge, and as the growth of the different varieties is very variable, it is important that all the plants should be from the same source. When variegated hollies are required, they are budded or grafted ; and as this is an operation requiring time, experience, and suitable stocks, it is better to purchase the varieties from a nursery. Messrs. Fisher, Son, and Sibray, of the Handsworth Nurseries near Leeds, have long been celebrated for their hollies, which may be transplanted successfully, if due care is taken, up to 5 or 6 ft. high. The holly, though an indigenous species, suffers from cold in very severe winters, as in 1838, when the thermometer fell at Chiswick on 2Oth January to -4!° Fahr. Lindley2 states that the holly in this year "was extensively affected in several places in the middle and north of England ; this plant, however, offered very different powers of resisting cold, some of the varieties proving much hardier than others, and, according to Mr. M'lntosh, those which are variegated more so than the plain kinds." In 1837-1838 /. eakaricawas not in the least hurt about London.2 Moreover, in 1905, when the temperature at Kew fell to 2° Fahr, on 8th February, none of the varieties of /. Aquifolium seems to have suffered.3 REMARKABLE TREES The holly occurs in woods and copses throughout the greater part of the British Isles, ascending* to 1000 feet in the Highlands, but seems to be commonest and of 1 Cf. Gard. Chron. 1848, p. 99. 2 In Trans. Hort. Soc. ii. 226, 275 (1842). 3 Cf. Kew Bull. 1896, p. 9. 4 Moss, in Tansley, Brit. Veg. 126 (1911), states that " the hollyis found in almost every oak wood on the Pennines, but although it sometimes produces flowers it rarely fruits." This is due no doubt to the altitude and prevalent low temperature in these woods. Ilex 1717 its largest size in the moister and milder parts of Ireland, Wales, and southern England. It is a very characteristic tree of the New Forest, and is not uncommon in parts of the Chiltern Hills, where large specimens may be seen on the roadside between Wyfold and Reading. There are few places in Great Britain where the holly grows in greater size and abundance, and forms such an important feature in the scenery, than in the New Forest, where, since the red deer were killed off, it has increased very fast ; and in some of the old woods of oak and beech forms almost impenetrable thickets, which not only add to the beauty of the scenery, but protect many young oaks and beeches from being eaten off by the ponies. I am inclined to think that but for these hollies the number of saplings which are coming up would be so small that the timber trees would in time disappear, and though deer, sheep, and rabbits all browse on and bark hollies in winter, they bear the shade of oak very well. In Mark Ash some of the hollies are over 50 ft. high, with straight trunks, and the annual growths here are over a foot long. In all the beautiful country about Midhurst and Haslemere the holly grows very well, and is cut by gypsies for whip-sticks, which when straight and slender are the best in England, being light and elastic. I have not seen here, however, any trees so remarkable for size as those which grow in Hertfordshire and Bedfordshire on sandy and gravelly loam. At Russells, near Watford, in 1907, I saw a group of trees in a thick shrubbery from 70 to 75 ft. high, but crowded by beech trees, and not well shaped. In Rod's Wood, Teppingley, near Ampthill, Henry measured in 1909 a fine specimen, 60 ft. high and n^ ft. in girth, at a foot from the ground, just below the point where it divided into about seven stems. On the oolite of the Cotswold Hills it is common on old downs and in hedge rows, but though often forming trunks of 4 to 5 ft. in girth does not attain as great height as on better soil. On the coast of Suffolk the holly grows remarkably well on sandy soil in Orwell Park, where in a natural wood of oak and holly I saw many 50 ft. high and more, with clean stems 15 to 20 ft. high, and over 5 ft. in girth. Mrs. Rivis tells us that part of Staverton Wood, near Butley, consists of numerous old holly trees crowded together, and with their stems clear of branches to a considerable height. At Rougham Hall, Norfolk, in 1907, I measured a splendid weeping silver holly 50 ft. by 4 ft. ii in., with a bole 15 ft. high, the finest of the sort I have seen. On the Steiperstone hills, south of Shrewsbury, a natural forest of hollies was said to have existed in which trees of great size were found ; one is mentioned 14 ft. in girth;1 but when I visited this place in 1909 I found that part of the land was now planted, and on the open part the hollies were injured by cattle, and by being lopped for Christmas decorations, so that I could find no old ones more than 20 to 30 ft. high, and few seedlings coming up. At Doddington Hall, near Lincoln, Lord Kesteven in 1907 measured a splendid holly about 50 ft. high, and 9^ ft. in girth at 4 feet from the ground. 1 Notes and Queries, ser. v., xii. 508. VII K iyi8 The Trees of Great Britain and Ireland At Mount Edgecumbe, Henry saw in 1911 a fine holly about 60 ft. high, with a clean stem of 25 ft., measuring 6 ft. in girth. The longest holly hedge in England is probably one bounding the park at Tyntesfield, near Bristol, the seat of G. A. Gibbs, Esq. Planted on a bank 3 feet high it extends by the side of the public road for about two miles. It is about 4 ft. high and 3 feet thick, and is very dense and even. At Kew there is a fine holly hedge surrounding the shrub nursery,1 which is 315 yards long, the greater part of it 9 feet high and 4 feet wide, but one portion as much as 12 feet high and 7 feet in width. A remarkable holly hedge2 of great length at Keele Hall, Staffordshire, is 25 feet high and 30 feet in thickness. Near Bagshot a fine holly hedge8 around a private garden is 100 yards long and 40 feet high. At Gorddinog, Llanfairfechan, North Wales, there is a remarkable avenue of yews, originally planted alternately with hollies, which the owner Colonel Henry Platt, C.B., considers to be very old, one yew tree bearing the date 1654 cut into the bark, with the figures stretched by age to a foot in length. Most of the hollies have died of old age, the largest one, which was 42 ft. high and 8 ft. 8 in. in girth, succumbing last year. A few remain, about the same height, all very decayed, and none exceeding 5^ ft. in girth. The finest golden variegated holly that we know of is growing in the Isle of Man at Kirby Park, the seat of G. Drinkwater, Esq. In January 1913 it was 46 ft. high and 6 ft. 8 in. in girth at one foot from the ground, above which it divides into three stems, forming a beautiful narrow pyramid of foliage. In Scotland the holly is quite as much at home as in England, and attains as great a size. Hutchison, who gives a complete list * of the remarkable holly trees in Scotland, states that they are most abundant in Morayshire and Aberdeenshire, in the basins of the Findhorn, Spey, and Dee, where the climate is mild, and there are numerous woods of ancient date. In Darnaway Forest, Morayshire, there are thousands of hollies, many of large dimensions, the finest, measured by the forester in 1891, being 42 ft. high with a clean bole of 16 ft., girthing 9 ft. 4 in. at five feet from the ground. These grow amidst oaks in a soil of reddish clayey loam, and are supposed to be 200 years old. The most remarkable collection of varieties of the holly is at Ochtertyre, Perthshire. Some of the oldest holly trees in Scotland appear to be at Glenkill, near Lamlash, in the Isle of Arran, the finest measuring in 1891 50 ft. in height and 8 ft. 3 in. in girth at three feet from the ground. An ash-tree, 30 ft. high and 2 ft. in girth, was growing, naturally grafted, on one of these hollies, the junction being about three feet from the ground, where the holly was 8 ft. i in. in girth. Hunter records5 a grand holly tree at Gourdiehill, Perthshire, which had a stem 1 Dallimore, Holly, Yew, and Box, 32 (1908). 2 Cf. Gard. Chron. xiii. 10, fig. 5 (1893). 3 Described in Gard. Chron. xxvi. 424 (1899). A photograph of this is reproduced by Robinson, English Flower Garden, 467 (1893). * Trans. Highland and Agric. Soc. Scotland, iv. 80-94 (1892). 6 Woods of Perthshire, 495 (1883). Ilex 1719 28 ft. long, as clean and smooth as a pillar, and holding a girth of 6| ft. throughout its length. Christison records 1 a holly at Fullarton House, near Troon, 30 to 40 ft. high, with a bole2 6 ft. long and 11 ft. 9 in. in girth, in 1891. Loudon8 mentions a holly at Blair Drummond which grew in sandy loam, and measured 59 ft. by 8 ft., but it is no longer living. Sir Joseph Sabine in Trans. Hort. Soc. vu. 194 (1830), gives a long account of some remarkable holly hedges in Scotland, of which those at Tyninghame, East Lothian, the seat of the Earl of Haddington, extended altogether to a length of 2952 yards. The most striking were those on both sides of a grass walk 36 feet wide, extending from the North Berwick road to the mansion. This walk was 743 yards long, and the hedges 15 feet high, and n feet broad at the base ; another was 170 yards long, 25 feet high, and 13 feet wide. Most of these were planted by Thomas, the sixth Earl, in 1712 ; but when I visited the place in 1905 I found that they had become old and ragged, many of the bushes having died. The largest tree that Sabine mentions here in 1830 was 54 ft. by 5 ft. 3 in. at three feet from the ground, but I measured one no less than 71 ft. by 4 ft. 9 in. drawn up among beech trees, but not a handsome specimen. The most remarkable groups of hollies I have seen anywhere are on the holly bank at Gordon Castle, growing on a moist gravelly old red sandstone soil facing west. These are in clumps, and many have evidently sprung from the same stool. Sabine counted seventy-three groups, containing 508 trees, of which eighty-seven had clean trunks from 8 to 14 ft. long. The largest he measured, which grew at the bottom of the bank, were 52 ft. by 5 ft. 7 in. with a bole of lO^ ft., and 43 ft. 9 in. by 4 ft. 9 in. with a bole of 8^ ft. He mentions one clump which then contained fifty-five trees growing on an area 134 ft. in circum ference, and from i^ to 3^ ft. in girth, which I believe to be the same that I saw in April 1904, not knowing that it had been previously described. I counted fifty-four trees 30 to 40 ft. high, averaging about 3 ft. in girth, and containing about 6 to 7 feet of timber, so that on an area of about a quarter of an acre there must have been over 300 feet of timber.4 Plate 377 gives a very good idea of this wonderful group, which appeared to me, as it did to Sir J. Sabine, to have been the work of nature, but Mr. Webster, the gardener, could give me no record of their age and origin. The Duke of Richmond, however, tells me that they were flourishing in 1760, as they are alluded to in an account of Gordon Castle written at that time. At Colinton House, Midlothian, the seat of J. Erskine Guild, Esq., there are some holly hedges, supposed to have been planted between 1670 and 1680, which are still in good health ; and as I am informed by the gardener, Mr. Bruce, are « Trans. Bot. Soc. Edinburgh, xx. 387 (1896). This tree was photographed by Paxton, who presented in 1894 to the library of the Edinburgh Botanic Garden a book of photographs and measurements of thirty remarkable trees in Ayrshire. * Renwick, in Tians. Nat. Hist. Soc. Glasgow, vii. pt. iii. 263 (1904), gives the girth in 1903 as ilj ft. at five feet from the ground. 3 Card. Mag. xvii. 507 (1841), where a list of measurements in 1836 of all the remarkable trees at Blair Drummond is given. * In Gard. Chron. iii. 51 (1875), Mr. Webster says that one of the trees here measures 7 ft. 9 in. at three feet, and 7$ ft. at ten feet, and another 8J ft. in the narrowest part of the trunk two feet from the ground. T 7 20 The Trees of Great Britain arid Ireland now in better condition than they were twenty-five years ago. Some parts of them are solid green, and quite thick from base to top. The height of the tallest is 40 ft. after clipping, which is done once in two years. The broadest part at the base is 2i ft. through, and where protected from rabbits the leaves touch the ground. They grow on light dark loamy soil, with a sandy or gravel subsoil, but hollies here thrive equally well on clay with a rocky subsoil. These are supposed to be the oldest and the tallest holly hedges in Scotland, and perhaps in the whole of Great Britain. The holly, according to Mr. R. A. Phillips, is distributed throughout Ireland, but is more abundant in the non-calcareous districts of the west, south, and south east than elsewhere. It also occurs on the islands1 off the west coast. In Ireland it formerly attained an enormous size, the most famous tree being one on Innisfallen Island, Killarney, which Hayes recorded2 as 15 ft. in girth in 1794; but I could find no trace of it in 1909. There were also remarkable woods in which holly grew nearly pure, and produced valuable timber in quantity. The late Earl Annesley informed Henry that out of a wood of this kind by the lake at Castlewellan his brother sold in 1871 more than ^500 worth of holly timber; but in 1906 the largest tree which remained was scarcely 6 ft. in girth. Near Mount Usher, in Wicklow, Henry measured in a wood in 1904 a tree 70 ft. in height and 6 ft. in girth. TIMBER, ETC. The wood of the holly is white, hard, and heavy ; and has a fine close grain, being very homogeneous in texture. It takes a good polish, and is used for making mathematical instruments, for inlaying, and for turnery. Pulley-blocks for ships were formerly made of holly. When dyed black, it is a cheap substitute for ebony. Baltimore states8 that snuff-boxes were, in the early part of the last century, made out of the knots and burrs that are sometimes found on the trunk. Young straight quickly grown shoots are used for making walking-sticks and whip handles. Bird-lime is made out of the mucilaginous bark of the young shoots. (H. J. E.) 1 Praeger, in Proc. R. I. Acad. xxxi. pt. IO, pp. 19, 21, 25, records it for Clare Island. 8 Treatise on Planting, 143 (1794). 3 Holly, Yew, and Box, 20 (1908). BUXUS Buxus, Linrueus, Syst. Nat. 9 (1735), Gen. PI. 284 (1737), and Sp. /7.983 (1753); Bâillon, Monog. Buxac. 58 (1859); Müller, in De Candolle, Prod. xvi. pt. i. p. 13 (1869); Bentham et Hooker, Gen. PI. Hi. 266 (1880); Pax, in Engler and Prantl, Nat. Pßanzenfam. iii. 5, p. 133 (1892); Van Tieghem, in Ann. Sei. Nat. v. 289 (1897); Hutchinson, in Kew Bulletin, 1912, p. 52. EVERGREEN trees or shrubs belonging to the family Buxaceae. Young branchlets quadrangular. Leaves opposite, coriaceous, simple, entire, penninerved, shortly stalked, without stipules. Flowers monoecious, without petals, in axillary clusters, which are composed of several staminate flowers, and usually a central single pistillate flower. Staminate flower : sepals four, in two opposite pairs ; stamens four, each oppo site a sepal, with a thick filament inserted beneath the rudimentary ovary, and with a two-celled anther, which dehisces longitudinally. Pistillate flower ; sepals four to seven, occasionally ten, some of which may represent bracts, imbricated ; ovary free, three-celled, crowned by three peripheral styles ; ovules, two in each cell. Fruit,1 a capsule, crowned by the three persistent styles, three-celled, each cell containing two seeds, or occasionally only one by abortion ; capsule, when ripe, splitting longitudinally, through the three styles and the dorsal sutures, thus producing three two-horned valves ; the endocarp afterwards opens down six lines, suddenly expelling the seeds, which are trigonous, shining, black, and smooth. The above description is restricted to the section Eu-buxus of the genus,2 which comprises about ten species, two, natives of tropical Africa and Socotra, and about eight species, natives of the Mediterranean region, Caucasus, Persia, Hima layas, China, and Japan. The following synopsis briefly deals with those which are in cultivation :— I. Branchlets glabrous, or occasionally with a few minute hairs near the nodes. i. Buxus japonica, Müller, in De Candolle, Prod. xvi. i, p. 20 (1869); Shirasawa, Icon. Ess. Forest. Japon, ii. t. 38, figs. 16-32 (1908). Buxus scmpervircns, Linnaeus, Mai. japonica, Makino, in Tokyo Bot. Mag. ix. 281 (1895), and xv. 169 (1901); Hayata, mjburn. Coll. Sei. Tokyo, xx. 3, p. 82, t. vi. c. (1904). A shrub or small tree, with glabrous branchlets. Leaves coriaceous, very similar to those of B. sempervirens in size and appearance, but usually more oval and occasionally almost orbicular, f to i in. long and £ in. broad, emarginate and rounded 1 Cf. Marshall Ward, Trees, iv. 154, fig. 147 (1908). 2 The other sections of the genus, which are sometimes regarded as distinct genera, are Buxella, which includes four species, natives of Madagascar, tropical and south Africa ; and Tricera, comprising about a dozen species, natives of the West Indies. I72I 172,2, The Trees of Great Britain and Ireland at the apex ; upper side of the short broad petiole pubescent, the leaf elsewhere being glabrous. Flowers of both sexes sessile, similar to those of B. sempervirens, but with the rudimentary ovary of the staminate flower much enlarged and as long as the inner sepals. This is wild in the mountains of Japan, where it has given rise in cultiva tion to several peculiar forms, one of which with almost orbicular leaves, puckered and uneven on the surface, has been introduced into Europe. The typical form is rare ; but is represented at Kew by a shrub about 3 ft. high.1 2. Buxus microphylla, Siebold and Zuccarini, in Abh. Ac. München, iv. 2, p. 142 Buxus sempervirens, var. microphylla, Blume, ex Miquel, in Ann. Mus. Lugd. Bat. iii. 128 (1867); Hayata, mfourn. Coll. Sei. Tokyo, xx. 3, p. 83, t. vi. D. (1904). Buxus japonica, var. microphylla, Müller, in De Candolle, Prod. xvi. i, p. 20 (1869). A low shrub, with very slender branchlets, which are glabrous or with a few minute hairs above the nodes. Leaves thin and membranous, spatulate, lanceolate or elliptic, about \ in. long and \ to £ in. broad, emarginate at the apex, slightly pubescent on the upper side of the very short broad petiole. Flowers, as in B. japonica. This species 2 occurs wild in Japan, in the provinces of Shimosa, Awa, and Tosa. It is a very distinct plant, with small thin leaves, and has lately been intro duced into Kew. 3. Buxus Harlandi? Hance, injourn. Linn. Soc. (Bot.} xiii. 123 (1873). A small shrub, attaining i to 2 ft. in height, with densely crowded foliage. Young branchlets very slender, glabrous or occasionally with a few minute hairs above the nodes. Leaves oblanceolate, very narrow in proportion to their length, f to i£ in. long, ^ to £ in. wide, emarginate at the apex, very tapering at the base, quite glabrous throughout, or with slight pubescence on the upper side of the petiole. Flowers similar to those of B. japonica, sessile in both sexes, with the rudimentary ovary of the staminate flower as long as the inner sepals ; but with the style of the pistillate flower as long as the ovary. This shrub, which is very distinct in appearance, occurs in central and southern China, where it is found growing in the shingly or rocky beds of rivers and streams. It has been in cultivation about ten years at Kew, where it is a dwarf shrub.4 II. Branchlets minutely puberulous. 4. Buxus balearica, Lamarck, Encyc. i. 511 (1753); Willdenow, Sp. PI. iv. 337 (1805). Buxus sempervirens, var. gigantea, Loiseleur, in Duhamel, Traité des Arbres, i. 82, t. 23 (1801). 1 B. Fortunei rotiindifolia, of some French nurseries, is probably a prostrate form of B. japonica. 2 B. stmophylla, Hance, in Jcurn. Bot. vi. 331 (1868), wild in the province of Fukien, in China, is closely allied to, if not identical with B. microphylla. B. sempervirens, var. riparia, Makino, in Tokyo Bot. Mag. xxvi. 293 (1912), a shrub 3 ft. high, growing beside the river Yoshino, in the province of Tosa, is said to be intermediate between B. japonica and B. microphylla. 3 This species was founded by Hance, on specimens gathered in Hong-Kong ; but his sheet, No. 322, so-named, includes two different plants, one, the true B. Harlandi, and the other, resembling B. japonica in the shape of the leaves, but differing in having pubescent branchlets. 4 B. Fortunei, Carrière, in Rev. Hort. xlii. 519 (1871), also sometimes known as B. longifolia, Hort, judging from specimens cultivated under this name in France, is identical with B. Harlandi. Buxus 1723 A shrub or small tree. Young branchlets with a minute pubescence, only visible with a strong lens. Leaves light green, coriaceous, larger than those of the common box, but similar in shape, elliptic, i to i| in. long, ^ to fin. wide, emarginate at the apex, minutely pubescent on the upper side of the short petioles, but glabrous elsewhere. Staminate flowers shortly stalked ; pistillate flowers with the styles as long as the ovary. This species * occurs in the Balearic Isles, and in the province of Granada, in Spain, on the coast at Nerja, near Malaga, and on the Sierra de Almijara, at 2000 feet elevation. Loudon2 states that it was introduced into England in 1780. It is perfectly hardy, the oldest specimen at Kew being about 23 ft. high. III. Branchlets densely pubescent. 5. Buxus sempervirens, Linnaeus. See p. 1724. 6. Buxus Wallichiana, Bâillon, Monog. Buxac. 63 (1859); Hayata, in Journ. Coll. Sei. Tokyo, xx. 3, p. 84, t. vi. E. (1904). Buxus sempervirens, J. D. Hooker. Fl. Brit. Ind. v. 267 (1887) (not Linnaeus); Gamble, Indian Timbers, 592 (1902). Buxus sempervirens, Linnaeus, var. liukiuensis, Makino, in Tokyo Bot. Mag. ix. 279 (1895), ar>d xv. 169 (1901). Buxus liukiuensis, Makino, in Tokyo Bot. Mag. xvi. 179 (1902); Schneider, Laubholzkunde, ii. 140 (1907). A shrub or small tree. Branchlets densely pubescent. Leaves lanceolate, i to 2\ in. long, \ to \ in. wide, rounded or emarginate, with a short point at the apex ; petiole densely pubescent, the pubescence spreading along the midrib on the upper side. Flowers of both sexes sessile ; rudiment of the ovary in the male flower very short, dilated and slightly four-lobed at the apex ; styles in the pistillate flower as long as the ovary, persisting elongated and erect on the fruit. B. Wallichiana is widely spread in Asia, occurring in India in the Suliman and Salt ranges, and in the Himalayas from Kumaon to Bhutan (but not in Sikkim), at 4000 to 8000 feet. It is common in the mountainous districts of central and southern China, and is the only species known at present in the Liu Kiu Islands and Formosa. Its distribution8 in the Himalayas is local and peculiar ; but it mainly grows along the banks of streams in moist and sheltered places, preferring a northerly aspect ; and is met with on sandstone as well as on limestone. It often occurs in considerable quantity, attaining a large size, trees being recorded as much as 5 ft. in girth ; but is slow in growth, averaging, according to Gamble, about twenty rings per inch of radius. In China,4 the box is known as huang-yang, and is of consider- 1 B. longifolia, Boissier, Diag. Fl. Orient. 107 (1853), and Fl. Orient, iv. 1144 (1879), °f which I have seen no speci men, is said to occur in the mountains of Syria ; and appears to be closely allied to B. balearica. It is entirely distinct from B. longifolia, Hort, which appears to be a synonym of ß. Harlandi. Cf. p. 1722, note 4. 2 Art. et Frut. Brit. iii. 1341 (1838). His account of the size and distribution of this species, which has been followed by Dallimore, Holly, Yew, and Box, 229 (1908), is erroneous. So far as I can learn, B. balearica is a small tree or shrub, not exceeding thirty feet in height ; and does not occur wild except in the localities mentioned ahove. Nyman, Consp. Fl. Eurcp. i. 647 (1878), records it for Sardinia ; but this is doubtful. s A dwarf box with very small leaves occurs in the Himalayas at high elevations ; and has been referred to B. japonica, var. microphylla, Müller, by Hooker, Flora Brit. India, v. 267 (1887) ; but is much more pubescent than the Japanese plant, and is probably a distinct species. 4 The species of box in China require further study. Buxus Henryi, Mayr, Fremdländ. Wald- u. Parkbäume, 451 (1906), described from specimens (No. 3387) collected by me on cliffs near Ichang, in central China, has glabrous branchlets, 1724 The Trees of Great Britain and Ireland able economic importance in the mountains north of Ichang, in Hupeh, where it is a tree of considerable dimensions. B. Wallichiana is very rare in cultivation ; but thrives at Kevv, where speci mens, about 6 or 7 feet high, produce flowers and fruit regularly. (A. H.) BUXUS SEMPERVIRENS, COMMON Box Buxus sempervirens, Linnseus, Sp. PI. 983 (1753); Loudon, Arb. et Prut. Brit. iii. 1333 (1838); Bâillon, Monog. Buxac. 59 (1859) ; Müller, in De Candolle, Prod. xvi. i, p. 19 (1869); Willkomm, Forstl. Flora, 802 (1887); Mathieu, Flore Forestière, 306 (1897). Buxus arborescens, Miller, Diet. ed. viii. No. i (1768). A shrub or small tree, attaining about 30 ft. in height and 3 ft. in girth. Young branchlets densely pubescent with short white hairs, which are more or less retained in the second and third years. Leaves persistent five or six years, coriaceous, opposite, oval or elliptic, averaging i to i£ in. long and ^ in. broad, rounded and usually emarginate at the apex ; shining and dark green above ; duller and yellowish green below ; secondary nerves pinnate, often forked, conspicuous on the upper surface ; margin entire ; tapering at the base to a very short petiole, which is pubescent like the branchlets, the pubescence extending along the midrib on the upper surface, and on the edges of the base of the blade. Flowers small, white ; both sexes sessile ; rudimentary ovary of the staminate flower scarcely half as long as the inner sepals ; styles of the pistillate flower short, about half as long as the ovary. Capsule ovoid, longer than broad, brown when ripe, crowned by short spreading styles ; seeds trigonous, smooth, shining, black. The seedling * has two oblong obtuse glabrous cotyledons about \ in. long, raised above ground on a glabrous caulicle about i^ in. long ; primary leaves opposite, decussate, elliptic, shortly stalked. VARIETIES A considerable number of varieties of the common box are in cultivation, most of which have arisen in gardens and nurseries—the variation in the wild state z being slight. i. Var. angustifolia, Loudon, Arb. et Frut. Brit. iii. 1333 (1838). Buxus angustifolia, Miller, Gard. Diet. ed. viii. No. 2 (1768). Leaves oblong-lanceolate, about I in. long and \ in. wide. This is said to occur lanceolate leaves as much as 3 in. long, and f in. broad ; staminate flowers on long pedicels, with a minute and linear rudimen tary ovary. This species, which is allied to B. balearica, has remarkably fine foliage and conspicuous flowers ; but has not yet been introduced. It has lately been fully described and figured by Diimmer, in Card. Chron. Iii. 423, fig. 182 (1912). 1 Cf. Lubbock, Seedlings, ii. 481, fig. 639 (1892). 2 The typical form of the species was distinguished as var. arborescent, by Linnaeus, Sp. PI. 983 (1753), a name kept up by Loudon, Arb. et Frut. Brit. iii. 1333 (1838). Var. grandifolia, Müller, in De Candolle, Prod. xvi. I, p. 19 (1869), wild in Spain, Greece, and the Caucasus, is scarcely distinguishable, though occasionally the leaves are longer and more lanceolate than in the type. Buxus I725 wild in Algeria ; and is commonly shrubby. An upright tall-growing form of this is known as var. salicifolia elata. 2. Var. myrtifolia, Loudon, Arb. et Frut. Brit. 1333 (1838). Buxus myrtifolia, Lamarck, Encyc. Meth. i. 511 (1783). Leaves dark green, oblong-lanceolate, smaller than in the last variety, about f in. long and \ in. wide. This is a low shrub, which was described by Lamarck, from specimens obtained from M. Gels' nursery at Paris, where it probably originated. 3. Var. myosotifolia, Dallimore, Holly, Yew, and Box, 227 (1908). Leaves resembling those of var. myrtifolia, but still smaller, about \ in. long and ^ in. broad, lanceolate, dark green. This is a dwarf shrub. 4. Var. rosmarinifolia, Bâillon, Monog. Buxac. 61 (1859). Leaves lanceolate or spatulate, more slender and thinner in texture than the last variety, about \ in. long and \ in. wide, marked beneath with whitish clots and tubercles. This variety, which is also known as var. thymifolia, grows to be a bush 5 or 6 ft. high. 5. Var. suffruticosa, Linnaeus, Sp. PI. 983 (1753). Buxus suffruticosa, Miller, Gard. Diet. ed. viii. No. 3 (1768). Leaves oval or obovate, \ in. long and \ in. wide. This is the well-known dwarf variety, which is used for edging beds in gardens. It has been in cultivation for several centuries at least, and is occasionally called var. nana or var. humilis. 6. Var. latifolia, Dallimore, Holly, Yew, and Box, 226 (1908). Under this name are included several forms, in which the leaves are broader than usual, averaging i in. long and f in. wide. In var. latifolia bullata, Späth, the leaves are uneven with peculiar swellings. Var. handsworthensis, Fisher, with broadly oval leaves, is vigorous in growth, and suitable for making hedges. 7. A considerable number of variegated forms are in cultivation. Those with leaves normal in size, or nearly so, are :—Var. argentea or var. argenteo-marginata, leaves white in margin ; var. aureo-marginata, leaves yellow in margin ; and var. aureo-maculata, leaves spotted with yellow. Var. elegantissima, Koch, Dendrologie, ii. pt. i. 477 (1872), is a distinct form with small oval leaves, variegated with white, and with many of the leaves deformed. 8. Var. peitdula, Simon-Louis, Cat. 1869, p. 21. Tree-like in habit, and graceful in outline, the secondary branches and branchlets being pendulous. 9. Var. pyramidalis, Simon-Louis, Cat. 1869, p. 21. Pyramidal in habit, with upright branches. DISTRIBUTION The common box is a native of western Europe, the Mediterranean region, the Caucasus, and northern Persia. It is probably a true native of England ; and in the south and east of France ' is widely spread in the Jura, Dauphiné, Languedoc, 1 Chatin, in Bull. Soc. Bot. France, viii. 364 (1861), considers the box to be naturalised in many localities in France, which are mainly in the neighbourhood of abbeys and castles that date from the middle ages. He mentions the forest of Marly, Vaux-de-Cernay, Neauphle-le-Chateau, Arthieul near Magny, Roche-Guyon, Chantilly, Nemours, Provins, and Jaux near Compiègne. He adds that it is abundant on the millstone grits and sandstones of Vaux-de-Cernay, near Taris, and on granite at Mauves-sur-Loire. VII L 172,6 The Trees of Great Britain and Ireland Provence,1 and Pyrenees,2 and occurs in a few scattered spots in the north of the departments of Meuse and Meurthe-et-Moselle, extending across the frontier into Belgium. It is recorded for Germany from one locality in Baden. Farther south it is common in the southern Tyrol, Istria, Dalmatia, the Balkan States, and Greece. It is also met with in Italy, Spain, and Portugal ; and is a rare plant in the cedar forests of Algeria. In Europe, it is usually a gregarious shrub, often growing on arid hills or mountains on limestone3 soil in the Mediterranean region. It is cultivated in Norway as far as lat. 67° 56' on the coast, and in Sweden up to lat. 59° 7'. The box attains its largest size in the Caucasus,4 where it is very common in the coast region along the Black Sea from Sotschi to Batoum, at elevations between sea-level and 4000 ft. It also grows wild in Talysch, but in other parts of the Caucasus is a doubtful native, though it is often planted, being considered a holy tree by the natives. The finest trees grow inland from the Black Sea, at about 2500 ft. altitude, where they are commonly 30 to 40 ft. in height, with stems 8 to 12 inches in diameter. Formerly still larger trees were known, Koppen4 mentioning 50 ft. in height and 2 ft. in diameter as the maximum size. Many large trees are of enormous age, 500 to 600 annual rings having been counted by Medwejew ; but nearly all these are rotten at the heart. The box grows usually in the Caucasus in narrow bands along rivers and streams as undergrowth in the great forests of oak, ash, and beech ; and is rarely found mixed with conifers. It thrives best in moist, rainy, sheltered, and shaded spots. The woods of box tree in Georgia are mentioned5 by Marco Polo. Until about 1890 the export of boxwood from the Black Sea to England, France, and Turkey, was enormous, averaging 2340 tons annually for the years 1883-1887. Consul Stevens, in his Trade Report* for Batoum for 1895, states : "Although all the private forests of boxwood have been exhausted, the Government up to the present still refuse to sell or allow boxwood to be cut in their extensive forests throughout Abkhasia ; consequently the total exports from the Caucasus have not exceeded 1200 tons." The box appears to be very common in the Elburz mountains in northern Persia, especially in the forests of Mazanderan, whence the export of boxwood in 1906 amounted to 125,864 pieces, weighing about 1560 tons.7 Whether the box is a native of England or not is doubtful ; but it is certainly 1 Tansley in Gard. Chroti. Hi. 113, fig. 51 (1912) describes the peculiar shrubby vegetation, which grows on the southern slopes of the hilly regions of Provence, at about 4500 ft. elevation. This consists of isolated bushes, with bare rock between them, of box, lavender, and Genista cinerea. The north slopes are covered with beech woods, in which there are holly and box, the latter being very abundant inland. 2 On the south side of the Pyrenees, above Venasque at 5000 ft. altitude, I saw, in 1912, much box, forming dense scrub on sunny slopes ; and it is the prevailing undergrowth in many of the valleys of the western I'yrenees on the French side. In the forest near Esterencnby, south of St. Jean-Pied-de-1'orl, box attained 30 ft. high on limestone, and was being felled for making prayer-beads. 3 It is most commonly found on limestone; but occurs frequently on other soils, as in the cases mentioned (p. 1725, note i) by Chatin ; and according to De Candolle, Geog. Bot. Kais. L 426 (1855), '* grows on schist in the Pyrenees, on granite in Brittany, and on volcanic soil in Auvergne. Gèze, in Bull. Soc. Bet. France, Iv. 464 (1908), states that at Villefranche it is not calcicole. 4 Cf. Radde, Pßanznt-verb. Katikasusländ. 145, 182, 201 (1899), and Koppen, Holzgeajächse Europ. Russlands, ii. 1-9 (1889). 6 Yule, Marco Pole, \. 50, 54 (1871). 6 Foreign Office Ann. Series, No. 1717, p. 27. ' Consular Report for Resht, No. 3864, p. 25. Buxus 1727 naturalised, if not truly indigenous, in a few localities. It was cultivated1 in Britain by the Romans ; and as it seeds itself freely in the south of England, it may have spread from abandoned villas. It was well known in Anglo-Saxon times, the earliest mention, I was informed by the late Dr. Skeat, being in the " Corpus Glossary" of Latin and English words, which was written about 750 A.D. In this work box is given as the Anglo-Saxon equivalent of the Latin word buxus. The word box begins to appear early in place-names,2 the oldest example known to Dr. Skeat being Box-ôra (i.e. box-bank), the old name of Boxford, Berks ; and there must have been box trees here at an early date. In the thirteenth century, numerous names of places occur with box, as La Boxe, Le Boxe, Hundred de Boxe, Boxen, Boxford, Boxhaie, Boxhey, Boxore, Boxley, Boxland, Boxstead. These place- names show that box was well known in former times ; but whether it was wild or cultivated, there is no means of determining. Professor Babington3 believed that the following extract from the beginning of Asser's Life of King Alfred showed that it was plentiful in Berkshire 1000 years ago : " Berrocscire ; quae paga taliter vocatur a Berroc sylva, ubi buxus abundan- tissime nascitur." Cough's Camden, 155 (1789), says: "The last remains of Boxgrove3 in Sulham parish, near Reading, whence the county probably took its name, were grubbed up about forty years ago." Gérard,4 writing in 1597, says : "The box tree groweth upon sundry waste and barren hills in England." Ray5 in 1696 records it growing at Boxhill near Dorking, at Boxley in Kent, and at Boxewell in the Cotswolds. At the present day, box is apparently wild in several places in the south of England, the most famous being Box Hill in Surrey, where many acres on the western slopes are covered with a mixture of yew, box, and other trees. The occurrence of the box-tree here was first recorded by Merrett6in 1666. Count Solms-Laubach suggests7 that the box and yew trees of Box Hill might probably be the remains of a native forest which originally clothed the North Downs. He urges the unlikelihood of such a soil as that of Box Hill being planted at all, and the improbability of any one hitting upon such a combination as yew and box for the purpose. Manning and Bray, History of Surrey, i. 560 (1804), give the follow ing account : " The Downs, which rise from the opposite bank of the Mole, are finely chequered with Yew and Box trees of great antiquity, to a considerable height. Of the latter of these in particular there was formerly such abundance that that part of the Downs which is contiguous to the stream, and within 1 Clement Reid, at a meeting of the Linnean Society, London, on 2nd December 1909, said that Box leaves have been found in three different rubbish heaps in the Roman remains at Silchester. The branches may have been used for wreaths. 2 The names of places with box, given by Spelman, Villare Anglüum (1653) and by Adams, Index Villaris (1680) are Box (Wilts), Boxend (Beds), Boxford (Berks and Suffolk), Boxley (Kent), Boxted (Essex and Suffolk), Boxwell (Gloucester), and Boxworth (Berks and Cambridge). Some of these, as we know from the old spelling, do not indicate the box tree ; thus Boxworth near Cambridge means the " farm of the buck." In some cases these places seem to be connected with the Roman occupation of Britain, as Boxmoor House (Herts), near which a Roman dwelling-house was discovered in 1851 ; but where the places occur on the chalk downs, the presumption is that the tree is indigenous. 3 Phytologist, iv. 873 (1853). Cf. Stevenson, Asser's Life of King Alfred, pp. I, 156, 157 (1904), who states that Berroc Wood was identified with Boxgrove by Francis Wise in 1738. * Herball, 1225 (1597). 6 Syn. Meth. ii. 310 (1696). Box does not appear to be growing wild at the present time near Boxley in Kent. 6 Pinax. Rerum Nat. Brit. 18 (1666). 7 Cf. article by G. R. 11. Murray, injourn. Bot. xxxix. 27 (1901). 172,8 The Trees of Great Britain and Ireland the precinct of this Maner, hath always been known by the name of Box Hill. Here was formerly also a Warren with its Lodge ; in a lease J of which from Sir Matthew Brown to Thomas Constable, dated 25th August 1602, the Tenent covenants to use his best endeavours for preserving the Yew, Box, and all other trees growing thereupon ; and in an account of the rents and profits for one year to Michaelmas, 1608, the receipt for Box trees cut down upon the Sheep Walk on the hill is £50. I have seen also an account of this Maner, taken in 1712, in which it is supposed that as much had been cut down2 within a few years before as amounted to .£3000." E. S. Marshall3 says : " He must be very sceptical who doubts it being native on the steep slopes of Box Hill, above Burford Bridge. I have also seen it growing rather plentifully a mile or more away towards Betchworth." Bromfield * says : " Box is profusely abundant on most parts of Sidon Hill, in Highclere Park (Hants), scattered over its shelving sides as if quite spontaneous, and said to disperse itself freely by seed ; " but he avers that it was certainly planted here. Thus the natural appearance now of the box on Box Hill, Surrey, is no sure proof of its being indigenous there. Another locality where the box occurs apparently wild is the Chiltern Hills; as on the chalk downs5 between Ashridge and Berkhampstead, where there are some very old-looking trees. Near here, on the top of Dunstable Downs, there is a place named Boxstead. On the Chequers Court estate, about half a mile from Ellesborough Church, near Wendover, box has all the appearance of being indigenous6 over a considerable area. Mr. Raffety of High Wycombe tells me that box is thickest here in two valleys, known locally as Ellesborough Warren and Kimble Warren, the bushes being about 20 ft., with numerous seedlings of all ages. It extends up over the adjoining spur of the Chiltern Hills to an altitude of 500 feet, the chalk subsoil being near the surface, and the exposure almost due south-west. Messrs. Sprague and Hutchinson state 7 that some of the stems are 8 in. in diameter, and that the only tree which has obtained a place amongst the box thickets is the elder. There is a place named Boxe in Domesday Book for Herts, section xxviii. ; but the village8 so-called no longer exists, being now part of Wymondley. Mr. H. Clinton-Baker tells me of a field at the Priory, Wymondley,9 around which is a broad 1 Barrington, in Phil. Trans, lix. 23 (1769) quotes from A Journey through England, printed in 1722 : "Box Hill was first planted by that famous antiquary the Earl of Arundel, with box wood, designing to have a house there ; but want of water made him alter his resolution and build one at Albury hard by." This is erroneous, as the Earl of Arundel was only sixteen years old in 1602, when Box Hill was already covered with box and yew, according to the lease cited above. 2 Ellis, Timber Tree Improved, 103 (1745), savs that "great quantities of box were felled off the Chalky Downs near Dorking in 1716, which paid its owner several hundred pounds." 3 In/ourn. Bot. xlv. 346 (1907). * In Phytologist, iii. 817 (1850). 6 Cf. W. G. Smith, in Jour». Bot. xxxix. 73 (1901). fi Loudon, Derby Arboretum, 50 (1840), says : "There are extensive native woods of the box tree on the estate of Sir Robert Russell at Chequers in Buckinghamshire." 7 Card. Chron. Hi, 404 (1912). The southern end of Ellesborough Warren is Velvet Lawn, a favourite meet of the hounds. 8 Chauncy, Hist. Antic. Herts, ii. 126 (1826), says : "The Vill or Palish of Box was situated between the parishes of Stevenage, Chivesfield, and Walkerne ; and this parish was called Box from a great wood which retains the name to this day." Boxbury Farm and Box Wood are in the parish of Walkern ; but there are no box trees in these places at the present time. 8 It is worth recording here that a chestnut tree, mentioned in Domesday book, still survives at Wymondley. Mr. II. Clinton-Baker tells me that it is a mere shell, no less, however, than 19 ft. in diameter, and still bearing fruit. Buxus 1729 some Duxus 17 belt (about 10 yards in width) of box trees, which average 20 ft. in height, s being 2 feet in girth. (A. II.) REMARKABLE TREES Though a native of countries much hotter than Britain, and known to most people only as a bush, box is capable of attaining large dimensions in England, and may under favourable circumstances become a small tree. The largest I have seen in this country are in the Hermitage Road in Hitchin, where there is a row of about forty trees, many of which attain 25 to 30 feet in height, and 2 to 2.^ ft. in girth, the largest being 35 feet by 2 ft. 10 in. Mr. Seebohm informed me that this line of trees formerly grew on private property1 which he bought, and opened a new road on which they now stand between an old wall and iron railings which protect them. At Beckford Hall, Gloucestershire, there is a walk 3 yards wide on each side of which a line of box trees grow which are from 30 to 31 feet high and 20 to 30 inches in girth. A monastery once stood here, and the trees are supposed to be about 800 years old, which is probably twice or three times their real age. At Boxwell Court,2 Gloucestershire, the seat of the Rev. O. Huntley, a wood of box trees exists which must be of very great age. Mr. Huntley showed me in the will of his ancestor, Henry Huntley, dated 1556, the following passage : " I will that it shall be lawful for the said Anne my wife, to cut and fell all my boxe, reserving the young store, at any time or times at her pleasure within the space of the said five years." It is thus clear that 350 years ago this wood was looked on as a valuable possession; and that the trees were coppiced, as they are now. It is possible that they were planted by some monk or returned Crusader, and are not, as some have supposed, wild. This wood lies on a steep slope just below the level plateau of the Cotswolds at an elevation of 500 to 600 feet and extends for about 800 yards along the slope which faces south. They grow very thickly and form a dense shade under which nothing grows, and show every sign of having been regularly cut over and reproduced from the stool ; in one place fresh plants have been planted in rows, 5 feet apart, which now form tall slender poles clean to 10 or 15 feet high, many 01 which have died from overcrowding. In other parts the shoots average 20 to 25 ft. high, and though the average girth is not above 6 to 8 in. yet there are a few stems of 2 ft. and over, the thickest that I measured being 2 ft. 10 in. and 3 ft. 4 in. in girth. Mr. Huntley tells me that, though an uncle of his is reputed to have sold a 1 W. Wilshere, M.P., who owned this property, states in Loudon, Gard. Mag. xv. 236 (1839), that these box trees were, in 1839, sixty in number, forming a hedge about I So ft. long. They then averaged 36 ft. in height, and 3 ft. 3^ jn. in girth at 2 ft. from the ground ; and were very old, thin, and ragged. They were supposed to have been planted as a border, and allowed to grow up through neglect ; but their exact history was unknown. Cf. J. E. Little, \nfourn. Hitchin Nat. Hist. Club, May 1891, who adds that in the park at Hexton, near Hitchin, box is very luxuriant, forming tall hedges along the drives. 2 Rudge, History of the County of Gloucester (1803), states: "Boxwell, anciently Boxewelle. This name is derived from a box wood of about 16 acres, within a warren of 40 acres, from which arises a considerable spring. This is the most considerable wood of the kind in England, excepting Uoxhill in Surrey ; and from the name, which has now been on record for more than seven centuries, it must have been of long standing." Ray, Syti. ii. 310 (1696), speaking of box trees, says: "At Boxwel in Coleswold in Gloucestershire, and at Boxley in Kent, there be woods of them.—Mr. Aubrey's Notes." Cf. p. 1727, note 5. 173° The Trees of Great Britain and Ireland quarter of an acre for £jo, between 1858 and 1863, yet now he can only get £i a ton for the wood, and at that low price no one seems to want it. In parts of the wood where rides have been cut, or openings made by the falling of trees, seedlings spring up abundantly ; but the growth of the shoots from the stool seems very slow, owing perhaps to the rabbits, which are hard to keep out. Mr. Cedric Bucknall describesJ another wood of box trees, between Wotton- under-Edge and Alderley, clothing the hill-side for a considerable distance, and with abundance of natural seedlings. In Ireland perhaps the best specimens of box are those growing in the grounds of the Earl of Rosse at Birr Castle. TIMBER The wood of the box tree is dense and homogeneous, with a very fine grain ; and is said to be the nearest approach to ivory of any wood known. Boxwood is unrivalled for wood-engraving, and is used for turnery and inlaying, and for making rules, scales, and other mathematical instruments. It is also employed in making shuttles and rollers that are used in textile factories. A good account of boxwood, with information about the best modes of felling, seasoning, and shipping, is given by Gamble,2 who quotes largely from a letter written by Messrs. J. Gardner and Sons, Liverpool. Boxwood from the Caucasus, whence formerly the main supply was drawn, is now being replaced, except for the very best articles, on account of its increasing cost, by "West Indian boxwood,"3 by Buxus Macowani from South Africa, and by other woods, belonging to different and often quite unallied genera. (H. J. E.) 1 In Journal of Botany, xxxix. 29 (1901). Cf. also J. II. White, Flora of Bristol, 523 (1912). 2 Manual of Indian Timbers, 592-593 (1902). 3 Sir David Train, Director of Kew Gardens, informs me that " West Indian boxwood " is not really shipped from the West Indies, but from Venezuelan ports ; and that its botanical origin is still unknown. It was erroneously stated in Keiv Ball., 1904, p. 11, to be Tabebuia pcntafhylla, Bentham and Hooker, a Bignoniaceous tree, which is known in the West Indies as " white cedar." II. Stone, Timbers of Commerce, 169, plate xii. fig. 105, gives an account of the so-called " West Indian boxwood," which he confuses with " white cedar," although he rightly questions the accuracy of the determination of Kew fSuIL, 1904, p. 11. " West Indian boxwood " is used for making parasol and umbrella handles, shuttles, rulers, thermometers, etc.— A.H. CRAT^EGUS Cratœgus, Linnœus, Sp. PI. 475 (1753); Bentham et Hooker, Gen. PL i. 626 (1865); Koehne, Deut. Dend. 227 (1893); Sargent, Trees N. Amer. 363 (1905); Schneider, Laubholzkunde, \. 766 (1906). Oxyacantha, Medicus, Phil. Bot. i. 150 (1789). Azarolus, Borkhausen, Forst. Bot. ii. 1224 (1803). Mespilus, sub-genus Cratœgus, Ascherson and Graebner, Syn. Mitteleurop. Flora, vi. pt. ii. 12 (1906). TREES or shrubs, belonging to the order Rosaceae, usually armed with simple or branched spines, which are either axillary accompanying a bud, or terminate a short shoot. Leaves usually deciduous, alternate, simple, stalked, usually lobed, serrate ; stipules often foliaceous and persistent on the long shoots. Buds small, globose, with numerous imbricated scales. Flowers, in corymbs, which are terminal on short lateral leafy branchlets ; with quickly deciduous linear bracts and bractlets ; pedicellate, regular, perfect ; calyx superior, with an urceolate, campanulate, or obconic calyx-tube, and five lobes, which are reflexed after the flower opens and either fall off or persist enlarged on the fruit ; petals five, inserted with the stamens on the edge of a disc lining the calyx-tube ; stamens 5, 10, 15, 20, or 25, with filaments broad at the base and incurved ; ovary composed of one to five carpels, concealed in the bottom of the calyx tube and adnate to it ; styles, one to five, free, with dilated truncate stigmas ; ovules two in each cell, erect. Fruit, a false berry or haw, usually umbilicate at the apex, and often crowned by the marcescent calyx-teeth, composed of the fleshy calyx-tube, which encloses one to five stones or nutlets, each containing one seed, the other ovule having aborted. This genus is widely spread in the extratropical regions of the northern hemisphere, occurring in Europe, Asia Minor, Siberia, Himalayas, China, and Japan ; and with numerous species in North America. Schneider admits about 150 species in all ; but Sargent and other American botanists have already described over 500 species in America alone, most of which may be regarded as varieties or hybrid forms. At least sixty species are in cultivation, all of which are either shrubs or small trees, not coming within the scope of our work. The native Whitethorn, which is described below, is now usually considered to comprise two species. Crataegus is closely allied to Mespilus, of which it has been made a section by some botanists. The following hybrids, one of which is doubtful, between the two genera are worthy of brief mention. I73I The Trees of Great Britain and Ireland 1. Mespilus grandiflora, Smith, Exot. Bot. \. 33, t. 18 (1814). Mespilus Mala, Poiret, in Lamarck, Encyc. Meth. Suffi, iv. 71 (1816); VV. J. Hooker, in Bot. Mag.l. 3442 (1835). Mespilus Smithü, De Candolle, Prod. ii. 633 (1825); Loudon, v4/-& ^ />•»*/. Brit. ii. 878 (1838). Cratcegus lobata, Bosc, Nouv. Cours. Agric. it. 223 (1821). Cratagus grandiflora, Koch, in Verh. Ver. Bef. Gartenb. i. 227 (1853). Cratcegus oxyacantho-germanica, Gillot, in Bull. Soc. Bot. France, xxiii. p. xiv. (1876). Pyrus lobata, Nicholson, Kew Hand-list Trees, 195 (1894) (not Koch). Cratœmespilusgrandiflora, Camus, in fourn. de Bot. xiii. 326 (1899). A tree, attaining about 25 ft. in height. Branchlets pubescent, with short aborted spines. Leaves very variable in shape, entire or three- to five-toothed near the apex, or five- to seven-lobed and finely serrate, pubescent beneath. Flowers large, white, fragrant like a hawthorn, solitary or two to five in a corymb ; calyx segments lanceolate, soon reflexed ; petals five ; stamens twenty ; styles one to four ; disc lobed. Fruit ovoid or globose, reddish brown, ^ in. in diameter, crowned by the sepals, with usually two or three nutlets, which are sterile. This tree, the origin of which is unknown, is considered by most botanists to be an accidental hybrid between Cratœgm oxyacantha and Mespilus germanica ; but Koehne1 considers it to be an independent species, possibly native of the Caucasus. Five apparently wild shrubs were found in 1875 at Saint-Sernin-du-Bois, near Autun, in Seine-et-Loire, in a hedge around the ruins of an old priory, by Dr. Gillot,2 whose interesting article should be studied. This remarkable tree, of which there is a good specimen3 at Kew, near the Director's office, was in cultivation at Paris about 1800 ; and possibly earlier in England, as Loudon mentions old trees at Syon and other places near London. 2. Two very remarkable graft hybrids4 originated about 1885 in the garden of M. Dardar, at Bronvaux, near Metz ; and have been propagated by Simon-Louis. On a very old medlar tree, that had been grafted on a stock of hawthorn, two peculiar branches 5 were observed to arise just beneath the graft. One of these branches, from which has been propagated the form known as Cratœgo-Mespilus Dardari? differed from the medlar in the branches being spiny, and the flowers in corymbs ; while the leaves and fruit were like those of the medlar but smaller. The other branch, which has been propagated as Cratœgo-Mespilus Asnieresi,6 was more like the haw thorn, the leaves being lobed and the flowers like Cratcegus monogyna in form and arrangement ; but the branchlets and leaves were pubescent as in the medlar. These two graft hybrids, which are now in cultivation at Kew, are said by Mr. Bean7 to be very different in appearance. The Asnieresi form has remained true to type, and is a small tree of great elegance and beauty. The Dardari form, according 1 Dent. Deiid. 230 (1893). 2 Cf. Rev. Hurt. Ixxi. 470 (1899), where Dr. Gillot states that it resembles Mesfihis more than Cratcegus, and is of undoubted hybrid origin. 3 There are several trees in the Green Park, London, and a fine one at Tortworth. *• The history of these graft hybrids has been given by Simon-Louis and by Liellair in Revue Horticole, Ixxi. 403, 482, 530 (1899) ; and by Koehne, in Gartenflora, 1. 628 (1901). R. P. Gregory, in Gard. Chron. 1. 185, fig. 86 (1911), gives Baur's explanation of their anatomical structure. 5 A third branch was subsequently produced on the original tree at Hronvaux, also from the junction of the stock and scion ; but on the opposite side to that occupied by the first two branches. It had at its base pure hawthorn ; but was trans formed towards the extremity into the Asnieresi form. 8 Jouin, in Le Jardin, 1899, P- 22- 7 Bean, in Kew Bull. 1911, p. 268, figs. I and 2. Cratsegus J733 to Mr. Bean, behaves to some extent like Laburnum Adami, and bore at Kew in 1911 three distinct kinds of foliage and flowers on the same specimen. One of its branches was like the Asnieresi form ; another branch was a pure medlar ; and all the other branches were the Dardari form. Neither of the two hybrids has as yet shown a branch of pure hawthorn. (A. H.) CRATŒGUS MONOGYNA, HAWTHORN, WHITETHORN Cratœgus monogyna, Jacquin, Fl. Austr. iii. 50, t. 292, fig. i (1775); Willkomm, Forstl. Flora, 835 (1887); Mathieu, Flore Forestière, 162 (1897); Schneider, Laubholzkunde, i. 781 (1906). Cratœgus oxyacantha? Linnasus, Sf. Pl. 477 (1753) (in part); Druce, List of British Plants, 26 (1908). Cratcegus oxyacantha, Linnaeus, var. monogyna, Loudon, Arb. et Frut. Brit. ii. 834 (1838). Cratcegus oxyacantha, Linnasus, sub-species monogyna, Hooker, Student's Flora, 127 (1870). Äfespilus monogyna, Allioni, Fl. Federn, ii. 141 (1785); Willdenow, Enum. Pl. Hort. Berol. i. 524 (1809); Ascherson and Graebner, Syn. Mitteleurop. Flora, vi. 2, p. 27 (1906). A shrub or small tree, attaining occasionally about 40 ft. in height. Bark greyish, broken on the surface into small scales. Young branchlets glabrous or with scattered pubescence. Leaves, variable in shape and size, usually broadly ovate, averaging if in. long, and \\ in. wide at the broad almost truncate or occasionally cuneate base ; pinnatifid with five or seven deep lobes, separated by narrow sinuses ; margin serrate ; upper surface with scattered white hairs ; lower surface pale green, with similar pubescence mainly on the midrib and nerves ; petiole slender, \ to i in. long, slightly pubescent. Flowers, variable in number in the corymb ; calyx-tube and pedicel pubescent ; sepals five, triangular, soon reflexed, persistent on the fruit ; petals five, usually white, but occasionally pink, even in the wild state ;2 stamens fifteen or twenty ; style one.3 Fruit, ellipsoid or ovoid, reddish, with one stone, which is either smooth or marked with shallow furrows. The seeds, when sown, do not germinate till the second year. The seedling4 has two obovate-oblong cotyledons, which are about | in. long, 3 in. broad, shortly stalked, glabrous, obscurely three-nerved, and raised above ground by a glabrous caulicle about i to i^ in. long. The pubescent stem bears alternate serrated leaves, the first three of which are small, cuneate, and three-lobed ; those succeeding becoming larger and deeply five-lobed. 1 C. oxyacantlia, Linnœus, included both species ; but this name was early limited to the two-styled species by Jacquin, who separated the one-styled species as C. monogyna. All botanists until lately have followed Jacquin's nomenclature of the two species, which is adopted by us. Recently much confusion has been caused by one or two writers, who restrict the name C. oxyacantha, Linnaeus, to the one-styled species. These authors are obliged to use another name, C. oxyacanthoides, Thuillier, for the two-styled species. 2 Briggs, Flora of riynwuth, 143 (1880), records a bush with pink flowers, and another with deep red flowers. In hedges near Cambridge shrubs with pink flowers are not uncommon. 3 C. kyrtostyla, Fingerhut, in Linnœa, iv. 372, t. iii. fig. I (1829), is a form of C. mcnogytia, in which the flowers have a peculiar curved or deflexed style. F. A. Lees, Flora of W. Yorkshire, 231 (1888), states that most old gnarled thorns in parks and pastures show this peculiarity. 4 Cf. Lubbock, Seedlings, i. 500, fig. 324 (1900), where the seedling of this species is described under the name C. oxyacantha. VII M 1734 The Trees of Great Britain and Ireland This species is more pubescent than C. oxyacantha, has a differently shaped leaf, and is readily distinguishable, apart from the single style, in the flower and solitary stone in the fruit. It is apparently much more variable in the wild state than the other species ; but some of the supposed varieties may be due to hybridising. Certain forms, which are plainly intermediate in various ways between the two species, are recognised as hybrids by continental botanists1 under the names : C media, Bech- stein, Diana, i. 88 (1797) ; and C. intermixta, Beck, Fl. Niederr. Oesterr. ii. i, p. 706 (1892). The commonest hybrid form in England has leaves like those of C. monogyna, but with flowers having a glabrous calyx-tube and pedicel. White states2 that there are several trees of this kind near Bristol, one of which on Leigh Down was noticed to be the last to bloom in 1881, a year remarkable for the abundant blossom of the hawthorn. In south-eastern and southern Europe there are peculiar races or allied species, which are not in cultivation in this country, and need not be more than alluded to.8 The following varieties have arisen in cultivation :— A. Differing from the type in habit? 1. Var. flexuosa, Dippel, Laubholzkunde, iii. 459 (1893). Cratœgus oxyacantha, var. flexuosa, Loudon, Arb. et Frut. Brit. ii. 835 (1838). Branches spinescent, dense, flexuose, twisted or like a corkscrew. This peculiar variety originated in Smith's nursery at Ayr, and is represented at Kew by a shrub about 15 ft. high. 2. Var. salisburifolia, Nicholson, Kew Hand-list Trees, 205 (1894). Cratœgus oxyacantha, var. salisburiœfolia, Späth, Cat., No. 59, p. 61 (1884). Branches similar to those of var. flexuosa, but without spines. Leaves with few and obtuse lobes somewhat like those of a Ginkgo tree in shape. This is repre sented at Kew by a shrub about 5 ft. high, which was planted in 1885. 3. Var. pendula, Dippel, Laubholzkunde, iii. 459 (1893). Cratœgus oxyacantha, var. pendula, Loudon, Arb. et Frut. Brit. ii. 832 (1838). Branchlets pendulous. Several forms are known. One is said by Loudon to have been picked out of a bed of seedlings at Somerford Hall. Anderson, curator of the Chelsea Botanic Garden in 1830, obtained several pendulous varieties by grafting shoots which were taken from the witches' brooms, that are occasionally 1 Focke, who describes the hybrid in Koch, Syn. Deutsch. Flora, i. 859 (1892), says that it is common in northern and central Germany, where it is more generally distributed in hedges and plantations than the true species. " Flora of Bristol, 300 (1912). 3 C. azarella, Grisebach, Spicil. Fl. Rum., i. 88 (1843) is a very pubescent form, which occurs in the Balkan States, Hungary, and Transylvania. C. hirsuta, Schur, Enum. PI. Trans. 206 (1866), is very similar and widely distributed in the Mediterranean region. The following are peculiar local forms :—C. Insegnœ, Bertolini, Fl. Ital. vii. 629 (1847), a native of Sicily ; C. grattalensis, Boissier, Elenchits, 41 (1838), wild on the Sierra Nevada in Spain ; and C. brcvispina, Kunze, in Flora, 1846, p. 737, a native of southern Spain. 1 In Rev. Hort. 1899, P- 489, it is stated that a form without spines was found in 1893 as a seedling in M. Hémeray- Proust's nursery at Orleans ; but I have not seen it in commerce. A compact dwarf spineless variety (var. inermis compacta) is advertised by Simon-Louis. Cratsegus found as conglomerations of slender branches on old trees.1 A weeping tree2 at Edinburgh, reputed to be over 300 years old and a favourite of Queen Mary, survived till 1836 ; and from it was propagated a form known as var. reginœ, or Queen Mary's Thorn. Var. pendula varïegata, a weeping form with variegated leaves ; and var. pendula rosea, a weeping shrub with pink flowers, are also in cultivation. 4. Var. ferox, Schneider, Laubholzkunde, i. 781 (1906). Cratœgus oxyacantha, var. ferox, Carrière, in Rev. Hort. 1859, p. 348. Cratœgus oxyacantha, var. horrida, Carrière, in Flore des Serres, xiv. 201, t. 1468 (1861); Regel, in Act. Hort. Petrop. i. 119 (1870); Lynch, in Gard. Chron. xxiv. 13, fig. 5 (1898). Branches pendulous, and armed with tufts of several spines. Carrière, who was unaware of the origin of this variety, states that seedlings raised from its seeds in the Jardin des Plantes at Paris reverted to the ordinary form. 5. Var. stricta, Nicholson, Kew Hand-list Trees, 205 (1894). Cratœgus oxyacantha, var. stricta, Loudon, Arb. et Frut. Brit. ii. 832 (1838). Fastigiate in habit, with upright branches. This was discovered in 1826 in a bed of seedlings in Ronalds's nursery at Brentford, and was said by Loudon to resemble the Lombardy poplar in shape. A tree so named at Kew, about 20 ft. high, is pyramidal, with ascending branches, but is not fastigiate. The truly fastigiate form is sold by Smith of Newry and by Späth of Berlin. 6. Var. ramulis aureis, Nicholson, Kew Hand-list Trees, 205 (1894). Var. xanthoclada, Zabel, ex Späth, Cat. No. 148, p. 91 (1911-1912). Branchlets of a bronze colour, conspicuous in winter. A shrub of this at Kew, obtained from Simon-Louis in 1885, is about nine feet high. B. Differing from tfie type in foliage. 7. Two variegated forms, mentioned by Loudon,—\zx.foliis argenteis, leaves mottled with white, and var. foliis aureis, leaves variegated with yellow,—are still in cultivation, but are of little ornamental value. 8. Var. laciniata, Dippel, Laubholzkunde, iii. 459 (1893). Cratœgus oxyacantha, var. laciniata, Loudon, Arb. et Frut. Brit. ii. 830 (1838); Regel, in Act. Hort. Petrop. i. 119 (1870). Leaves deeply pinnatifid, with irregularly serrated lobes. This occurs in hedgerows, in company with the typical form. F. A. Lees states 3 that in York shire, Worcestershire, Lincolnshire, and Berkshire, it flowers much less freely than the other forms. 9. \zx.ßlicifolia, Koehne, Deut. Dend. 238 (1893). Cratœgus oxyacantha, var.ßlicifolia, Van Houtte, Flore des Serres, xx. 51, t. 2076 (1874). Leaves broad, fan-shaped, deeply divided into numerous curled segments. This beautiful variety, which resembles Adiantum farleyense in foliage, does not 1 Cf. Loudon, Gard. Mag. ix. 596 (1833). 2 Figured by Loudon, Arb. et Frut. Brit. ii. 833,'%. 556 (1838). 3 Flora W. Yorkshire, 231 (1888). It is recorded for Warwickshire by Bagnall, Flora of Warwickshire, 107 (1891). 1736 The Trees of Great Britain and Ireland seem to be in cultivation1 in England. There are several other varieties2 with peculiar cut leaves, none of which I have seen. C. Differing from the type in flowers. IG. Var. semperflorens, Dippel, Laubkolzkunde, iii. 460 (1893). Cratœgus oxyacantha, var. semperflorens, André, in Rev. Hort. liv. 354 (1882) and Iv. 140, fig. 26 (1883). A low bushy shrub, which numerous flowers, which appear more or less con tinuously throughout the season from May to autumn. This was found about 1879 at Poitiers by M. Bruant, in a bed of seedlings of the common hawthorn, and was subsequently propagated by grafting. A shrub at Kew is about 2 ft. high. 11. Var. prœcox, Dippel, Laubholzkunde, iii. 459 (1893). Glastonbury Thorn.3 Cratœgus oxyacantha, var. prœcox, Loudon, Arb. et Prut. Brit. ii. 833(1833). Flowers, usually appearing at Christmas or early in January, and not ripening into fruit, the leaves being somewhat later ; a second crop of flowers, which produce fruit, is borne in May and June. The original tree/ which grew at Glastonbury, was mentioned by Turner in 1562, and by Gerard in 1597, and had the appearance of a very old tree, when it was seen by Withering5 in 1793. From this tree the variety was propagated. A specimen, growing near the Temperate House at Kew, is irregular in the time of flowering, which depends upon the nature of the season. Bean states6 that, with a mild November and December, the tree at Kew will flower about Old Christmas Day (6th January) ; but if cold weather sets in before New Year, the flowers may not open till March or April. In 1908, owing to the unusual warmth of the autumn, it was in full blossom in the first week of Nov ember, before the leaves had fallen, so that it was carrying flowers, fruit (derived from the flowers of the preceding May), and foliage simultaneously. D. Differing from the type in fruit. 12. Var. eriocarpa, Dippel, Laubholzkunde, iii. 460 (1893). Cratœgus oxyacantha, var. eriocarpa, Loudon, Arb. et Frut. Brit. ii. 831 (1838). 1 Var. flKcifolia is advertised by Späth, Cat. No. 148, p. 91 (1911-12). 2 Var. pinnatiloba, Dippel, Laubhohkunde, iii. 458 (1893), 's identified by Schneider, Laubhohkunde, i. 785 (1906), with C. microphylla, Koch, Die Weissdorn, 68 ( 1884), which is a Caucasian species, not apparently in cultivation, though it is given in the Kew Hand-list. Koehne, jDeut. Dendr. 238 (1893), however, considers it to be a hybrid between C. monogyna and C. oxyacantha. 3 Chevallier, in Ann. Soc. Agric. Sei. Départ. Indre et Loire, xxx. 7° (I8So), describes a similar sport of Prunus spinosa, to which is attached a similar legend. At the Château du Chabrol, St. Patrice, on the Loire, midway between Saumur and Tours, there is a large blackthorn, called Vépine miraculeuse, which flowers every year in the last week in December, even in the severest seasons. The legend is that St. Patrick, while on his way to Tours in A.D. 395, reposed one night in winter under the shade of this tree, which burst forth into flowers and leaves to shield him from the cold. The tree did not appear to be very old in 1850, but is now of considerable size, judging from a photograph sent me by M. Hickel. This curious variety, which may be named Prunus spinosa, var. prœcox, does not seem ever to have been propagated. 4 Parkinson, Theat. 1025 (1640), mentions other trees of the same kind at Romney Marsh and at Nantwich. Plot, Nat. Hist. Oxfordshire, 159 (1705), mentions a very old tree, which flowered at Christinas, in Lord Norrey's park in Oxfordshire ; but was uncertain whether it was a graft from the Glastonbury tree or an original specimen of the variety. 6 Arr. Brit. PL 459 (1793). Cf. Loudon, Gard. Mag. ix. 122 (1833). However, James Howel, Dodond's Grove, 55 (1644), implies that the original thorn was destroyed by Puritan fanatics, one of whom "was wel served for his blind zeale, who, going to cut doune an ancient white Hawthorne tree, which, because she budded before others, might be an occasion of Superstition, had some of the prickles flew into his eye and made him Monocular." 6 Kew Bull. 1908, p. 452. Cf. also J. W. White, Flora of Bristol, 302 (1912), who mentions a specimen at Ipswich, which flowered I4th November 1885, and another at Evesham, which was in flower on 26lh November, 1899. Cratasgus Fruit woolly pubescent. This variety is rare ; but is occasionally found in the wild state, as in the Isle of Wight, where it is recorded by Bromfield.1 It is said2 to occur near Breslau, in Silesia. 13. Var. maurianensis, Didier, in Bull. Soc. Dauph. ix. 385 (1882). Fruits very large, \ to i in. long, and | in. or more wide. This variety was described from specimens found in Savoy, and occurs also in hedges near Toulon.3 Mr. J. W. White records4 two trees, with branches bending down under the weight of numerous large haws, which were found in 1909 near Bristol, one on a low cliff near Walton-by-Clevedon, the other in Chelvey Batch wood. The fruits were very handsome, and four times larger than the typical form, averaging \ in. long and \ in. wide. As both C. monogyna and C. oxyacantha have been much confused, it is impossible to give an accurate account of their separate distribution. The common hawthorn, comprising both species, is a native of Europe, and of the mountains of Algeria and Morocco; and extends from Asia Minor and the Caucasus, through Armenia, Persia, and Afghanistan to the western Himalayas, where it grows between 6000 and 9000 feet elevation.5 It grows wild in Norway as far north as lat. 62° 55', in Sweden as far as Upsala, lat. 59° 52', and in Finland to lat. 61° 30'. In Russia, it is common in Livland, Kazan, and Orenburg, and throughout the southern provinces except in the Steppes. It occurs mainly in hedges, waste places, and on the margins of woods, ascending in the Alps to about 3000 feet altitude. C. monogyna is by far the commoner of the two species in Britain, where it is found in hedges and woods from Moray and Islay southwards ; and it is met with in all districts in Ireland. It is the most commonly planted hawthorn either for hedges or for ornament ; and most of the large trees in parks are referable to this species, though some of them look as if of hybrid origin. The hawthorn lives to a great age, probably to three or four hundred years.6 Old trees often grow irregularly, so that ribs are formed upon their stem, which assume a vertical or a spiral direction. As years go on, these rib-like projections become larger, and the intervening channels deeper. Ultimately, when decay begins at the heart and spreads outwards, the projecting parts become separated and appear to be a number of subordinate stems, which are, however, united at the base, and bear on their inner surface, instead of bark, remains of the decayed heart- wood.7 (A. H.) 1 In Phytologist, iii. 288 (1848). 2 Ascherson and Graebner, Syn. Mitteleurop. Flora, vi. pt. ii. 30 (1906). 3 Described as var. macrocarpa, Reynier, ex Albert and Jahandiez, Plantes Vase, du Var, 185 (1908). 1 Flora of Bristol, 300 (1912), where these two trees are assigned to var. splendent, Druce, a name which cannot be retained, as var. splendens, Koch, Dendrologie, i. 159 (1869), much earlier in date, was applied to forms with pink and scarlet flowers. Mr. White identifies this large-fruited variety with Oxyacanthits folio et friictu majore from Oxfordshire of Merrett's Pinax (1667). Ray, Syn. 454 (1724), states that it was also found by Sherard in Northamptonshire. 6 The forms in southern Europe, northern Africa, and Asia, are apparently very distinct varieties or allied species ; and require further study. Cf. Schneider, Laubhohkunde, i. 782 (1906). 6 Lees, in Gard. Chron. iii. 688 (1875), states that he counted over 300 annual rings in the stem of a tree about I ft. in diameter, that grew on the Malvern Hills. 7 Cf. Purchas, \njourn. of Bot. iii. 366 (1865), who points out that stems growing close together, which have commenced as independent trees in a hedge, are surrounded on all sides by bark, and are thus readily distinguishable from the peculiar stems described above. Lees, in Card. Chron. iii. 688, figs. 142, 143 (1875), gives illustrations of trees with divided stems at Garnstone, Herefordshire, and at Upper Wyck, near Worcester. 1738 The Trees of Great Britain and Ireland CULTIVATION The hawthorn, though it grows on almost all soils, succeeds best in a rich loam, and does well on strong clay. It is propagated by seeds, which as they lie over for a year, should be mixed with ordinary soil or sand in pits or heaps, where they are kept until the second year following, when they may be sown in Feb ruary or March. Young plants should be removed from the seed bed at the end of the first year, and after standing in nursery lines for two years, be planted out. The varieties are budded or grafted on seedlings of the common species. The most important use of the hawthorn is as a hedge plant. Though hedges appear to have been in use in England from the time of the Romans, they were not generally planted to enclose ordinary fields and meadows till about the end of the seventeenth century. Dr. Walker statesl that the first hedges in Scotland were planted by Cromwell's soldiers in East Lothian and Perthshire. Fine old hawthorns, with trunks of great girth and wide-spreading branches, exist in many parks throughout the country ; and tall slender specimens are occasionally seen, drawn up in woods. We may mention a few, remarkable for age, though doubtless there are many quite as large that we have not heard of. An immense old thorn, at Hethel in Norfolk, was first mentioned by Marsham, who, in a letter to the Bath Society about 1740, made its girth 9 ft. i in. at four feet from the ground. Grigor 2 states on the authority of H. Gurney, Esq., that the first Sir Thomas Beevor, who owned the place towards the end of the eighteenth century, put a railing round it, which was subsequently repaired, and the spreading limbs propped up by Mr. Gurney. Grigor says that the trunk measured 12 ft. i in. at one foot, and 14 ft. 3 in. at five feet, whilst the branches, though several large ones had been lost, spread over an area 31 yards round. Both the trunk and the large branches were then hollow, but the wood sound and hard. Sir Hugh Beevor in 1895 found it to be 13^ ft. at eighteen inches from the ground where the girth was least. Miss Eaton sent me a photograph in 1903 which showed branches supported by numerous props. Mr. Edwards, who photographed the tree in September 1912, tells me that the tree now consists of several stems formed by the splitting of the original trunk. The branches, which now measure 37 yards round, are sound and covered with leaves and fruit, though bearing many tufts of mistletoe. It is protected from cattle by a rail ; and the branches are supported by numerous props (Plate No. 378). In the park at Hoi wood House, Kent, Mr. A. D. Webster records3 a tree, which in 1888 was 14 ft. 6 in. in girth at three feet from the ground, above which it divided into six limbs, measuring at a yard from the fork, 4 ft. 2 in., 4 ft., 5 ft. 8 in., 2 ft. 8 in., 4 ft. 4 in., and 3 ft. 5 in. respectively. Its height was 42 ft., with a spread of branches 63 ft. in diameter. This tree was growing in strong clayey loam, and was in perfect health. Mr. Edwin Lees described4 a remarkable hawthorn at Lenchford, in Worcester- 1 Essays on Nat. Hist. 54 (1812). 3 Trans. Roy. Scot. Arbor. Soc. xii. 311 (1889). 2 Eastern Arboretum, 282 (1841). 4 Card. Cftran. iii. 688, figs. 141, 146 (1875). Cratsegus shire, which was 60 ft. high and 9 ft. in girth in 1875 ; and another in Downton Park, Herefordshire, which was over 50 ft. in height. He also mentioned others of con siderable size in the same counties. At Chideock Manor, Dorset, there is a remarkable thorn, about 25 ft. high and 6 ft. in girth, with wide-spreading branches and very pendulous branchlets, the diameter of the spread being about 40 ft. In Lilford Park, Northamptonshire, where there are many fine thorns, I measured, in 1906, a tree no less than 51 ft. in height, but only 5 ft. in girth. In the pleasure grounds of Hatfield House, Herts, there is a tree, which Henry measured in 1911 as 48 ft. by 3 ft. 8 in. In Ware Park, in the same county, Mr. H. Clinton Baker found in 1910 a tree 30 ft. by 12 ft. 5 in. Within the walls of Rothesay Castle, in the island of Bute, there was growing in 1878 a remarkable thorn which Mr. James Kay described1 as follows: " Though the tree was blown down thirty-nine years ago, it is still vigorous and healthy. The extreme length of the tree as it now lies, measuring from the original surface of the root, is 47 ft. ; present vertical height, 28 ft. ; circumference, three feet up, 6 ft. 8^ in. ; four and a half feet up, 6 ft. 6J in. ; six feet up, 6 ft. 10 in." This tree dates from some time after 1685, when the Castle was burnt ; and its age in 1878 did not probably exceed 190 years. TIMBER The wood of the hawthorn is white, often tinged with red ; and is hard, heavy, and difficult to work, but with a fine grain and susceptible of a good polish. It is not much used, as it seldom can be obtained of sufficient size, and is usually spoiled by defects or knots. It is occasionally employed by turners, and was formerly found suitable for teeth of mill-wheels. Mr. W. G. Smith, who made many woodcuts for the Gardeners' Chronicle, states2 that hawthorn wood is quite as good for engraving as ordinary boxwood, and possesses a far better colour. The best box, however, cuts a little smoother, as it has a somewhat closer grain. (H. J. E.) CRATSEGUS OXYACANTHA, HAWTHORN, WHITETHORN Cratœgus oxyacantfia? Linnseus, Sp. PL 477 (1753) (in part); Jacquin, Fl. Austr. iii. 50, t. 292, fig- 2 (!77S); Loudon, Arb. et Frut. Brit. ii. 829 (1838); Willkomm, Forst!. Flora, 838 (1887); Mathieu, Flore Forestière, 163 (1897); Schneider, Laubholzkunde, i. 780 (1906). Cratœgus oxyacantha, var. vulgaris, De Candolle, Prod. ii. 628 (1825). Cratœgns oxyacantha, sub-species oxyacanthoides, Hooker, Student's Flora, 127 (1870). Cratœgus oxyacanthoides, Thuillier, Fl. Envir. Paris, 245 (1799). Mespilus oxyacantha, Crantz, Stirp. Austr. i. 39 (1763); Allioni, Fl. Pedem. ii. 141 (1785); Willdenow, Enurn. PL Hort. Berol. i. 524 (1809); Ascherson and Graebner, Syn. Mitteleurop. Flora, vi. 2, p. 25 (1906). A shrub or small tree, similar to C. monogyna in bark and habit. Young branchlets glabrous. Leaves obovate or ovate, usually with three shallow lobes, the terminal 1 Trans. Roy. Scot. Arbor. Soc. ix. 76 (1879). 2 In GarJ. Chron. iii. 689, note (1875), it is stated that figures 142 and 143 on p. 688 were engraved on hawthorn wood. 3 Cf. p. 1733, note i. 174° The Trees of Great Britain and Ireland one being often tabulate ; margin irregularly serrate ; upper surface dark green, shining, glabrous ; lower surface pale green, glabrescent or slightly pubescent on the midrib ; petiole slender, usually glabrous. Flowers, variable in number in the corymb ; peduncle, pedicel, and calyx-tube glabrous ; sepals five, triangular, spreading, persistent on the fruit ; petals five, white ; stamens fifteen or twenty ; styles usually two, rarely in some flowers one or three. Fruit ovoid or globose ; stones two or three, rarely one, flattened on the inner surface, convex with deep longitudinal furrows on the outer surface. This species is very distinct in appearance, and in England as a rule comes into flower a fortnight earlier than C. monogyna. VARIETIES This species is apparently much less variable in the wild state * than C. monogyna, but has given rise to some remarkable garden forms. 1. Var. multiplex, Loudon, Arb. et Frut. Brit. ii. 832 (1838). Var. flore plena albo, Rodigas, in Flore des Serres, xv. t. 1509, fig. 2 (1862). Cratœgus monogyna, var. alba plena, Rehder, in Bailey, Cyt. Amer. Hort. i. 396 (1900). Flowers white, double, produced in great profusion and dying off a beautiful pink colour.2 This variety, the origin of which is unknown, differs little from the type in other respects, having two-styled flowers with glabrous pedicels and calyx- tubes, and glabrescent three-lobed leaves.8 2. Var. rosea, Loudon, Arb. et Frut. Brit. ii. 832 (1838). Petals pink, with white claws ; in other respects similar to the type. This is a single pink variety, which is occasionally found wild. 3. Var. punicea, Loddiges, Bot. Cat. t. 1363 (1828); Loudon, Arb. et Friit. Brit. ii. 832 (1838). Var. ßorepuniceo, Rodigas, in Flore des Serres, xv. t. 1509, fig. 2 (1862). Petals larger than in var. rosea, dark red and without white on the claws. This is the handsome single crimson variety, which was first raised in Scotland, and afterwards propagated by Loddiges, who budded it upon the common whitethorn. 4. Var. Gumpperii bicolor, Van Houtte, in FI. des Serres, xvi. t. 1651 (1866). Flowers single ; petals white, edged with a pink margin. This originated at Stuttgart about 1860, and is probably of hybrid origin, as the flowers, while having a glabrous calyx-tube, bear only one style. 1 Var. integrifolia, Wallroth, Sched. Crit. 219 (1822), a name given to shrubs having obovate leaves with three very shallow lobes, can scarcely be retained as a variety, as such leaves are common in the typical form of the species. Similarly, var. auriculata, Dippel, Laubhokkunde, iii. 457 (1893), said to have persistent large stipules, is doubtfully distinct, as the retention of the stipules depends on the vigour of the branches, and is common enough in the ordinary form of the species. 2 Späth, Cat. No. 148, p. 91 (1911-1912), advertises var. candido-pkna, a new variety with double flowers that remain pure white. 3 M'Nab, in Trans. Bot. Soc. Edin. vi. 284 (1860) gives an account of a tree with double white flowers in the Edinburgh Botanic Garden, which retained most of its leaves during the preceding winter, some remaining green till izth May 1859. It then flowered and produced normal single flowers. Crataegus 1741 5. Var. pwiicea flore pleno, Loudon, Arb. et Frut. Brit. ii. 832 (1838), and Trees and Shrubs, 377 (1842). Var.ßore ritbro pleno, Rodigas, in Flore des Serres, xv. t. 1509, fig. 3 (1862). Flowers double, pink. This is said by Loudon to have been imported about 1832 by Masters of Canterbury, and to bear flowers not so brilliant in colour as the single crimson variety. It differs from typical C. oxyacantha in the pubescent leaves and calyx-tube, and is possibly a hybrid. 6. Var. coccinea flore pleno, Paul, in Florist and Pomologist, vi. 117 (1867). Var. floribus coccineisplenis, Lemaire, in Illust. Hort. t. 536 (1867). Cratœgus monogyna, var. fauli, Render, in Bailey, Cyc. Amer. Hort. \. 396 (1900). Flowers double, deep scarlet. This originated about 1858 in Mr. Christopher Boyd's garden near Waltham Cross, as a single branch on a tree of the double pink variety, which was about 25 years old and nearly 30 feet high. This branch was observed to bear flowers of a deep scarlet colour, year after year, whilst all the other branches on the tree continued to produce flowers of the original pink colour. It was propagated by Messrs. Paul, who showed it at the International Horticultural Exhibition in 1866, under the name "Paul's New Double Scarlet Hawthorn," by which it is still known. 7. Var. Gireoudi, Späth, Cat. No. 104, p. 89 (1899-1900). Young shoots pink, bearing new leaves, which are mottled with white and pink. This is represented at Kew by a spreading bush about 6 ft. high, received from Späth in 1899. 8. Var. aurea, Loudon, Arb. et Frut. Brit. ii. 831 (1838). Var. xanthocarpa, Lange, Rev. Spec. Cratagi, 71 (1897). Fruits yellow. This variety, which has been in cultivation over a hundred years, bears freely and is very showy. 9. Var. leucocarpa, Loudon, Arb. et Frut. Brit. ii. 831 (1838). Fruit white. Plot, Nat. Hist. Oxfordshire, 158 (1705), mentions1 a tree with white haws in a hedge near Bampton ; but it does not appear to have been propagated, and this variety was unknown to Loudon. The following is either a geographical form or a closely allied species :— IQ. Cratœguspolyacantha, Jan, Elench. Hort. Farm. 8 (1826). Cratœgus oxyacantha, sub-species monogyna, var. polyacantha, Nicholson, Kew Hand-list Trees, 343 (1902). Leaves small, about f in. long and broad, tri-lobed. Young branchlets, petioles, peduncles, pedicels, and calyx-tubes densely covered with white woolly pubescence. This is a small shrub, native of Sicily and Calabria. It is in cultivation at Kew. 1 Cf. Ray, Syn. Meth. 453 (1724). A form with fruit larger than usual, occurring in the south of France and Switzer land, has been distinguished as var. macrocarpa. Le Grand, Stat. Bot. Forez, 119 (1873), ex Rouy and Camus, Flore de France, vii. 4 (1901), identical with C. macrocarpa, Hegetschweiler, Fl. Schweiz, 464 (1840). VII N I742 The Trees of Great Britain and Ireland DISTRIBUTION C. oxyacantha is widely spread throughout Europe, occupying nearly the same territory as C. tnonogyna, but being much less common. Willkomm believes it to be more prevalent in the north than in the south of Europe. In France, according to Mathieu, it does not grow in the region of the olive ; while elsewhere it is comparatively rare, and scarcely ever attains the dimensions of a tree. I saw it in 1912 in the interior of the Forest of Orleans, forming a large shrub, and growing in shade, as it does in the Gamlingay Wood, near Cambridge. It is doubtfully wild in Scotland and Ireland; and is apparently indigenous only in the midland, eastern, and south-eastern counties of England. It is recorded for many stations in Kent, Surrey, Middlesex, Oxfordshire, and Cambridgeshire. It is not mentioned as a native plant for the northern or western counties in any of the published floras in which the two species are distinguished. In England, as far as we know, it is usually a shrub ; and we have no records of any large trees of this species. (A. H.) SALIX Salix, Linnaeus, Sp. PI. 1015 (1753); Forbes, Salic. Woburn., 1-294 (1829); Andersson, Monog. Salic. 1-180 (1863), and in De Candolle, Prod. xvi. 2, p. 191 (1868); Bentham et Hooker, Gen. PI. iii. 411 (1880); Buchanan White, mjourn. Linn. Soc. (Bot.} xxvii. 333-457 (1890); Camus, Saules d'Europe, 9-40 (1904); Schneider, Laubholzkunde, i. 23 (1904). TREES or shrubs, with scaly bark, and slender terete branchlets, which are often easily separated at the joints. True terminal buds are not developed,1 the top of the branchlet dying off in summer, and leaving a minute scar close to the uppermost axillary bud, which prolongs the branch in the following season ; buds apparently covered by one scale, which is composed of two scale-like leaves fused together, as indicated by their keeled margins. Leaves deciduous,2 alternate, rarely sub-opposite, simple, variable in shape, penni-nerved, stalked ; stipules oblique, serrate, either small and early deciduous, or large, leafy, and persistent, often conspicuous on barren vigor ous young branches. On branches from which the leaves have fallen the leaf-scars are crescentic and ß-dotted, and accompanied on each side by a minute stipular scar. Flowers, appearing in some species before the leaves, in others after the leaves, dioecious, in catkins, each of which terminates a short shoot, and bears numerous flowers on a slender axis ; each flower, with one or two honey-glands, placed front and back at its base, and subtended by a scale, which is usually entire in margin. Staminate flowers, with two or three to twelve stamens, inserted on the base of the scale, with slender filaments, free or more or less connate, and two-celled anthers opening longitudinally. Pistillate flowers ; ovary free, stalked or sessile, one-celled, with four to eight ovules on each of the two placentae ; crowned by a style, which is often extremely short or obsolete, with two stigmas, which are either entire or bifid. Fruit, an acuminate capsule, separating when ripe into two recurved valves. Seeds minute, narrowed at the ends, dark brown or nearly black, furnished with long silky hairs. About 160 species of Salix are known, distributed from the Arctic regions southwards to the Andes of Chile in the New World, and to South Africa, Mada gascar, Himalayas, Burma, Malay Peninsula, Java, and Sumatra in the Old World. Most of the species are shrubs or small trees; and in the following account we have only dealt with the few species in cultivation which attain a large size, including S. Caprea, on account of its interest to foresters. These may be distinguished as follows :— 1 However, in § Chamatia, Dumortier, in Rijdr. Natuurk. Wetemch. i. 56 (1826), which includes Arctic and Alpine under-shrubs, like S. reticulata, Linnieus, there are true terminal buds, giving rise to catkins in the following year. The species of this section in some respects are intermediate between Salix and Pofulus. 2 S. Bonflandiana, II. B. K., a native of Mexico, is said to have persistent leaves. Cf. Dode, in Ball. See. Dend. France, 1909, p. 151. 1743 1744 The Trees of Great Britain and Ireland I. Leaves ovate or oval. 1. Salix Caprea, Linnaeus. Seep. 1745. Leaves oval, elliptic or ovate, 2 to 3 in. long, i^ in. broad, irregularly crenate ; lower surface bluish grey, tomentose. 2. Salixpentandra, Linnaeus. See p. 1747. Leaves fragrant when bruised, ovate to ovate-lanceolate, about 3 in. long and ij in. broad, abruptly acuminate, minutely serrate, glabrous on both surfaces. II. Leaves lanceolate. A. Mature leaves not ciliate in margin, green above, green or glaucous beneath. * Leaves broadly lanceolate, with coarse serrations, which are not close together. 3. Salix speciosa, Host. Seep. 1756. Leaves, broader than in S.fragilis, 4 to 6 in. long, i£ to \\ in. broad, glabrous except for a few scattered hairs on the slightly glaucous under surface. Young branchlets pubescent or glabrescent. 4. Salix fragilis, Linnaeus. See p. 1754. Leaves about 4 in. long, f to i in. broad, glabrous except for a few scattered hairs on the glaucous under surface. Young branchlets pubescent. ** Leaves narrowly lanceolate, with fine close serrations. (a) Young branchlets crimson, glabrous ; clay-coloured in the second year. 5. Salix decipiens, Hoffmann. Seep. 1755. Leaves, 2 to 3 in. long, \ to f in. broad, similar to those of S. fragilis, but smaller and green beneath. (b) Young branchlets green, becoming a brilliant yellow in winter and tlte following year. 6. Salixvitellina, Linnaeus. Seep. 1768. Leaves, 2 to 2^ in. long, f to \ in. broad, green above, glaucous beneath ; with scattered appressed hairs, sparse above, more abundant beneath. See No. IIA. (c) Young branchlets green, becoming olive green or brownish grey in the second year. 7. Salix babylonica, Linnaeus. See p. 1749. Leaves, z\ to 4 in. long, \ to f in. wide, tapering to a long slender filamentous apex; when mature, glabrous and glaucous beneath. Branchlets pendulous, always injured in England by spring frost. 8. Salix elegantissima, Koch. Seep. 1751. Leaves similar to those of S. babylonica, but broader, and more coriaceous, and less glaucous on the lower surface. Branchlets pendulous, uninjured in England by spring frost. 9. Salix Salamonii, Carrière. See p. 1750. Leaves similar to those of S. babylonica, but with scattered appressed long hairs on both surfaces. Branches ascending, with pendulous branchlets, which are not injured by frost in England. Salix B. Mature leaves ciliate in margin, covered more or less on both surfaces with appressed silky hairs. 10. Salix alba, Linnaeus. Seep. 1759. Leaves 2 to 2^ in. long, f to ^ in. wide, covered with silky pubescence, densest on the whitish under surface. A wide-spreading male or female tree, with pendulous or spreading branchlets ; ovary and fruit sessile. 11. Salix c cerulea, Smith. Seep. 1763. Leaves similar to those of S. alba, but thinner in texture, more translucent, less pubescent, the lower surface being bluish grey and not white. A pyramidal female tree, with ascending branches and erect terminal branchlets ; ovary and fruit shortly pedicellate, i IA. Salix vitellina, Linnaeus. See No. 6. Leaves occasionally ciliate till autumn, with both surfaces more or less appressed-pubescent. This is readily distinguishable by the bright yellow branchlets in winter. (A. H.) SALIX CAPREA, SALLOW, GOAT WILLOW Salix Caprea, Linnaeus, Sp. PL 1020 (1753); Loudon, Arb. et Frut. Brit. iii. 1561 (1838); Anders- son, Monog. Salic. 75 (1863), and in De Candolle, Prod. xvi. 2, p. 222 (1868); Willkomm, forstliche flora, 487 (1887); Buchanan White, in Journ. Linn. Soc. (Hot.) xxvii. 385 (1890); Mathieu, flore forestière, 465 (1897); Camus, Monog. des Saules, 202 (1904). A small tree, occasionally attaining 40 ft. in height. Bark smooth and greenish at first, ultimately ridged and fissured. Young branchlets, with a minute tomentum, becoming more or less glabrescent in the second year. Buds ovoid-conic, minutely tomentose or glabrous. Leaves oval or elliptic, about 2 to 3 in. long and i£ to i J in. broad ; apex acuminate, with the point usually directed to one side ; base broadly cuneate, often unequal ; upper surface light green, slightly shining, glabrous except for slight pubescence on the midrib and nerves ; lower surface bluish grey, reticulate, covered with a whitish tomentum ; margin irregularly crenulate, undulate, or rarely almost entire ; petiole more or less tomentose ; stipules oblique, reniform or half cordate, dentate. Flowers appearing early, before the leaves ; catkins sub-sessile, with scale-like leaves at the base, very silky, on account of. the spatulate scales, brown or blackish towards the apex, and pubescent with long hairs. Staminate flowers ; stamens two, free, glabrous. Pistillate flowers ; ovary tomentose on a long pedicel, much surpass ing the gland in length ; style short, with two stigmas, which are usually emarginate or rarely bifid. Capsules on long pedicels, narrow, elongated, covered with greyish silky hairs. VARIETIES AND HYBRIDS I. S. Caprea is variable in the wild state. The catkins, which are usually sessile or sub-sessile, are occasionally provided with leafy peduncles. The branchlets and 1746 The Trees of Great Britain and Ireland buds, which are normally glabrous, are occasionally pubescent. Varieties, differing in the leaves, have been distinguished :— 1. Var. orbiculata, Kerner, in Verh. Zool. Bot. Ges. Wien, x. 248 (1860). This is the typical form of the species, in which the leaves are broadly oval, slightly cordate at the base, and with the apex bent to one side. 2. Var. elliptica, Kerner, loc. cit. Leaves elliptic and tapering at both ends. This is said to be the more common form in northern and mountainous districts. 3. Var. sphacelata, Wahlenberg, Fl. Carpat. 319 (1814). Var. alpina, Gaudin, Fl. Helv. vi. 240 (1830). Salix sphacelata, Smith, Brit. Bot. iii. 1066 (1805), Eng. Bot. t. 2333 (1812), and Eng. Flora, iv. 224 (1828); Loudon, Art. et Frut. Brit. iii. 1563 (1838). Leaves small, tomentose on both surfaces. This is an alpine form, which occurs in the Highlands of Scotland. 4. Var. pendula, Petzold and Kirchner, Arb. Muse. 576 (1864). Salix Kilmarnocki, Nicholson, Kew Hand-list Trees, ii. 213 (1896). Pendulous in habit, usually grafted on a stock about 4 ft. high, and forming a weeping shrub, which is known as the Kilmarnock Willow. This was discovered* in 1840 on the banks of the river Ayr, and was propagated by Lang, nurseryman at Kilmarnock. The original tree died of old age about 1884. II. S. Caprea is closely related to both S. cinerea and S. aurita, which are bushy species. Intermediate forms, which are often difficult to discriminate, have been referred to the following hybrids :— 5. Salix Reichardtii, Kerner, in Verh. Zool. Bot. Ges. Wien, x. 249 (1860). This hybrid between S. Caprea and S. cincrca is rare, only a few examples from Perthshire, Fifeshire, Worcestershire, and Kent being recorded by Buchanan White. Both parents are common ; but they do not flower at the same time, and are rarely met with together, as S. Caprea is most frequent in woods, whilst S. cinerea grows chiefly on river banks. 6. Salixcapreola, Kerner, ex Andersson in De Candolle./Vöd'. xvi. 2, p. 223 (1868). This hybrid between S. Caprea and S. aurita is also rare, as the periods of flowering of the two species are not identical. There are pistillate specimens at Kew from Derbyshire and Surrey ; and Buchanan White mentions other localities in Perth, Worcestershire, and Somerset. DISTRIBUTION The common sallow or goat willow is widely distributed, occurring in Europe, the Caucasus, Siberia, Amurland, Manchuria, and Korea. In Europe it exists in every country, extending as far north as Iceland and Lapland, and as far south as southern Spain, Italy, and Greece. It is most commonly found in woods on the plains and lower hilly regions, but reaches 3300 feet elevation in Norway and 4500 feet in the Carpathians. It grows on almost all soils, on those which are moist, marshy, or even peaty, as well as in dry rocky or stony ground. It attains its largest 1 Cf. Rev. Dr. Lanclsborough, in Ann. Kilmarnock Glenfield Raniblerf Sec., 1893-1894, p. so. Salix size in East Prussia, Lithuania, and the Russian Baltic provinces, where, according to Willkomm, it is often a fine tree, 30 to 50 ft. in height. It is naturally regenerated by seed, and when cut down produces vigorous coppice shoots. It can be propagated by cuttings and by sets. (A. H.) The sallow is common in all parts of the British Isles, growing in woods and copses, and in waste places, but rarely attains a large size. It is often 20 to 30 ft. in height, and sometimes produces a trunk a foot in diameter, but seems to be a short lived tree. The finest which I have seen is growing by the roadside two miles below the lodge at Guisachan, Inverness-shire, and measured in 1910 about 50 ft. by 6 feet. It is the only willow which commonly grows from seed, and in some of my planta tions is so abundant that it may be called a forest weed. It is usually looked upon by foresters as a useless tree ; but it has proved valuable in fixing loose and shifting soil on river embankments and similar situations, as it is so readily propagated by cuttings. Mitchell1 says : " It is the best underwood that we have. It makes good fences, and sheep hurdles made of it will always last a year or two longer than those made of hazel ; and no soil or situation comes wrong to it, wet or dry." In Sussex it is used for making truck-baskets and handles of rakes, and also for fencing, as it is light and tough, and splits easily.2 Though the wood is of a nice pinkish colour, it is too small, as a rule, to have any recognised value. In northern Russia (as well as formerly in Scotland) the bark, which contains 7 per cent of tannin, is sometimes used for tanning leather. (H. J. E.) SALIX PENTANDRA, BAY WILLOW Salixpentandra, Linnseus, Sp. PL 1016 (1753) ; Loudon, Arb. et Frut. Brit. iii. 1503 (1838) ; Andersson, Monog. Salic. 35 (1863), and in De Candolle, Prod. xvi. 2, p. 206 (1868); Willkomm, Forstliche Flora, 475 (1887); Buchanan White, \r\Journ. Linn. Soc. (-Dot.) xxvii. 359 (1890) ; Mathieu, Flora Forestière, 449 (1897); Camus, Monog. des Sattles, 84 (1904). Salix fragrans, Salisbury, Prod. 393 (1796). A tree, occasionally attaining 30 to 40 ft. in height, but often shrubby in the wild state. Young branchlets glabrous, dark brown, shining as if varnished. Buds ovoid, pointed, dark brown, shining, viscid. Leaves fragrant when bruised, ovate, ovate-oblong, or ovate-lanceolate, averaging 3 in. long and i\ in. broad, glutinous when young, coriaceous when fully grown ; rounded at the base, abruptly acuminate at the apex ; glabrous on both surfaces, dark green and very shining above, pale and dull beneath ; margin closely and finely serrate, the serrations tipped with dark red glands ; petiole, about \ inch long, glabrous, with two or three glands near its junction with the blade, and expanded at its origin from the branchlet, where there are one or two glands probably representing stipules. Catkins, appearing with or after the leaves, terminating a branchlet which bears four or five leaves, spreading ; axis pubescent. Staminate catkins i \ in. long, densely 1 Dendrologia, 56 (1827). Cf. Smith, Eng. Flora, iv. 227 (1828), who states that the wood and branches make the best hurdles, being tough, flexible, and durable. The wood was also used for the cutting-boards of shoemakers. 1 Card. Chron. xlvi. 19 (1909). 1748 The Trees of Great Britain and Ireland crowded with flowers ; scales pubescent at the base ; stamens usually five, rarely four to ten, unequal in length, with long hairs on the base of the filaments ; and with two glands, the one beneath the scale often three-lobed, the other quadrate or broadly crescentic. Pistillate catkins, i £- in. long ; scales about half as long as the ovary, fringed with long hairs ; ovary stalked, glabrous, about \ in. long, narrowly conic- subulate, with a short style, dividing into two arms, each of which is bifid ; glands1 two, one minute beneath the scale, the other quadrate and half the length of the pedicel. Capsules, J inch long, on distinct pedicels, glabrous, narrow and elongated. This species varies in the size and shape of the leaf, broad-leaved and narrow-leaved forms being distinguished by Andersson as vars. latifolia and angustifolia ; but intermediate forms are common. Buchanan White states that in Britain, S. pentandra is a bushy shrub in the wild state, but that when cultivated, it becomes a tree with broader and larger leaves than those of the wild plant. The bay willow is a very distinct-looking species, on account of its broad glabrous shining leaves, which resemble those of a Prunus ; but is readily recog nised to be a willow by its buds with a single scale. The flowers are fragrant, with an odour similar to that of the bay or true laurel (Laurus nobilis) ; and the leaves exhale the same fragrance, especially when bruised. The following hybrids have 5". pentandra as one of the parents :— 1. Salix Meyeriana, Rostkovius, ex Willkomm, Berlin Baumz. 427 (1811). Salix cuspidata, Schultz, Prod. Fl. Starg. Suppl. 47 (1819) ; Borrer, in Smith and Sowerby, Eng. Bot. Suppl. v. tt. 2961-2962 (1863); Buchanan White, in Journ. Linn. Soc. (Bot.) xxvii. 360 (1890). Salix titictoria, Smith, in Rees, Cyd. xxxi. No. 13 (1815). Leaves similar to those of 5". pentandra, but narrower and more cuspidate at the apex. Pistillate catkins more slender and more tapering, and bearing narrower and more cylindrical capsules with longer pedicels. Stamens usually four. This hybrid, of which the parents are supposed to be S. pentandra and S. fragilis, is said to be somewhat common on the Continent ; but in England is rare in the wild state. It is recorded as a pistillate plant in Shropshire by Buchanan White. There are also specimens in the Kew Herbarium collected in 1895 on Wybunbury Bog in Cheshire by Linton. Dr. Moss adds to this distribution Herefordshire and Westmoreland in England, and Co. Mayo and Co. Kildare in Ireland ; and tells me that it is occasionally planted in osier beds in Cambridgeshire and Suffolk. There is a fine tree in the Glasnevin Botanic Garden, about 40 feet high by 4 ft. in girth in 1912. Sir F. Moore states that it was obtained from Smith of Worcester, as the Purple King Willow, a name given to it on account of its purplish shoots. He adds that it is a quick grower, flowering early in spring, with beautiful large staminate catkins, the twigs being useful for house decoration. 2. Salix hexandra, Ehrhart, Beit. vii. 138 (1791); Buchanan White in Journ. Linn. Soc. (Bot.} xxvii. 361 (1890). Salix Ehrhartiana, Smith, in Rees, Cyd. xxxi. No. 10 (1815). Leaves lanceolate, long acuminate at the apex, silky pubescent at first, but 1 Fraser, Proc. Roy. Hort. Soc. xxxv. p. cxv (1909), showed specimens of flowers of S. pentandra, with proliferation of the posterior gland, giving rise to two or three additional pistils. Salix 1749 soon becoming glabrous ; dark green and very shining on the upper surface ; minutely glandular-serrate ; petiole slender, nearly ^ in. long. Stamens four to six. This hybrid, the parents of which are supposed to be S. pentandra and 5". alba, is known on the Continent in the staminate form. Buchanan White refers to it two specimens : one, a shrub with pistillate catkins, growing near Duddingston, Edinburgh, and the other, a barren specimen from a bush at Restennet, near Forfar. It has recently been found1 in the Lake District in Cumberland and Westmoreland. 5". pentandra is a native of nearly the whole of Europe—except the extreme south—the Caucasus, and northern Asia, as far east as Kamtschatka and Amurland, and apparently extending to the province of Yunnan, in China. It grows mainly on river banks, and in marshy places, ascending in peat mosses in the Alps to 4000 ft. elevation. It grows in similar situations in Britain2 from Argyle and Moray southwards to Derbyshire, ascending in Northumberland3 to 1300 ft. In Ireland,4 it is frequent and native in the north, becoming less common southwards, till in Kerry and Cork it appears only as an introduction. (A. H.) The bay willow makes a handsome tree, with very distinct foliage, but is rather slow in growth. The finest specimens we have seen are :—one at Kew, on the lawn near the Palm House, which measured, in 1907, 58 ft. by 9 ft. 8 in. at eighteen inches above the ground, its trunk forking at three feet. Another at Woburn was 50 ft. by 5 ft. in 1908, with a short bole. A handsome tree at Beauport, Sussex, measured in 1911 about 35 ft. by 6 ft., with a bole 6 ft. long, another near the keeper's lodge not being quite so large. (H. J. E.) SALIX BABYLONICA, WEEPING WILLOW Salix babylonica, Linnaeus, Sp. PL 1017 (1753); Louden, Arb. et Frut. Brit. iii. 1507, and iv. 2588 (1838); Andersson, Motiog. Salic. 50 (1863), and in De Candolle, Prod. xvi. 2, p. 212 (1868); Willkomm, Forstliche Flora, 471 (1887); Mathieu, Flore Forestière, 453 (1897); Burkill, in Journ. Linn. Soc. (Bot.) xxvi. 526 (1899); Camus, Monog. des Saules, 65 (1904). Salixpendula, Moench, Meth. 336 (1794). Salixpropendens, Seringe, Saules de la Suisse, 73 (1815). Salix Napoleonis, Schultz, Arch. Fl. 239 (1856). A tree, attaining 30 to 40 ft. in height, with rough ridged bark, and usually with a short trunk, and wide-spreading branches, the ultimate branchlets being very long and pendulous. Branchlets slender, glabrous except near the nodes. Leaves linear-lanceolate, about 2^ to 3^ in. long, and 5 to f in. broad, tapering at the apex into a long slender acuminate thread-like point, cuneate at the base, slightly pubescent when young, perfectly glabrous when fully grown, bright green and shining above, pale and covered with a glaucous bloom beneath ; margin finely serrate, the serra tions often ending in minute sharp incurved points ; petiole \ in. long, without glands, glabrous or slightly pubescent ; stipules early deciduous. 1 Journ. Bot. xxviii. 229 (1900). • 2 Hooker, Studenfs Flora, 355 (1878). 3 Baker and Täte, Fl. Northumberland and Durham, 248 (1868), state: "Frequent in damp woods and by stream sides, ascending in Coquctdale to Ilarbottle, in Allendale to 450 yards, and in Teesdale to the junction of the Whey Sike with Harwood Beck." 4 Praeger, in Proc, Roy. Irish Acad. vii. 282 (1901). VII O 175° The Trees of Great Britain and Ireland Only female trees are known in cultivation. Catkins terminating branchlets with one, two, or three leaves, very slender, green, curved, about i in. long ; axis pubescent ; scale about frds the length of the ovary, pilose at the base, ciliate, ovate-acuminate ; ovary sub-sessile, about j^ in. long, ovate, glabrous, ending in a very short style, which is divided into two stigmatic arms, each of which is bilobed ; gland posterior, quadrate, emarginate or bilobed. VARIETY AND HYBRIDS I. The following variety is known in cultivation :— i. Var. annularis, Ascherson, Fl, Brandenburg, 630 (1864). Var. crispa, Loudon, Arb. et Frut. Brit. iii. 1514 (1838). Salix anniilaris, Forbes, Sal. Woburn. 41, t. 21 (1829). Leaves folded and rolled up, so as to form a ring or spiral, otherwise as in the type. This remarkable variety, which is called the ring-leaved willow, is of unknown origin ;x but undoubtedly is a sport of S. babylonica, and like it is a female tree. W. Masters of Canterbury statedz that his father gave fifteen shillings for a plant 6 in. high, but had no clue as to where or how the variety had originated. He mentions an instance of a ring-leaved willow, which after being planted twenty years, produced a single branch with leaves of the ordinary form, which continued to be borne for years afterwards on the same branch and its ramifications. There is a fine specimen on the lawn near the Palm House at Kew, which was, in 1912, 56 ft. high with a trunk n ft. in girth at three and a half feet from the ground, above which it divides into two stems. Lord Kesteven measured in 1906 a fine specimen, 57 ft. by 7 ft., on a farm near Caythorpe, Grantham. II. The following trees, often considered to be varieties of S. babylonica, are probably hybrids :— 2. Salix Salamonii, Carrière, in Rev. Hort. xl. 463 (1869) and xlix. 444 (1877). Salix babylonica Salamonii, Simon-Louis,Cat. 1869, p. 85 ; Carrière, va Rev. .flo/Axliv. 115 (1872). A tree with a tall straight stem, and ascending branches, forming when young a pyramidal crown ; ultimate branchlets pendulous, but not so long as in S. baby lonica. Young branchlets pubescent near the nodes, becoming glabrous. Leaves similar in shape, size, and colour to those of S. babylonica, but pubescent with scattered appressed long hairs on both surfaces. Only pistillate trees are known ; catkins similar to those of S, babylonica, but with the axis more pubescent and the scales furnished with long cilia. This remarkable tree, which is supposed to be a crosss between S. babylonica and S. alba, is very distinct in habit, forming when young a handsome pyramidal tree, which grows with astonishing vigour and is not injured in our climate by frost 1 Dode, in Bull. Sue. Dend. France, 1909, p. 153, believes that this may have originated from a witches' broom ; and states that he obtained a similar sport as a cutting, which was taken from an abnormal growth on a Salix alba. * Card. Chron. 1855, p. 726. Cf. Darwin, Animals and Plants under Domestication, i. 408 (1890). 3 Schneider, Laubhohkumle, i. 36, note (1904), supposes it to be identical with S. sepulcralis, Simonkai, in Termes. Füz. xii. 157 (1890), said to have been found in Hungary. Camus, Monog. Saules, 235 (1904), identifies with the last-named, 5. alba, var. triais, Trautvetter, Fl. Alt. iv. 255 (1833), described from an Altai specimen. Salix in spring, as is almost invariably the case with the common weeping willow. Though not so pendulous in habit as the latter, it has graceful drooping branchlets. This tree originated on the estate of Baron de Salamon, near Manosque (Basses Alpes) some time before 1869, when it was put on the market by Simon-Louis of Metz. Carrière recommended this tree, on account of its vigour,1 for the production of timber ; and stated that it grew on all soils, even on dry soils and on limestone, where the weeping willow refused to grow, or remained stunted and yellow. S. Salamonii comes nearly as early into leaf and retains its foliage almost as late in the season as S. babylonica. There are several old trees of S. Salamonii on the borders of the lake at Kew. The exact age of these is unknown ; but they much surpass in size the true weeping willows beside them.2 At Casewick, where Lord Kesteven has planted S. Salamonii as a park tree, it thrives well, and has attained 35 ft. in height at eighteen years old. 3. Salixpendulina, Wender, in Schrift. Natf. Ges. Marburg, ii. 235 (1831). Salix blanda, Andersson, Monog. Salic. 50 (1863), and in De Candolle, Prod. xvi. 2, p. 212 (1868); Camus, Monog. Saules, 232 (1904). Salix elegantissima, Koch, in Wochschrf. Ver. Bef. Gartb. xiv. 380 (1871), and Dendrologie, ii. pt. i. 505 (1872). Under these names are possibly included three slightly different hybrids of the same parentage, S. babylonica and S. fragilis. They are wide-spreading trees, with long and pendulous branchlets, differing only slightly in habit from S. babylonica, but much more hardy than this species, and on that account often cultivated in Germany where the true weeping willow is killed by severe frost. I have not been able to study the three forms ; but a tree at Kew, labelled S. elegantissima? obtained from Dieck in 1889, has leaves more coriaceous and slightly broader than those of S. babylonica ; both surfaces glabrous, slightly glaucous beneath. Catkins occasionally androgynous,4 usually only pistillate, shorter than in S. babylonica ; axis pubescent ; ovary shortly pedicellate, about T\, in. long, slightly pubescent at the base, with two stylar arms, each arm bilobed ; scale two-thirds the length of the ovary, very pilose ; glands irregular, usually two,—one narrowly oblong between the scale and the ovary,—the other, posterior, nearly quadrate, and occasionally bilobed. Koch states that this tree, which on the Continent is often sold in commerce as S. Sieboldii, is fast in growth, with very long branchlets, almost reaching to the ground. There is a thriving specimen at Glasnevin, about 25 ft. high, which is labelled S. blanda. In the Botanic Garden at Leyden there is a handsome tree, about 40 ft. by 3 ft. in 1912, which is labelled S. Petzoldii pendula? 1 In Garden and Forest, x. 497 (1897), it is said to be the fastest-growing willow in California, where, at the Chico Forestry Station, stems cut back in February 1896 to 2 ft. from the ground, were 31 to 32 ft. high in August 1897. 2 A vigorous young tree on the bank of the Cam, Trinity College, Cambridge, was 35 ft. high by 4 feet in girth in 1912. Its pyramidal crown with ascending upper branches contrasts much with the older but lower adjacent weeping willows, which have broad flattened crowns and spreading branches. 3 Both .S1. blanda and .S1. elegantissima were described as having perfectly glabrous ovaries ; but notwithstanding this, the Kew tree, received from Dieck, is probably .S1. elegantissima. Ascherson and Graebner, Syn. Mitteleurop. Flora, iv. 73 (1908), correctly describe 5. elegantissima as having a pubescent ovary ; and on that account doubt its being a hybrid between 5. babylmica and 5. fragilis. Following Koch, tbey believe that it was introduced from Japan by Siebold ; and if this is the case, 5. fragilis could scarcely have been one of the parents. 4 As Camus points out, androgynous catkins are common in the hybrids .S1. blanda and .S1. sepulcralis which he describes. 6 Cf. Lauche, Haupt-Katalog Muskaiier Baumschulen, 1905, p. 32. 1752* The Trees of Great Britain and Ireland DISTRIBUTION S. babylonica was the name given by Linnaeus1 to the common weeping willow cultivated in Europe, which he erroneously supposed to have been identical with the trees growing by the rivers of Babylon, which are mentioned in Psalm cxxxvii. 2. The latter are, without doubt,2 a species of poplar, Populus euphratica. The original home of the weeping willow appears to be central and southern China, where it is commonly found on the banks of rivers and canals, as well as in gardens. Fortune3 observed this tree, identical in all respects with that cultivated in Europe, in the neighbourhood of Shanghai and Canton, and also near Ningpo, where it is sometimes planted around graves. The typical form of the species, that long cultivated in Europe and prevalent in central and southern China, is always a female tree, with long pendulous branchlets. In the neighbourhood of Peking4 a variety is more common, which may be dis tinguished as var. pekinensis, Henry. This is an upright tree, with ascending branches, which is known in both sexes. The foliage is practically identical6 with that of the typical form ; but there is a marked difference, not only in habit, but in the pistillate catkins, which are extremely short and compact in this variety, not exceeding ^ in. in length, often on leafless peduncles ; ovaries wider, ending in two short undivided stigmas ; scales nearly glabrous. Var. pekinensis is represented at Kew by a tree, about 15 ft. high, originally from Peking, which was obtained from the Arnold Arboretum in 1905. This tree, like the common weeping willow, is injured regularly by spring frosts. From China the weeping willow was early introduced into Japan, where it is now cultivated and naturalised in many places.6 During the Middle Ages it was probably carried westward to Persia, Asia Minor, and Turkey ; and it is now also "cultivated in Baluchistan, Northern India, and the Punjab, and less commonly in the plains farther east, and also in Kurdistan,"7 The first mention of the tree in European literature is by Petiver,8 who refers to a specimen gathered in China by Cunningham in 1701, but which cannot now be found in the British Museum. The first mention of the tree in the Levant was by Tournefort9 in 1719; and it is possible that either he or Wheler, who travelled in Asia Minor and Greece in 1675-1676, introduced it into western Europe. 1 First described by Linnœus, in Hort. Cliff. 454 (1737). 2 Cf. Koch, Dendrologie, ii. pt. i. 507 (1872). S. babylonica is not now found in Babylonia. See p. 1771, note 4. 3 Wanderings in China, 118, 136 (1847), and Residence Among the Chinese, 52 (1857). The weeping willow is apparently depicted by Nieuhoff, VEmbassade des Provinces Unies vers China, \. 189 (1665), in a view of Tonnau, a village on the river Wei in the province of Shantung. 4 Cf. Bretschneider, in Journ. N. C. Branch, R. Asiat. Soc. xv. 15, 30 (1880), and Bet. Sinic. ii. 359 (1892). The name of 6". babylonica in China is yang-liu. The weeping form is rare at Peking, where it is distinguished as ck'tii-yang-litt. 5 The serrations of the leaves in var. pekinensis are more distantly placed, and without the cilia, which are present in the typical form. These slight differences may be due to the different ages of the specimens. 8 Franchet et Savatier, Emtm. PL Jap. i. 459 (1875). 7 Brandis, Indian Trees, 637 (1906). 8 Mus. Petiv. Cent. 997 (1703), where it is referred to as follows : " Yang-din chinensitus. Arbor salicis folio ramulis pendulis. Frequently painted on Japan work. Of the wood they make arrows." 6 Corollarium, 41 (1719), where it is described as follows: " Salix orientalis, ßagellis deorsitm pulchre pendentibus, hvjns etiam meminit IVheler I tin." Wheler, Journey to Greece and Asia Minor, 217 (1682), saw near Brusa, and not far from Mt. Olympus in Asia Minor, a tree which appears from his descripiion to have been a weeping willow. Salix '753 The weeping willow was introduced into England some time before 1730, as it was on sale in London, according to a catalogue* published in that year. Collinson states2 that it was introduced by Mr. Vernon, Turkey merchant at Aleppo, who planted it at his seat in Twickenham Park, where Collinson saw it in 1748. The latter says that this tree was the original of all the weeping willows in England ; and adds that he measured one in 1765 at Mr. Snelling's at Godalming, which, though only fifteen years old, was 6 ft. in girth. There was a famous weeping willow, planted by Pope in front of his villa at Twickenham, which was felled3 in 1801, when the story was given in St. James's Chronicle, August 25-27, of that year,4 that this tree was the first one planted in England, having been introduced as a withy round a package from Spain ; but doubtless Pope's tree was a cutting from Mr. Vernon's willow. Another celebrated tree was the weeping willow in St. Helena, which was planted in 1810, and was a favourite of Napoleon. After his death cuttings were brought to England, and planted in many places, where they were called Napoleon's willow, but differed in no respect from the ordinary form.5 There is a weeping willow at the Fountain Pond, Cassiobury, which formerly bore a plate 6 stating that it had been a cutting from the tree in St. Helena. This tree fell and sustained considerable damage ; but it has been replanted, and, according to Mr. Daniel Hill, who measured it in 1912, is 36 ft. high and 4 ft. 7 in. in girth. (A. H.) The weeping willow strikes freely from cuttings, and grows rapidly in good soil beside water ; but is very liable to have the young shoots killed by frost, and is not nearly so hardy as the hybrid S. Salamonii. It is one of the earliest trees to come into leaf, and the latest in retaining its foliage, being frequently green in December. The finest trees known to Loudon were those at various places on the banks of the Thames, which were 50 to 60 ft. high in 1838 ; but it is doubtful if any of these now survive, as it is not a long-lived tree. The best that I have seen is perhaps a tree (Plate 379) on the Promenade, Cheltenham, which was planted about 1860, and is still thriving, although its limbs have been supported by iron rods for some years. It measured in 1911 about 75 ft. by 9 ft., and on 24th November 1911, after a severe frost, it still retained most of its leaves. There are several picturesque trees,7 but of no great height, growing on the banks of the Cam, behind the Colleges of Cambridge. These are exceeded in size by one 1 Miller, Catalogus Plantarum, a Catalogue of Trees, Shrubs, Plants, and Flowers, which are propagated for sale in Gardens near London, p. 71 (1730), where it is mentioned as "S. orientalis, T. Cor. 41. The weeping willow vulgo." 2 Dillwyn, Hortus Collins. 48 (1843), quoted by Loudon, Gard. Mag. xix. 64 (1843). Alton, Hort. Kew. v. 356 (1813), states that the weeping willow was first cultivated in 1692 in the Royal Garden at Hampton Court, but gives no authority for this. If true, the introduction into Western Europe was probably made by Wheler. Aiton quotes Plukenet, Phytographia, t. 173, fig. 5, which is not the weeping willow. 3 Corbett, in Mem. Twickenham, 285 (1872), states that this tree perished and fell to the ground in 1801. The wood was worked up by an eminent jeweller into all sorts of trinkets and ornaments, which had an extensive sale. The Empress of Russia took cuttings from Pope's willow in 1789 for the gardens at St. Petersburg. « Phillips, Sylva Florifera, ii. 263 (1823). » Loudon, Arb. et Frut. Brit. iv. 2588 (1838), and Trees and Shrubs, 758 (1842). Forbes, Salic. Woburn. 43 (1829) states that a plant raised from a cutting of the St. Helena tree was identical with the common weeping willow. 8 D. Hill, in Trans. Herts. Nat. Hist. Soc. xiv. pt. ii. 132 (1911). 7 These trees are comparatively young, and have replaced the original trees, which were planted in 1760. Cf. Willis and Clarke, Archil. Hist. Univ. Camb. ii. 646 (1886). 1754 The Trees of Great Britain and Ireland in the Fellows' Garden of King's College, which was about 45 ft. high and 10 ft. in girth in 1912, with the trunk decayed and mended with cement. The weeping willow attains * a much greater size and beauty in warm countries than it does in England. I have seen none finer than in Chile, where it is often planted by the sides of the irrigation canals, and enjoys a long and warm summer. (H. J. E.) SALIX FRAGILIS, CRACK WILLOW Salix fragilis, Linnaeus, S{>. PI. 1017 (1753); Smith, Fl. Brit. iii. 1051 (1804), Eng. J3at.z t. 1807 (1808), and Eng. Flora, iv. 184 (1828) ; Andersson, Monog. Salic. 41 (1863), and in De Candolle, Prod. xvi. 2, p. 209 (1868); Willkomm, Forstliche Flora, 472 (1887); Buchanan White, in Journ. Linn. Sec. (ßof.) xxvii. 368 (1890); Mathieu, Flora Forestière, 450 (1897); Camus, Monog. des Saules, 76 (1904). A tree, attaining about 70 ft. in height. Bark rough, strongly ridged, and divided into broad deep fissures. Young branchlets slightly pubescent, glabrous in the second year. Buds appressed to the branchlet, compressed, shining, glabrous. Leaves lanceolate, about 4 in. long and f in. broad, gradually tapering above into a long caudate oblique acuminate apex, cuneate at the base, silky when young ; upper surface in summer glabrous, shining ; lower surface glaucous or glaucescent, with scattered silky appressed hairs ; margin coarsely serrate, each serration tipped with a conspicuous reddish brown gland ; petiole f in. long or more, pubescent, usually with two glands at the insertion of the blade. Catkins appearing with the leaves, terminating short branchlets, which bear three or four usually entire leaves ; axis densely pubescent. Male catkins, about 2^ in. long ; scales concave, oblong, truncate or cuspidate, glabrous within, pubescent without, margin fringed with long hairs ; stamens two, filaments as long as the scale, slightly pubescent at the base, anthers yellow ; glands two, the posterior transversely oblong, the anterior half its size and narrowly oblong. Female catkins about 2 in. long ; scales concave, lanceolate, pilose at the base, ciliate, with long hairs ; ovary distinctly stalked (pedicel ^5 in. long), one-third longer than the scale, fusiform, glabrous, £ in. long, gradually narrowing to the apex, which ends in a short style, divided into two arms, each of which is bilobed ; only one gland,3 which is posterior, usually present, quadrate, entire in margin, much shorter than the pedicel. Fruiting catkins about 3 in. long, with a pubescent axis, and distinctly stalked capsules. S. fragilis is called crack willow, on account of the ease with which the branchlets disarticulate, especially in spring. It is readily distinguishable by its 1 Salix Sa/sa/, Dode, in Bull. Soc. Demi. France, 1906, p. 62, fig. (not Forskal), received from Palermo, where it was introduced by Schweinfurth from Egypt, is indistinguishable from .S1. balylonica ; but is remarkably fast in growth and may be a hybrid. A specimen in the nursery of the Jardin des Plantes at Paris, which I saw in 1912, though only three years old from a small cutting, was 2O ft. high and 15 in. in girth. M. Dode states that this is perfectly hardy at Paris, none of the shoots being damaged by frost. 2 One of the male flowers in Eng. Bat. t. 1807, has three stamens ; and it is possible that the male plant figured is a hybrid like S. speciosa. 3 The presence of two glands in the pistillate flower of S. fragilis is very rare, and abnormal. Salix nearly glabrous coarsely glandular-serrate large leaves, ending in a long point directed to one side. When adult the stem is covered with a much rougher bark than that of 5". alba, the depressions between the ridges being broad and deep ; on this account buyers of willow timber usually designate it as the "open-bark willow." It is of no value for making cricket bats, as its wood lacks the necessary strength, lightness, and elasticity. VARIETIES AND HYBRIDS True S. fragilis does not seem to be a very variable plant ; but peculiar forms, more or less resembling it, are known, which are supposed to be of hybrid origin. These are described below. The twigs in some trees become dull grey in the second year, whilst in others they assume a brilliant orange colour. Varieties founded on the size of the leaf or on the colour of its lower surface are not sufficiently distinct to be worth naming ; and the two forms, vars. britannica and genuina, Buchanan White, m Journ. Linn. Soc. (Bot.) xxvii. 368 (1890) based on the relative length of the bract and flower, are untenable, as this depends simply on the age at which individual flowers are observed. The following are supposed to be hybrids :— i. Salix decipiens, Hoffmann, Hort. Sal. ii. 9, t. 31 (1791); Smith, Eng. Bot. t- *937 (1808); Forbes, Sal. Wobtirn. t. 29 (1829); Buchanan White, in Journ. Linn. Soc. (Bot.) xxvi. 348 (1890). Salix fragilis, var. decipiens, Koch, Syn. Fl. Germ. 643 (1837). Salix cardinalis, Veitch, Cat. Trees and Shrubs, 1910, p. 75. A shrub or small tree. Young branchlets crimson or dark red, glabrous ; becoming in the second year polished and yellowish white or of a clay colour. Leaves glabrous, smaller than those of 5". fragilis, 2 to 3 in. long, \ to f in. broad, more oblong, more parallel-sided, usually less narrowed and often rounded at the base ; under surface pale dull green, scarcely glaucous ; serrations finer, sharper, and closer together than in 5". fragilis. This is supposed to be a hybrid between 5". fragilis and 5". triandra, and occurs in both sexes. The catkins of the female plant are intermediate between these two species ; while the number of stamens in the male plant is usually two, rarely three. 5". decipiens, according to Buchanan White, is widely spread, but not abundant in Britain ; and is usually found growing in company with S. triandra and 5". fragilis. He mentions various localities from Perth to Somerset. There are specimens in the Kew Herbarium from Dorset, Sussex, and Essex. This species was said by Forbes1 to be cultivated for basket-work, producing when cut crimson-coloured annual shoots, which are very remarkable in appearance. It appears2 to be largely used at the 1 Sal. Wolurn. 57, t. 29 (1829). Cf. also Loudon, Arl. et Frut. Brit. iii. 1515 (1838). Smith, Eng. Bot. t. 1937 (1808), states that Crowe found it in several osier beds in Norfolk and Cambridgeshire, where it was known as white Welsh osier. 2 Cf. Ellmore and Okey, injettra. Board Agric. xviii. 915 (1912), who refer to it as Salix alba, var. cardinalis. 1756 The Trees of Great Britain and Ireland present time for the same purpose in Belgium ; and when imported is called " Belgian red willow." It is occasionally planted for ornament, appearing in some nursery catalogues under the name S. cardinalis. 2. Salixspeciosa, Host, Sal. 5, t. 17 (1828) ; Buchanan White, mjourn. Linn. Soc. (Bol.) xxvii. 353 (1890). Salix fragilis, var. latifolia? Andersson, in De Candolle, Prod. xvi. 2, p. 209 (1868). A large tree, similar in bark and habit to S. fragilis. Young branchlets slightly pubescent, glabrous in the second year. Leaves longer and broader than those of S. fragilis, up to 6 in. long and if in. wide ; broadly lanceolate, with a long acuminate often curved apex ; pubescent when young ; when mature, upper surface dark green, shining, glabrous ; lower surface pale, glabrescent, or with scattered appressed hairs ; margin coarsely glandular-serrate ; petiole f in. long, often with one to four glands at the junction with the blade. Staminate catkins, 2^ in. long ; flowers crowded on the densely pubescent axis ; scale, shorter than the stamens, and covered with long white silky hairs ; stamens usually two, rarely three, glabrous or pilose at the base ; glands two, variable, entire or lobed. Pistillate flowers not seen. The above description is drawn up from an old tree1 on the side of the lake at Kew, which has been named S. triandra x fragilis? by Linton. It is identical with some trees, which have been called " open-bark willow," by Carter, and is perhaps not very rare in Britain. 3. Salix viridis, Fries, Nov. PI. Suec. 283 (1828) ; Andersson, in De Candolle, Prod. xvi. 2, p. 210 (1868); Buchanan White, in Joîirn. Linn. Soc. (Sot.) xxvii. 364, 371 (1890). (?) Salix rubens, Schrank, Baier Fl. i. 226 (1789) Salix excelsior, Host, Sal. 8, tt. 28, 29 (1828). Salix palustrts, Host, Sal. 7, tt. 24, 25 (1828). Salix montana, Forbes, Sal. Woburn. 37, t. 19 (1829). Salix fragilis-alba, Wimmer, Denkschr. Schles. Ges. 156 (1853). A tree similar in size to the reported parents, S. aida and S. fragilis, with spreading branches, and usually more or less pendulous branchlets. Young branchlets variable, glabrous or pubescent. Leaves intermediate between these two species, variable in size, shape, colour, and pubescence ; darker green, less shining, more finely serrated, and less oblique at the apex than those of S. fragilis; longer and broader than those of S. alba, and soon becoming glabrescent. Staminate catkins narrower and larger than in S. fragilis. Pistillate catkins more slender than in that species ; ovary shortly but distinctly pedicellate ; capsules intermediate in size between S. alba and S. fragilis. 1 Specimens which were sent from Kew to Buchanan White in 1887 ancl 1888, and are now preserved in his herbarium at Perth, show that this tree was then labelled .?. fragilis, var. latifolia, a name accepted by White. 2 .£. alofecuroides, Tausch, Ind. fforf. Canal. (1821), ex Andersson, in De Candolle, Prod. xvi. 2, p. 203 (1868); A. Kerner, in Verh. Z. B. Ges. Wien, 69 (1860); Koch, Dendrologie, ii. pt. i. 516 (1872), is the name given to the hybrid A triandra x fragilis, which occurs on the continent. This, judging from Tausch's specimen, preserved in the Cambridge Herbarium, is quite distinct from X sfeciosa, Host. Salix Ï757 Numerous forms of S. viridisl occur, some being close to S. fragilis, others nearer to S. alba ; and no exact definition of this hybrid is possible ; but it is easy to recognise when S. alba, S. fragilis, and S. cœrulea are excluded. According to Buchanan White, S. viridis is widely distributed in Britain, occurring from Cornwall and Surrey to Perth ; but it is less abundant and more local than the parents, S. alba and S. fragilis. The second-class bat willow is often S. viridis. S. viridis doubtless grows to as large a size as either of the parents, with which it is generally confused. The largest2 we know are growing beside a stream at Thornbury, Gloucestershire. One tree is 60 ft. high by 18 ft. in girth. Near it are two trees growing from the same root, one of which is 65 ft. by 11 ft. 6 in. and the other 60 ft. by 15 ft. 7 in., according to measurements which were kindly taken for us by Mr. Samuel Fudge in July 1912. 4. Salix Russelliana, Smith, Fl. Brit. iii. 1045 (1804), Eng. Bot. t. 1808 (1808), and Eng. Flora iv. 186 (1823); Forbes, Sal. Woburn. 55, t. 28 (1829); Loudon, Are. et Frut. Brit. iii. 1517, fig. 1311 (1838). Salix fragilis, var. Russelliana, Koch, Syn. Fl. Germ. 643 (1837). A tree, remarkably vigorous in growth, with long, straight and slender branches, not angular in their insertion like those of S. fragilis. Leaves in the young state white and silky pubescent beneath ; when fully grown, lanceolate, acuminate, 3\ in. long, f in. wide, finely serrate, with scattered appressed hairs on the under surface near the midrib. Pistillate trees only known, with flowers similar to those of S. fragilis, having pedicellate ovaries and bipartite styles, but with looser catkins. This seems, judging from the type specimens in Smith's herbarium at the Linnean Society, to be a form of S. viridis, which is not clearly known to botanists3 at the present time. The normal leaves are well figured by Loudon (fig. 1311); those depicted by Sowerby (Eng. Bot. t. 1808) from coppice shoots are larger than the normal leaves, and have the glands on the petiole often developed into slender leaflets. 5. Russelliana was sent to Woburn about 1800 by Mr. Bakewell from Leicester shire, and was called the Dishley or Leicester Willow. It seems to have been extraordinarily vigorous, as Lowe, Agric. Survey of Notts, p. 118, is quoted4 for the fact that " this willow in a plantation yielded at eight years' growth poles which realised a net profit of ^214 per acre." It was also remarkable for the large per- 1 Scaling mentions in his pamphlet, Salix or Willow, i. Cat. p. 8 (1871) and ii. 19 (1872) two willows, which he sent out in 1871 from his nursery at Basford, Notts :—(l) .£ sanguinea, the branches of which were brilliant red in winter. This was obtained by Scaling some years previously in the Ardennes, where it was known as the red willow. (2) .S. basfordiana, with branches of a deep orange colour. This appeared as a seedling in the nursery, and grew with great vigour. Salter, in Card. Chron. xvii. 298, figs. 41, 42 (1882), apparently confused these two distinct trees, which he described as S. basfordiana. There are iwo trees at Kew under this name—a male, No. 58, which is a form of S. viridis, and a female, No. 80, which is identical with .?. vitellina. Whether these correspond to Scaling's .S. sanguinea and .5. basfordiana is uncertain. 2 These trees have pubescent leaves, simulating those of .£. alba, but larger and thinner. They also differ from the latter species in having distinctly pedicellate fruit. 3 .S. Russelliana has been misunderstood by most botanists since Smith's time. White's view that it was true S. fragilis, while Smith's S. fragilis was S. viridis, is untenable. 4 Duke of Bedford, in Forbes, Sal. Wobutn., p. v (1829). The frontispiece represents Johnson's Willow. VII P 1758 The Trees of Great Britain and Ireland centage of tannin contained in the bark, which exceeded, according to Mr. Biggin's analysis,1 even that of the oak. The Rev. W. Dickenson assured the Duke of Bedford and Sir James Smith that the tree known as Johnson's Willow,2 which grew between Lichfield and Stow Hall, was 5. Russelliana, This tree, which was so called because Dr. Johnson frequently rested under its shade, was figured by Loudon, Arb. et Friit. Brit. iii. figs. 1312, 1313, who states that it was about 60 ft. high, with a great bole, about 20 ft. in length, which girthed,8 in 1810, 21 ft. at six feet from the ground. It contained 130 cubic ft. of timber, and was perfectly sound. It was blown down in 1829, when it was supposed to have been 130 years old. Mr. L. Fosbrooke, of Ravenstone Hall, Ashby de la Zouch, believes that 5. R^l,sselliana still occurs in Leicestershire in the Trent valley, where the trees are supposed to be female 5. fragilis. They differ in foliage from male trees of true 5. fragilis, according to Mr. Fosbrooke, who tells me that though they grow fast when young, they become round-topped as they approach maturity, and are certainly not nearly as vigorous as 5. cœrulea. I have not been able to examine specimens of these trees ; and cannot say whether they are identical with the original 5. Russelliana, which may have been a solitary sport or hybrid of exceptional vigour, now lost to cultivation. DISTRIBUTION The distribution of 5. fragilis in the wild state cannot be determined with certainty, as it has been largely planted outside of its original area ; but it is supposed to be indigenous in the greater part of Europe and in western Asia. It is not wild in Norway, but occurs there as a planted tree as far north as lat. 64° 5'. In Sweden it is also only known in cultivation, the male tree being seen occasionally in Werm- land and Schone, and the female tree in the district round Kalmar. It is wild in Jutland, in the island of Oesel, and in Russia,4 where it extends as far north as Esthonia, Livland, Kostroma, and Kazan ; but is absent from the Crimea. It is widely spread in the Caucasus, Persia, and Asia Minor. Farther east it is cultivated5 in the Kuram valley, in Gilgit, Ladak, and Lahaul, as well as at Quetta, where it was found by Lace at 5600 ft. elevation. In Europe it extends southwards to Spain and Portugal, Sicily, and Greece. It is planted largely in northern and central Germany, where it is wild in many localities ; but in southern Germany, Austria, and Hungary, it is only seen in river valleys, where it prefers a deep loamy soil ; and ascends along the edges of streams in the mountains, as high as 1700 ft. altitude in the Bavarian Alps, and 2500 ft. in Tran sylvania. In order to succeed it requires considerably more moisture in the soil than 5. alba, and on this account is most often seen on the Continent generally, on the banks of rivers, streams, and lakes, being rare in the interior of the forests. 1 Davy, Agric. Chem. 89 (1814) analysed the bark of a large tree of the "Leicester Willow" and found it to contain more tannin than any other British tree, having a little more than coppice oak, and three times as much as the " common willow." 2 See note 4, p. 1757. 3 Withering, Arr. Brit. Plants, ii. 68 (1818) gives these measurements on the authority of Rev. W. Dickenson. Loudon, Derby Arboretum, 55 (1840), states that a young tree, raised from the branches of Johnson's Willow, was growing at Lich field in 1836, when it was 20 ft. high. 4 Koppen, Hokgew. Europ. Russlands, ii. 257 (1889). 6 Brandis, Indian Trees, 637 (1906). Salix The crack willow is probably wild in marshy ground in Britain from Perthshire southwards ; but it is extremely difficult to decide in what stations it is really indigenous. It is supposed to have been introduced1 into Ireland, where however it is often seen as a planted tree. (A. H.) This species occasionally attains as great dimensions as 5. alba. In Gard. Chron. i. 447 (1874), E. Lees figures some remarkable old willows. The largest of these (S. fragilis) grew in the Wye valley near Ross, and was over 70 ft. in height, with a girth of 24 ft. at two feet from the ground. Mr. H. Marshall can find no trace of this tree at the present time. Jackson, Syon House Trees and Shmbs, 29 (1910), mentions an immense tree on the south side of the lake in a decaying condition. "Judging from the length of the bole, which is now prostrate, this may well have been the specimen mentioned by Loudon2 (p. 1521) as being 89 ft. high and about 13 ft. in girth, and called by him Salix Russelliana." In Messrs. Samson's nursery at Kilmarnock there was a crack willow, which in 1904 measured, according to Mr. Renwick, 80 ft. high, with a bole 22 ft. in length and girthing 16 ft. i in. at five feet from the ground. This tree was blown down in November 1911, when the trunk was found to be much decayed at the heart. No shoots have sprung from the root since. Mr. Renwick8 records a tree at Bruntwood Mains near Galston, Ayrshire, which measured 13 ft. i in. in girth in 1902, when it was reputed to have been 62 years old. (H. J. E.) SALIX ALBA, WHITE WILLOW Salix alba, Linnaeus, Sp. PI. 1021 (1753); Smith, Eng. Bot. t. 2430 (1812), Eng. Flora, iv. 231 (1828); Forbes, Sal. Woburn. 271, t. 136 (1829); Loudon, Arb. etFrut. Brit. iii. 1522 (1838); Andersson, Monog. Salic. 47 (1863), and in De Candolle, Prod. xvi. 2, p. 211 (1868) ; Willkomm, Forstlich Flora, 469 (1887); Buchanan White, in Journ. Linn. Soc. (Bot.) xxvii. 370 (1890); Mathieu, Flore Forestière, 451 (1897); Camus, Monog. des Saules, 69 (1904). A tree, attaining about 90 ft. in height and 20 ft. in girth, with more or less ascending branches, but with spreading or pendulous ultimate branchlets. Bark less deeply fissured than in 5. fragilis, the depressions between the ridges being narrower and shallower than in that species. Young branchlets covered with whitish appressed pubescence, partly retained in winter and in the following year. Buds flattened, appressed against the twig, silky pubescent. Leaves lanceolate, 2 to 2^ in. long, rarely exceeding ^ in. in width, tapering to a long acuminate, straight or curved apex ; upper surface greyish green, more or less covered with silky white appressed pubescence ; lower surface whitish, densely covered with similar pubescence ; margin densely ciliate, and with minute glandular serrations ; petiole short, pubescent. Catkins, appearing with the leaves, on short lateral leafy branchlets ; axis densely tomentose ; flowers crowded. Staminate catkins, about i^ in. long; 1 Traeger, in Proc. Key. Irish Acad. vii. 283 (1901). 2 Loudon's identifications of these large willows are very uncertain. 3 Trans. Nat. Hist. Soc. Glasgaw, vi. pt. iii. 353 (1902). 1760 The Trees of Great Britain and Ireland stamens two, united and pubescent at the base ; glands two, anterior quadrate and entire, posterior small and ligulate ; scale concave, ciliate, half the length of the stamens. Pistillate catkins, i^ in. long; ovary sessile, glabrous, about -^, in. long, ovoid, abruptly narrowed towards the apex ; style short, the two arms apparently undivided (really slightly bilobed) ; scale, fringed with long hairs, and as long as the ovary ; gland, one, posterior, quadrate. Fruiting capsules sessile, ^ in. long. S. alba appears to be very variable in the size and pubescence of the leaves ; but doubtless some of the varieties attributed to S. alba are forms of the hybrid, S. viridis, in which the characters of S. alba are dominant. The following form is in cultivation :— i. Var. argentea, Wimmer, Sal. Europ. 17 (1866). Var. splendens, Andersson, in De Candolle, Prod. xvi. 2, p. 211 (1868). Var. leucophylla, Simon-Louis, Cat. 1869, p. 85. Var. regalis, Beissner, in Gartenßora, xxvi. 40 (1877). Salix splendens, Bray, ex Opiz, Boehm. Gew. no (1823). Sau'x regalis, Wesmael, in Bull. Cong. Bot. Brux. 1864, p. 280. Young branchlets and leaves on both sides covered with dense silvery white tomentum. This is highly effective in beds in gardens when cultivated in masses, the single stems being kept about 12 ft. high by pruning. At Glasnevin there is a thriving tree about 35 ft. high. DISTRIBUTION S. alba is widely spread through central and southern Europe, extending southwards to Algeria and Morocco, and eastwards to western and northern Asia. It is impossible to define its northern limit in Europe, as its natural area has been much extended by planting ; but, according to Schübeler, it is not wild in Scandinavia, where it is occasionally planted as far north as lat. 63° 52' in western Norway. In Russia,1 it is probably wild as far north as southern Livonia, Vitebsk, Smolensk, Moscow, Vladimir, Kostroma, Viatka, and Perm ; and is common in the Crimea, where it attains a large size. It is widely spread in the Caucasus, north Persia, and Asia Minor ; and occurs in Siberia as far east as Lake Baikal ; but it is only cultivated2 in the northern Himalayas and in western Tibet. It is doubtfully wild in northern Germany, where it is, however, planted to a considerable extent. Farther south, it is undoubtedly indigenous in southern Germany, Austria, the Balkan peninsula, Italy, Spain, and the greater part of France.3 In central and southern Europe, it is often the dominant tree in the forests of the alluvial plains and in the woods on the banks of rivers, as on the Danube and its great tributaries, where it either forms pure stands or grows in mixture with Quercus pedunciilata, Populus nigra, and Salix fragilis. It grows in similar situations to the last species, but thrives in lighter and more sandy soils. It ascends the river valleys in the mountains to 1400 ft. in the Bavarian Forest, to 2700 ft. in the Bavarian Alps, to 2900 ft. in Transylvania, to 3200 ft. in the Caucasus, and to 5400 ft. in the Sierra Nevada. The comparatively low altitude 1 Koppen, Hokgew. Europ. Russlands, ii. 259 (1889). * J. D. Hooker, Flora Brit. India, v. 629 (1888). 3 It is common on the banks of all the large rivers of western France from the Seine to the Adour ; and appears to be the sole native large willow in these situations where I saw no S. fragilis in 1912. Salix 1761 which it attains in the mountains is not due to the fall in temperature, as it ripens seed and grows to be a fine tree in Livonia. The white willow is supposed to be indigenous in Britain in marshy ground from Sutherland southwards ; and is considered by Praeger * to be probably in Ireland an original tree of river banks, but now generally planted. Witches' brooms 2 on this species are apparently formed by the irritation set up by a mite (Eriophyes salicis), which causes a shoot to branch repeatedly and produce small narrow soft leaves, the whole mass often measuring a foot in diameter, and turning bright red in autumn. The white willow was early introduced into North America, where it is now planted both in Canada and the United States. Together with S. fragilis, it has proved very useful as a windbreak in the prairie regions, and where timber is scarce is valuable for fuel. It is also a useful tree for reclaiming and holding the soil along streams. The wood is fairly durable in contact with the soil, and has been employed for fence-posts in the north-western plains of the United States. Pinchot3 recommends that it should be grown as coppice, when required for posts or fuel ; and says that plantations should be tilled frequently till they are well shaded. This cultivation destroys weeds, and prevents excessive evaporation of moisture from the soil. (A. H.) REMARKABLE TREES A tree at Bury St. Edmunds, figured by Strutt, Sylva Britannica, plate xxiii, as the Abbot's Willow, is one of the largest white willows of which we have record. It was measured by a surveyor, named Lenny, in 1822, when it was 72 ft. high by i STJ feet in girth, and was estimated to contain 440 feet of timber. Loudon was informed that it was almost dead in 1836. The largest white willow known to us is at Haverholme, near Sleaford, in the park east of the priory, and measured4 in 1907 about 80 ft. high by 25^ ft. in girth, with a spread of branches about 80 ft. in diameter. A photograph, for which I am indebted to Miss F. H. Woolward (Plate 380) shows it to be past its prime ; but though partly decayed, it is said to be still increasing in girth. At Water Hall farm, Bayfordbury, Herts, there is a tree, about seventy years old, which Mr. H. Clinton-Baker measured in 1912. It is about 65 ft. high ; and has a short bole, 27 ft. in girth at one foot from the ground, and divided at four and a half feet up into two stems, the larger of which is 1 7 ft. in girth at six feet from the ground. At Compton Wynyates, Kineton, a beautiful old mansion belonging to the Marquess of Northampton, there is a group of sound healthy trees from 80 to 90 ft. high, one of which girthed 10 ft. 4 in. in 1905. At Highclere, there is a fine tree, a picture of which by Alfred Parsons, now in the possession of Lady Carnarvon, was reproduced by Robinson, Wild Garden, 258 (1895). It formerly had three stems, but the two largest were blown down some 2 proc Roy_ IIort_ A._ xxxy; pt_ ;; ^ cxvi; (IQIO). 1 Proc. Key. Irish Acad. vii. 283 (1901). 3 U. S. Dept. Agric. Forest Circ. No. 87 (1907). * Measured by Mr. ]. Cowance. In Woods and Forests, 1884, p. 642, this tree was said to have been, in 1881, 27 ft. 4 in. in girth at one foot from the ground, 20 ft. 5 in. at four feet, and 28 ft. at seven feet. The branches spread 40 ft. on one side, and 28 ft. on the other. Cf. Card. Chron. xiv. 362 (1893). 1 762 The Trees of Great Britain and Ireland years ago. Mr. Storie informs us that the stem remaining, which leans to one side, is over 50 ft. in height and girths 1 3 ft. at five feet from the ground. At Fawsley Park, Daventry, a tree, of which a photograph was reproduced in the Journal of the Northampton Natural History Society for 1882, was said to have been no less than 102 ft. high and 9 ft. 2 in. in girth. This tree was blown down in February 1882 ; and apparently all that now survives Js a small tree, grown from a cutting, specimens of which were kindly sent us by Lady Knightley of Fawsley. In Scotland, Henry measured in 1904 two trees at Palnure, Kirkcudbright, which were 86 ft. by 10 ft. 8 in., and 82 ft. by 12 ft. 9 in. At Coodham House, Kilmarnock, there are two great trees, the largest of which was 17 ft. i in. in girth in 1904, when it was measured by Mr. Renwick. Mr. J. M 'Gran, the gardener, informs us that in 1910 it was about 60 ft. high, and girthed 17 ft. 8 in. at three feet from the ground, 19 ft. 8 in. at five feet, and 21 ft. at six feet. At Moncreiffe I saw in 1907 a remarkable old willow, of which the original trunk, now broken, measured 19 ft. 4 in. in girth. The branches had become rooted in several places ; and one of these, now severed from the trunk, is over 6 ft. in girth. The total circumference of the branches is 106 paces. The foliage of this tree, of which we have not seen flowers, is not so white as usual ; and it may be a form of S. viridis. In Ireland we have no records of any very large trees ; but in Mucksna Wood, a mile from Kenmare, I saw by the road-side in 1910 some very fine willows, which the Marquess of Lansdowne believes to have been planted eighty or ninety years ago. I measured one of these 85 ft. by 9 ft. 10 in. The white willow attains a large size in Germany. T. Schube * gives reproduc tions from photographs of two enormous trees which are growing in Silesia, one at the Primkenauer factory, 20 ft. in girth, and the other, with a taller stem not quite so thick on the road between Stronn and Korschlitz. The largest willow2 in Berlin, which grew beside a canal and had a girth of 23 ft., fell to the ground in 1894. Beissner reports3 a tree, 92 ft. high and 9 ft. in girth in 1907, which is growing in a park at Hohenmistorf in Mecklenburg. As the park was laid out in 1854, this willow probably dates from that year. TIMBER The wood of this species is used for making cricket-bats of an inferior kind, those of the best class being mainly obtained from S. cœrîilea. Apart from this special use, the wood of the white willow is of considerable value, though it is not so much esteemed now as formerly. It is tough, and indents without splintering from blows or hard usage, and is used on that account for brakes on railway waggons, and for the sides and bottoms of carts ; it is also used for the rims of riddles and milk- pails, and by turners for making toys. It is also valuable for hurdle-making. G. W. Newton states* that George Stephenson had a high opinion of willow, as forming durable blocks for paving. Gorrie states5 that "in roofing it has been known to stand one hundred years as couples, and with the exception of about ^ in. on the 1 Mitt. Deut. Dend. Ges. 1910, p. 52, figs. 8 and 9. * Timber Trees, 34 (1859). 2 Ibid. 1894, p. 29. 3 Ibid. 1907, p. 56. 6 Loudon, Gard. Mag. i. 45 (1826). Salix 1763 outside, the wood has been found so fresh at the end of that period as to be fit for boat-building." Boards of willow were laid for floors in I7OO.1 (H. J. E.) SALIX CŒRULEA, CRICKET-BAT WILLOW Salix cœrulea, Smith, Eng. Bot. t. 2431 (1812), and in Rees, Cycl. xxxi. 141 (1819); Aiton, Hort. Kew. v. 365 (1813) (not Forbes). Salix alba, var. cœrulea, Smith, Eng. Flora, iv. 231, 232 (1828); Loudon, Arl. et Frut. Brit. iii. 1523 (1838); Bean, in Kew Bulletin, 1907, p. 312, plate, and 1912, p. 205. Salix viridis, Pratt,2 in Journ. Roy. Agric. Soc. Ixvi. 22 (1905), and Quart. Jonrn. Forestry, \. 325, fig. No. 02 (1707) (not Fries). A tree, attaining 100 ft. in height, with ascending branches, making a narrow angle with the stem, and forming a pyramidal crown.3 Terminal branchlets erect, not spreading or drooping. Bark smoother than in S. alba. Young branchlets appressed pubescent, becoming reddish brown in winter and the following year. Leaves similar to those of S. alba, but thinner in texture, more translucent, and less densely pubescent ; lower surface not white, but bluish grey ; margin ciliate, with minute glandular serrations. Pistillate catkins, differing from those of S. alba as follows: ovary4 slightly stalked, more tapering at the apex, about \ in. in length ; each of the two style- arms distinctly bilobed ; scale more concave, about two-thirds the length of the ovary. Fruiting capsule, nearly \ in. long, on a distinct short pedicel. This remarkable tree, which is best distinguished by its pyramidal habit, with stiff ascending branches and branchlets, is par excellence the true cricket-bat willow, as it exceeds the other kinds in rapidity of growth. Bean calls it the " best close- bark willow." In addition to its different habit, the leaves are readily distinguishable from those of S. alba by their different colour, and also by their translucency, as when viewed against the light with a lens, the tertiary venation is always plainly visible. In the forms5 of S. viridis, which approach S. alba in foliage, the leaves simulate those of S. cœrulea in colour and translucency, but are considerably larger and are also more coarsely serrate in margin ; and no form of S. viridis, known to me, has either the peculiar habit or rapid growth of S. cœrulea. S. cœrîilea is considered by most botanists to be a variety of S. alba ; but it differs from the latter in the flowers, which approximate in their size and shape and stalked ovaries to those of S. fragilis ; and it is possible that S. cœrulea may be the first cross between these two species, most of the characters of S. alba, if this hypo thesis is correct, being dominant. S. cœrulea was first distinguished by Smith, who was unable to find any botanical 1 Ellis, White Woods, quoted by Mitchell, Dendrologia, 56 (1827). 2 Mr. E. R. Pratt at first accepted, on Mr. Linton's identification of certain specimens, the name S. viridis for the true cricket-bat willow, though iu Jotim. Roy. Agric. Soc. Ixvi. 22 (1905), he agrees that the East Anglia cricket-bat willow has leaves indistinguishable from those of S. cœrulea. 3 Owing to the fastness of growth of S. cœrulea, the nodes are at greater intervals, and the crown of foliage in consequence is remarkably sparse. The bark is darker in colour than that of S. alba. * Bean, Kew Bull. 1907, p. 313, stales that the ovaries of S. carulea are identical with those of S. alba. A careful examination shows that the pyramidal tree with bluish foliage has always the distinct flowers described above. 6 These forms of S1, viridis are often referred to S. alba, var. carulea, especially on the Continent ; but I restrict the name S. cccrulea to the pistillate pyramidal tree here described. 1764 The Trees of Great Britain and Ireland character to separate it from S. alba, except " that the under side of the leaves loses at an early period most of the silky hairs." However, he went on to say : " Its qualities are of the highest importance. The superior value of the wood and bark, the rapid growth as well as handsome aspect of the tree, its silvery blue colour, its easy propagation and culture, in dry as well as wet situations, all render it so superior to our common white willow, that a cultivator might justly think lightly of any one who should tell him that there was no difference between them." He states that a cutting planted in Norfolk " became in ten years a tree 35 ft. high and 5 ft. 2 in. in girth, which is a rapidity of growth beyond all comparison with the common white willow." Succeeding botanists have applied the name cœrulea to pendulous slow-growing trees, of which the foliage is bluish white beneath ; but I consider that Smith distinctly described the quick-growing pyramidal tree which is now recognised by cultivators as the true cricket-bat willow ; and his specimen of 5". cœrulea in the herbarium of the Linnean Society, London, is undoubtedly this tree. Apparently no staminate tree1 of S. cœrulea exists ; and Smith knew it only as a female tree. The origin of S. cœrulea is obscure ; but it appears to be confined to the eastern counties of England, where it has been known2 since 1804 at least. Bean,8 relying upon Shaw, states that it is only found at the present time in Essex, Hertford, and Suffolk ; but it undoubtedly occurs also in Norfolk and Cambridgeshire ; and is said to have been formerly a rare tree in Kent and Surrey. Dealers restrict their purchases to the eastern counties, and have not yet ever found any suitable willow for making bats in other parts of Britain or on the Continent. It is probable, however, that S. cœrulea, when planted in other parts of the British Isles, will prove satisfactory, as there is no reason for believing that the peculiar qualities of the wood of this tree are dependent upon the climatic conditions of the eastern counties. Mr. J. A. Campbell, who planted in 1904 about 150 trees of 5. cœridea at Arduaine, Lochgilphead, Argyllshire, has received a favourable report4 from Mr. D. J. Carter, willow-dealer at Waltham Cross. One of these trees, which was cut down in 1912, when it was 6 in. in diameter, would, if it had been large enough for making bats, have fetched the normal price. This tree was too small to give a certain result ; but judging from Mr. Carter's report, 5. cœrulea, grown in the west of Scotland, where the rainfall is 60 inches annually, apparently retains the qualities which render its timber so valuable in East Anglia, where the climate is dry and sunny, and the mean summer temperature much higher than in the west.5 Loudon identified S. cœntlea with the upland or red-twigged willow of Ponteyc ; 1 Forbes, Sal. Wobum. 273, t. 137 (1829), describes as S. cœrulea, a male tree, which is doubtless a form of S. viridis. 2 Smith, Fl. Brit. iii. 1072 (1804), refers to it as a bluish, quick-growing variety of S. alba. 3 Kew Bulletin, 1907, p. 313. * Cf. Bean, in Kew Bulletin, 1912, p. 205. 6 Notwithstanding tbe above statement, I should hesitate to advise the planting of this willow in any part of Scotland or the west of England, until it has been proved by actual sale that the timber is of equal value to that grown in the eastern counties. Even supposing that it should prove to have the same qualities when large enough, it is not easy to convince leading manufacturers that their reputation is worth risking ; and though eminent cricketers whom I have consulted do not seem to judge bats so much by the appearance of their wood, as by their balance, handle, and weight, they would rather pay a high price for a bat which is guaranteed by the maker than use a substitute whose durability and driving powers are more or less un certain.—II. J. E. e Profitable Planter, 72 (1814). Sang, Nicol's Planter's Kaleiidar, 68 (1812), says that the red-twigged willow "forms a large tree and has a fine silvery foliage ; it is probably the same as the upland willow of Mr. Pontey." Salix but the latter's description is vague and uncertain ; and I have been unable to find any accounts of this remarkable tree as early as the eighteenth century. S. cœrulea is much faster in growth than either S. alba or any of the varieties of S. viridis. It is true that well-shaped trees of the latter kinds, grown with clean stems on proper soil, are frequently purchased at a fair priceT by bat-makers ; yet 5. cœriilca should always be preferred for planting. It comes to market earlier, on account of its rapid growth ; and produces a wood light in weight,2 very elastic and tough, which is found to be the most suitable for making the best kinds of cricket bats.3 CULTIVATION The best account of the cultivation of this tree is given by -Mr. E. R. Pratt, from whose article4 I shall quote largely. The choice of soil is most important, as many failures are due to the prevalent idea that willows will thrive in any wet or marshy situation. This is erroneous, and soils sodden with stagnant water or of a peaty nature should be avoided. The best ground is undoubtedly rich light alluvial land by the side of a running stream ; but good willows are also seen growing in fertile loam, where there is a good supply of moisture. Clay and gravel soils are usually quite unsuitable. If willows are planted in grass land beside a stream, they must be protected against cattle, during the whole period of their growth ; and, in all cases, rabbits must be excluded. In peaty soil, except in very rare cases where the drainage is good and the properties of the peat modified in consequence, willows never thrive, and after a few years often die. In such cases the soil, when examined, proves to be very acid in reaction ; and the willows are frequently attacked by a fungus, Physalospora gregaria, Saccardo, which produces cankerous spots on the young stems. The epidermis at first looks as if scorched, then dries up, turns brown, and becomes cracked by the protrusion of very small black spots (the fruit of the fungus) ; it ultimately peels off, exposing the inner part of the stem. Dr. T. Johnson, who has given a good account,6 with figures, of this fungus, believes that its ravages are much favoured by raw peaty soil ; and certainly, in some cases, it is the cause of extensive failures in willow plantations. In wet or marshy situations, where there is great growth of grass, the willow seems more liable also to the attacks of the beetle, Saperda carcharias, Linn. Though rooted cuttings are frequently advertised by nurserymen, there seems to be no doubt that it is much more advantageous to plant large sets. Scaling,6 who had great experience, says : " All varieties of tree willows grow better and more 1 Two trees in Suffolk, growing in good soil, which I examined before they were cut down—both S. alba, but one male, and the other female—were sold at 6s. 8d. per cubic foot in 1910 ; but these trees, twenty years after planting, had only attained 14 in. in diameter, and had short stems which made only four bat lengths. 2 Mr. L. Fotbrooke, Ravenstone Hall, Leicester, states in Card. Chron. xxxix. 46 (1906), "The close-barked white willow is Salix cœrulea. It is the quickest grower of all the tree willows in a moist soil, reaching 18 in. in diameter in as many years. The quicker the growth, the lighter the limber, and the better the price." 3 Mr. E. R. Trait, in Quait. Journ. Forestry, i. 336 (1907), shows that the specific gravity of the wood of S. alba (male trees) exceeds that of S. ccerulea (female trees) by 14^ per cent. 4 Journ. Roy. Agric. Soc. Ixvi. 19-34 (1905). 6 Proc. Roy. Dublin Soc. x. pt. ii. 153-166, plates 13-15 (1904). 6 Salix or Willow, i. Cat. p. 8 (1871). VII Q 1766 The Trees of Great Britain and Ireland vigorously from cuttings than from rooted plants." Gorrie] found that shoots of the white willow "6 to 8 ft. long and about 2 in. in diameter succeed better than rooted plants ; they require to be put from 18 in. to 2 ft. deep, in marshy soil, which should be drained." At the present time, it is the invariable custom amongst expert growers of cricket-bat willow to use large sets, up to 20 ft. long and 3 in. in diameter, and scarcely smaller than 6 to 10 ft. long by 2 in. in diameter. The sets are best obtained from young trees that have either been felled for sale or that have been specially pollarded for the purpose ; and should never be taken from the tops of old trees, as these are seldom straight, and require much trimming. A young tree, about ten years old, gives the best sets when pollarded, and can be again pollarded every five years for six or seven times till decay sets in. The sets should be cut in early spring. Mr. Pratt, after trimming them for four-fifths of their length, has them tied up in bundles of ten, and kept in water for a month, after which they are planted out. As the willow is a light-demanding tree, and the grower's object is to produce, as quickly as possible, a short stem clean of branches for about 12 to 15 ft., a good crown of foliage must be preserved from the start, and the trees should be planted so wide apart that they do not interfere with each other by lateral shade. If closely planted, they grow more slowly, and often develop an elliptical instead of a circular stem. The distance apart along the side of a stream should not be less than 10 yards. Close planting to kill undergrowth or grass is a mistake, as the latter, if necessary, can be removed by cultivation, though this is seldom done. Holes may be made for the sets by driving in a stake two or three feet deep, which can afterwards be levered out. The sharpened end of the set is then dropped into the hole, and tightly rammed in position. It is very important to insert the set deeply and firmly, so that it may not be shaken by the wind. The after care consists in rubbing off in the first three years, with the gloved hand, as high up as possible, the buds which appear on the stem, so as to prevent the development of side shoots. Mr. Pratt advocates the pruning of the stem afterwards to a height of 25 ft., but this is seldom if ever done in Herts or Essex, where growers are content with a stem clear of branches to 12 or 15 ft. Willow trees become saleable to bat-makers, when they are about 15 in. in diameter at six feet from the ground, or about 13 in. at twelve or fifteen feet up. The original set remains as a useless core in the centre of the stem. As the width of a bat is about 5^ in. and the clefts are taken radially, the minimum diameter should be 5^ in. + 5^ in.+ 2 or 3 in. (the diameter of the set) = 13 or 14 in. The trees may of course be allowed to grow for another period of years, until a second ring of clefts is formed around the first ring, or even for a further period ; but it is usually most profit able to dispose of the trees when they are young. The trees are generally sold standing, and are deemed of first quality when the stem is straight, clear of knots or branches, and covered with a smooth scaly bark, which is indicative of rapid growth. As the length of a bat is about 28 in., the trees 1 Loudon, Card. Mag. i. 46 (1826). J. Harrison, New Method, 20-24 ('766) recommended, in planting willows, the use of sets 16 ft. long, with all the side branches pruned off; to be preserved from cattle by three stakes two feet from the set, tied up with thorns ; and an after treatment of disbudding. Salix 1767 as soon as felled, which is done in winter, are cross-cut into lengths of 28 to 30 in., and these in turn are split into clefts. The clefts are split up along the radii so that the annual rings run from the front to the back of the bat. The best clefts come from the lower part of the tree, which is far tougher than the upper portion. In the best bats I have counted from seven to nine annual growths. The clefts are ultimately fashioned into blades, which are subjected to hydraulic pressure ; and it is here that the value of 5. cœrulea shows itself, as blades made of 5". fragilis are unable to stand the requisite amount of pressure. The further process of manufacture is detailed in a pamphlet* written by W. E. Bussey, which should be consulted by those interested in the growth of willow. The extraordinary value of the true cricket-bat willow is not exaggerated in the following statement,2 made by Mr. John Barker of Pishiobury, Sawbridgeworth : " A good set costs is. to is. 6d., and, if planted in a suitable soil and does well, is worth from .£5 to £S in fifteen years." He instances a case where a piece of land was bought for ^"50, and planted with willows, which were sold, when sixteen years old, for ^2000 in 1905. The following figures have been given to us by a reliable grower of willows in Herts. In December 1910 he sold twenty-four trees which grew on the bank of a stream, and had been thirteen years planted, for an average of ^5 per tree. In January 1912 he sold eleven willows, which had been planted fourteen years previously, for £Si. These trees averaged 55 to 60 ft. in height, and were clear of branches for about 18 ft., their stems ranging from 42 to 46 in. in girth at five feet from the ground. The best tree contained 12 cubic feet of timber, available for making bats, and as it sold for ^8, the price per cubic foot came to 135. 4d. In 1906 fifty-three trees growing on the same property were sold for ^190. These had short stems, averaging 13 in. in diameter, and yielding only three bat lengths. Many trees of remarkable size, but comparatively young, have been felled for conversion into cricket bats. One of the largest on record3 was a tree at Boreham, Essex, which was planted in 1835, and felled in 1888, when it was 101 ft. high and 5 ft. 9 in. in diameter. It weighed upwards of eleven tons, and was perfectly sound. It was felled by B. Warsop and Sons, who made from it no less than 1179 bats.4 Mr. H. Clinton-Baker tells us of another willow which grew in a field at Aspenden, near Buntingford, and was purchased by the same firm. It was 6 ft. in diameter, at three feet from the ground, and divided at five feet up into two stems, which were clear to a height of 50 ft, where they still measured 2 ft. in diameter. Mr. Stuart Surridge purchased for £25 in 1910 a tree near St. Albans, which was about 80 ft. high and measured 5 ft. in diameter at three feet from the ground. Judging from a photograph, it had a clean stem of about 16 feet. He states that the largest tree known to him grew at Robertsbridge, in Sussex, and measured 21 ft. in girth. This was felled in 1902, and produced over 1000 cricket bats. Sir Thomas B. Beevor made the following note in his copy of Evelyn's Sylva : ' Published in 1910 by Geo. G. Bussey and Co., Queen Victoria St., London, B.C. 2 Gard. Chron. xxxix. 62 (1906). 3 Trans. Eng. Arbor. Soc. iv. 122 (1899). * Mr. Edwin Savill informs us that he sold a tree in 1911 for £jo, but he has no measurements. The dealer told him that the number of feet utilized worked out at £i per cubic foot. 1768 The Trees of Great Britain and Ireland "Mar. 12, 1808. Willow tree, planted about eighteen years in a meadow at Wortham, in Suffolk, in length about 22 ft., and measured, at four feet from the ground, 7 ft. IO in., and in its bark appearing as green as when young. This willow is now found to be the ccerulean willow, and was taken down in 1816, and cut into boards which measured 22 in. broad ; and increased (in the last) eight years in girth to ii ft. 8 in." There are many fine trees on the Copped Hall estate, near Epping, where the agent, Mr. P. W. Dashwood, informed me in 1907 that he had refused a short time previously an offer of ^"1500 for 100 trees to be selected by the purchaser. He sold one tree for ^£25 which was afterwards re-sold for ^£40 to another dealer. This tree had a stem 3^ ft. in diameter, and was clean of branches for about 25 ft.; it yielded eleven lengths of bats, twenty-eight inches being allowed for a bat-length. In 1910 I measured at Ryston, Downham, Norfolk, a true S. cœmlea which was twenty years old and growing in good soil—alluvial silt over clay. It was 80 ft. high by 5 ft. 4 in. in girth. A male S1, alba of the same age growing beside it was 65 ft. high by 3 ft. n in. in girth. Mr. E. R. Pratt sent me a photograph of a tree, growing on the bank of the river Wissey at Hilgay, south of Downham, which, in October 1912, measured 84 ft. in height, and 6 ft. 5 in. in girth. This tree was planted in 1889 ; and is growing very rapidly, as it was only 5 ft. 6 in. in girth in February 1910. Elwes saw in 1907 a fine tree growing in a meadow near Spains Hall, Braintree, Essex, which was considered by Mr. A. W. Ruggles-Brise, to be a true cricket-bat willow. It was quite sound and healthy, and measured 90 ft. by 12^ ft. The illustration (Plate 381) is reproduced from a negative for which we are indebted to Mr. M. C. Duchesne. It was taken from a tree in a meadow called Hartham, near Hertford, of which the grazing is annually let by the municipality. The soil and situation are considered ideal for growing willow. (A. H.) SALIX VITELLINA, GOLDEN WILLOW Salix vitellina, Linnaeus, Sp. PI. 1016 (1753)5 Smith, Eng. Bot. t. 1389 (1805), and Eng. flora, iv. 182 (1828) ; Host, Salix, tt. 30, 31 (1828) : Forbes, Salic. Wobnrn. 39, t. 20 (1829); Loudon, Arb. et Frut. Brit. Hi. 1528 (1838). Salix alba, Linnaeus, var. vitellina, Stokes, Bot. Mat. Med. iv. 506 (1812); Seringe, Ess. Saule Suisse, 83 (1815) ; Andersson, in De Candolle, Prod. xvi. i, p. 211 (1868) ; Buchanan White, in Journ. Linn. Soc. (Bot.) xxvii. 371 (1890); Camus, Monog. Saules, 75 (1904). (?) Salix basfordiana* Scaling, Salix or Willow, Cat. p. 8 (1871), and ii. 19 (1872). A tree, attaining about 60 ft. in height, with spreading branches. Young branchlets, pubescent at the tips and near the nodes, becoming glabrous and bright yellow in winter, and in the following year. Leaves lanceolate, averaging z\ in. long and f in. wide, gradually tapering to a slender acuminate caudate apex ; shining green above, bluish white beneath, with scattered appressed silky hairs, usually slight 1 Cf. page 1757, note I. Salix 1769 on the upper surface, more abundant beneath ; serrations minute with incurved glandular tips ; margin either ciliate throughout the season, or with the cilia deciduous in summer ; petiole slightly pubescent. Catkins terminating short leafy branches and appearing with the leaves ; axes white tomentose, densely flowered. Staminate catkins, i to if in. long, curved ; stamens two, rarely three, base of filament slightly pilose ; scales concave, ovate- lanceolate, almost glabrescent, slightly ciliate, nearly twice as long as the stamens ; glands two, posterior quadrate and usually bilobed, anterior smaller and usually entire. Pistillate catkins, very slender, if in. long ; ovary with a distinct pedicel, which is twice as long as the posterior bilobed quadrate gland, conic, glabrous, about \ in. long, ending in a short style ; stylar arms spreading and bilobed ; scale as long as the ovary, pubescent on both surfaces. S1. vitellina is related to S1, alba, but is distinct in the flowers, and has narrower and less pubescent leaves. The flowers are occasionally unstable, three stylar arms and three stamens being present instead of two, and the ovary is often peculiarly inflated towards the apex. S1, vitellina is variable as regards the amount of the pubescence ; in one form the leaves are scarcely pubescent on the upper surface, and the margin becomes non-ciliate ; while in another form there is appressed pubescence on both surfaces, often dense beneath, and the cilia are retained on the margin till late in the season. S1, vitellina is possibly of hybrid origin, and though long known in cultivation and naturalised in many parts of Europe is very doubtfully wild. Smith states1 that Crowe found it wild in pastures at Ovington, near Watton, Norfolk. The following varieties are known :— 1. Vzx.pendula, Späth, Cat. No. 69, p. no (1888). Branchlets pendulous. Leaves narrow, non-ciliate. This is a beautiful weeping tree,2 which is often sold under the erroneous name of S1, babylonica, var. aurea. 2. Var. britzensis, Späth, Cat. No. 57, p. 67 (1883). Young branchlets bright red, pubescent. Leaves appressed-pubescent on both surfaces, ciliate till autumn. This is a staminate tree. It is the finest of all the coloured willows, the twigs assuming a beautiful red colour in winter. A thriving specimen at Glasnevin is about 40 ft. high, and is narrowly pyramidal in habit. The golden willow is planted for ornament both in England and the Continent, and is occasionally cultivated in osier beds.3 It is very striking in winter, and seems to thrive in this country, though Smith states that the twigs are often killed by severe cold, like those of 51. babylonica. We have seen no trees of great size, the finest being probably two trees at Glasnevin, which are about 65 ft. in height and 8 ft. in girth. These are probably of considerable age, and have rough bark like that of S1, fragilis. It was early introduced into North America, where it is now very common in New England. (A. H.) 1 Fl. Brit. iii. 1050 (1804). 2 Uode, in Bull. Soc. Bot. France, Iv. 655 (1910), considers this to be a hybrid between 51. vitellina and 51. balylonica, and names it 5. chrysocoma, Dode. 3 Cf. Ellmore and Okey, in Journ. Board. Agric. xviii. 914 (1912), who say that it is one of the toughest willows grown, if used with the bark on in a green state. Hence it mostly produces rods, which are used for tying purposes. POPULUS Populus, Linnaeus, Gen. PI. 307 (1737) and 456 (1754), Sp. PI. 1034(1753); Wesmael, in De Candolle, Prod. xvi. 2, p. 323 (1868), in Mem. Soc. Sei. Hainaut, iii. 186-250 (1869), and in Bull. Soc. Bot. Belg. xxvi. pt. i. p. 371 (1887); Bentham et Hooker, Gen. PI. iii. 412 (1880); Schneider, Laubholzkunde, i. 2-23 (1904), and ii. 869, 870 (1912); Dode, in Mem. Soc. Hist. Nat. Autun. xviii. 1-76 (1905); Ascherson and Graebner, Syn. Mittelettrop. flora, iv. 14-54 (1908); Gombocz, in Math. Termes. Ko'zl. xxx. 5-238, with two maps (1911). DECIDUOUS trees, belonging to the natural order Salicaceae. Terminal buds large, with one or two outer pairs of opposite scales at the base, and several imbricated scales above ; axillary buds smaller, with fewer scales, the lowest of which is short, broad, and open next the stem ; scales accrescent, marking when they fall the base of the branchlet with ring-like scars. Branchlets terete or angled, with five-angled pith, and showing, when the leaves have fallen, three-dotted leaf-scars, on the sides of which are visible two minute scars left by the early deciduous stipules. Leaves simple, alternate, penninerved, usually long-stalked ; entire, dentate, or lobed ; often different in shape, pubescence, and margin on the long and on the short shoots. Flowers without honey, fertilised by the wind, appearing before the leaves in early spring ; dioecious, in pendulous stalked catkins, which arise from buds in the axils of the leaf-scars of the previous year or at the ends of short shoots. Catkins bearing on a slender axis numerous flowers, each of which is subtended by a caducous stalked lobed, dentate, or laciniate scale (bract). Perianth absent, replaced by a stalked disc. Male flowers densely crowded ; stamens four to twelve, or twenty to sixty, with short white filaments, and purple or red two-celled anthers, arising from the oblique, flat or concave disc. Pistillate flowers not so dense in the catkin as the male flowers ; ovary sessile in the oblique cup-shaped disc, one- celled, with two, three, or four placentae ; style short or obsolete, with as many entire or bifid stigmas as there are placentae. Fruiting catkins elongated, ripening early before the leaves are fully grown ; capsule one-celled, separating when ripe into two to four recurved valves, girt at the base by the persistent disc.1 Seeds numerous, minute, without albumen, elliptic, compressed, acuminate at the apex, surrounded by tufts of long white silky hairs, attached to their short stalks and deciduous with them. Seedling,2 with two stalked suborbicular cotyledons, sagittate at the base, thick and succulent ; primary leaves either in opposite pairs as in P. canescens, or alternate as in the black poplars. 1 In section Turanga the disc is deciduous, and does not persist on the fruit. 2 An account of the germination of poplars, with figures of the seedlings, is given by Miss F. II. Woolward, in Journ. Bot. xlv. 417, t. 487 (1907). Cf. also Ilickel, in Bull. Soc. Deiid. France, No. 25, p. 88 (1912). 1770 Populus 1771 The poplars are typical light-demanding trees, incapable of bearing shade, their branchlets and leaves dying when not exposed to full light. This is well seen when two Lombardy poplars are planted close together, the shade of the taller of the two killing the branches on the adjacent side of the other. In connection with this demand for light, which necessitates sparse branches and foliage, the poplars normally shed many of their smaller branchlets in autumn. The process1 by which this is effected is similar to that by which the leaves are cast off,—a zone of corky tissue being formed at the point where the rupture subsequently takes place,—the branchlet leaving when it falls a circular scar on the main branch to which it was attached. The genus Populus comprises about twenty-five species, inhabiting the extra- tropical regions of the northern hemisphere from the Arctic Circle southwards ; in North America extending to Lower California and northern Mexico ; throughout Europe and northern Africa ; and in Asia, extending as far south as Asia Minor, Syria, Mesopotamia, Persia, Afghanistan, the Himalayas, China, and Japan. Towards the extreme north certain species often form great forests ; elsewhere poplars are most common in alluvial land bordering rivers, streams, and swamps ; but occasionally they form a part of the deciduous forests. The genus is divided into five sections ; and the following key, based mainly on the characters of the leaves and buds, includes all the species in cultivation, with the more important hybrids ; and in addition, three species, mainly of botanical interest, which we have not seen in England in the living state. I. TURANGA, Bunge, Beit. Kennt. Fl. Russ. 498 (1848). This section differs from the others in the remarkable polymorphic leaves, and in the deeply cleft disc of the flowers, which does not remain persistent at the base of the fruit. i. Populus euphratica, Olivier.2 Northern and eastern Africa, Syria, Mesopotamia, Persia, Turkestan, Afghanistan, north-west India, Mongolia, north China. A tree 50 ft. high. Leaves coriaceous, greyish green, of the same colour on both surfaces ; on young trees, linear or oblong, entire, short-stalked, and willow- like ; on old trees extremely diverse, ovate, oblong, rhombic, or orbicular, lobed or cut, long-stalked. Not hardy in Great Britain.3 This poplar, and not a willow,4 is the 'arabim of the Psalms, cxxxvii. 2, the trees growing by the rivers of Babylon, on which the Jews in captivity hanged their harps.6 1 This natural fall of branchlets, effected by a vital process, was termed cladoptosis by Berkeley, in Gard. Chron. 1855, p. 596. It has been observed in oaks and willows, as well as in poplars, and Shattock gave a complete account of it in the case of the aspen 'mjourn. Bot. xxi. 306 (1883). This phenomenon appears to have been first noticed by J. Main, Hort. Register, iv. 193, fig. A (1835), where a fallen branch of the black poplar is figured. 2 Voy. Emp. Othom. figs. 45, 46 (1807). This is a very variable species, which has been variously treated by botanists. Gombocz, in Math. Termes. Kösl. xxx. 71, 72 (1911), recognises several varieties, and treats P. pruinosa, Schrenk, as a distinct species. P. illicitana, Dode, in Bull. Soc. Dencl. France, 1908, p. 163, lately found near Elche in Spain, appears to be only naturalized there, and is identical with the ordinary form of the species in Morocco and Algeria. 3 Spath, Catalogue, No. 91, p. 51 (1893-1894), states that a plant sent to Lauche from Turkestan in 1881 soon died. It was reintroduced by General Korolkow, who sent it to Späth in 1892. 1 Salix babylonien, Linnœus, was so called, because it was erroneously supposed to be the tree of the Psalms. Cf. p. 1752. *> Cf. Koch, Dendrologie, ii. pt. i. p. 507 (1872), and Ascherson, in Sitzb. Ges. Nat. Fr. Berlin, 1872, p. 92. I772 The Trees of Great Britain and Ireland II. LEUCE, Duby, in De Candolle, Syn. PI. Fl. Gall. i. 427 (1828). White poplars and aspens. — Bark smooth on young stems, ultimately breaking on the surface into rough rhombic cavities. Buds variable, tomentose and dry, or glabrous and viscid. Leaves variable, tomentose or glabrous ; lobed, dentate, or serrate. Flower scales fringed with long hairs ; stamens few, about ten ; capsules slender, oblong. A. White poplars.—Leaves densely tomentose on the long shoots ; less tomentose or glabrescent, and different in shape on the short shoots. 2. Popitlus alba, Linnaeus. Europe, North Africa, Asia Minor, Caucasus, Central Asia, Himalayas. See p. 1777. Leaves on the long shoots palmately lobed, snowy white beneath ; on the short shoots oval, dentate, greyish beneath. 3. Popiilus tomentosa, Carrière. North China. Seep. 1786. Leaves on the long shoots triangular-ovate, without lobes, biserrate, grey tomentose beneath ; on the short shoots, with a few sinuate teeth, glabrescent beneath. 4. Populus canescens, Smith. Western Europe. Seep. 1780. Leaves on the long shoots ovate-deltoid, greyish white beneath, with irregular triangular serrated teeth ; on the short shoots suborbicular or broadly ovate, glabrescent beneath, with a few sinuate teeth. B. Aspens.—Leaves on the long and short shoots not markedly different, glabrous or glabrescent beneath. * Branchlets glabrous. 5. Populus tremula, Linnaeus. Europe, North Africa, Asia Minor, Caucasus, Siberia. See p. 1787. Leaves suborbicular, i^ to 2 in. in diameter, thin in texture, acute or rounded at the apex ; margin with rounded or sinuate teeth. 6. Populus tremuloides, Michaux. North America. See p. 1791. Leaves orbicular or ovate, i^ to 2 in. in diameter, thin in texture, cuspidate at the apex ; margin ciliate, finely and regularly serrate. * * Branchlets slightly tomentose. 7. Populus Sieboldii, Miquel. Japan. See p. 1794. Leaves ovate, 3 in. long, 2 in. wide, thick in texture, shortly acuminate at the apex ; margin with minute sinuate teeth or glandular serrations. 8. Populus grandidentata, Michaux. North America. See p. 1792. Leaves ovate-deltoid, 3 to 4 in. long, 2 to 3 in. wide, thin in texture, acuminate ; margin with a few sinuate triangular large teeth. III. AIGEIROS, Duby, in De Candolle, Syn. PI. Fl. Gall. \. 427 (1828). Black poplars.—Trees with furrowed bark. Buds viscid, but not very odorous. Leaves green on both surfaces, with a clearly defined translucent border. Populus 1773 A. Leaves never ciliate in margin. 9. Populus nigra, Linnaeus. Europe, Caucasus, Siberia. See p. 1795. Leaves rhomboid, about 3 in. long, 2 in. broad, cuneate at the base, gradually tapering above into a long acuminate apex ; glands never present at the base. Stigmas always two. Stamens fifteen to thirty, purple. Branchlets and petioles glabrous in var. typica, pubescent in var. betulifolia,. The Lombardy poplar (p. 1798) is the fastigiate form of var. typica ; and var. plantierensis (p. 1802) is the fastigiate form of var. betulifolia. B. Leaves ciliate in margin. * Glands absent at the base of the leaf. ID. Populus Fremontii, Watson. North America. See p. 1794. Leaves on young cultivated trees, reniform or rhombic, with a cuneate base ; on old trees deltoid with a truncate base ; about ï\ in. wide ; apex cuspidate ; serrations few, coarse, and incurved. Stigmas three. Stamens about sixty, with dark red anthers. * * Glands always present on the base of the leaf. 11. Populus monilifera, Aiton. North America. Seep. 1807. Branchlets rounded. Leaves deltoid-ovate, about 3 in. wide, shallowly cordate or truncate at the base, cuspidate at the apex ; serrations sinuate, with incurved tips, fewer and coarser than in the hybrids. Stigmas three or four. Stamens fifty to sixty. 12. Populus angulata, Aiton. North America. Seep. 1810. Branchlets angled. Leaves triangular-ovate, longer than broad, up to 7 in. long, and 5 in. broad, truncate or cordate at the base, acute or shortly acuminate at the apex. Stigmas four. Stamens fifty to sixty. *** Glands variable at the base of the leaf, absent or one or two in number. Hybrids between the European P. nigra, and one or other of the American P. monilifera and P. angulata. Leaves with irregular marginal cilia, which are often sparse, and usually deciduous in summer ; serrations not coarse and sinuate. * Branchlets glabrous. 13. Populus serotina, H artig. Seep. 1816. Branches ascending, wide-spreading. Leaves unfolding latest of all the poplars and with a red tint, ovate-deltoid, about 3 in. wide, with a broad truncate base, and a short cuspidate or acuminate apex. A staminate tree ; stamens twenty to twenty-five. 14. Popitlus regenerata, Schneider. See p. 1824. Similar to P. Eugenei in habit, with foliage like that of P. serotina, but unfolding a fortnight earlier. A pistillate tree ; stigmas usually only two. 15. Populus Eugenei, Simon- Louis. Seep. 1826. A narrow columnar tree, with ascending short branches. Leaves unfolding early with a reddish tint, smaller than those of P. serotina, broadly cuneate at the base, with a slender sharp-pointed non-serrated acuminate apex. A staminate tree ; stamens about twenty, vu R 1774 The Trees of Great Britain and Ireland 16. Populus marilandica, Bosc. See p. 1828. A tree with irregular branches. Leaves early in unfolding, rhomboid, cuneate at the base, tapering above into a long acuminate apex. A pistillate tree ; stigmas variable, two, three, or four. 17. Populus Henryana, Dode. See p. 1829. A tree with irregular branches. Leaves opening early, not tinged with red, ovate-triangular, cuneate at the broad base, and ending in a long acuminate apex. A staminate tree ; stamens thirty to thirty-five. Flower buds pubescent. * * Branchlets pubescent. 18. Populus robusta, Schneider. Seep. 1829. Branchlets grey. Leaves unfolding early with a deep red tint, variable in shape, ovate-deltoid or rhomboid ; cuneate, rounded, or truncate at the base ; acuminate at the apex ; petioles pubescent. A staminate tree ; stamens twenty. 19. Populus Lloydii, Henry. Seep. 1830. Branchlets yellow. Leaves similar to those of P. robusta, but smaller ; petioles pubescent. A pistillate tree ; stigmas usually two. IV. TACAMAHACA, Spach, in Ann. Sei. Nat. xv. 32 (1841). Balsam poplars.—Trees with rough furrowed bark, and fragrant foliage. Buds very viscid, exhaling a strong balsamic odour. Leaves whitish beneath, without a clearly defined translucent cartilaginous border ; petioles rounded or quadrangular, channelled on their upper side. A. Branchlets rounded, without projecting ribs, except occasionally on vigorous shoots of young trees. * Branchlets and petioles pubescent. 20. Populus candicans, Aiton. North America. See p. 1834. Leaves broad, ovate-deltoid, palminerved and cordate at the base, 5 to 6 in. long, 4 in. wide, ciliate in margin. 21. Popîilus tristis, Fischer. Himalayas (?). Seep. 1840. Leaves narrow, ovate, palminerved and cordate or rounded at the base, 4 in. long, 2 in. wide, ciliate in margin. 22. Populus Maximowiczii, Henry. Japan, Amurland, Manchuria, Korea. See p. 1838. Leaves nearly orbicular, oval, or elliptic ; palminerved and subcordate at the rounded base ; 4 in. long, 3 to 3^ in. wide ; densely ciliate in margin, and pubescent on the midrib, nerves, and veinlets of both surfaces. 23. Populus suaveolens, Fischer. Siberia, Mongolia. Seep. 1841. Leaves ovate or ovate-lanceolate, palminerved and rounded at the base, abruptly contracted into an acuminate apex, 3 to 3.^ in. long, \\ to 2 in. broad ; margin ciliate ; glabrous on both surfaces. * * Branchlets and petioles glabrous. 24. Popuhis balsamifera, Linnaeus. North America. See p. 1832. Leaves ovate, palminerved and rounded at the base, 4 to 5 in. long, 2 to 3 in. wide, sparsely and minutely ciliate in margin. Populus 25. Populus angustifolia? James. Rocky Mountains of North America. See P. 1831. Leaves lanceolate, resembling those of Salix fragilis in shape, cuneate at the base, greenish beneath, 2 to 4 in. long, f to i in. wide ; lateral nerves, fifteen pairs, all pinnate. B. Branchlets with projecting linear ridges. * Branchlets pubescent. 26. Populus laurifolia, Ledebour. Altai Mountains. See p. 1842. A wide-spreading tree. Leaves lanceolate, 3 to 5 in. long, i to 2 in. broad, rounded at the base, gradually tapering to an acuminate apex, finely and regularly serrate. 27. Populus berolinensis? Dippel. A hybrid. See p. 1844. A narrow columnar tree. Leaves ovate or ovate-rhombic, 3 in. long, 2 in. wide, rounded or cuneate at the base, contracted at the apex into an acuminate point, crenately and occasionally irregularly serrate. * * Branchlets glabrous. 28. Populus Wobstii, Schrceder. A hybrid. See p. 1843. Branchlets only slightly ribbed in the first year, the ribs more apparent in the second year. Leaves lanceolate, tapering to a narrow rounded base, 4 to 6 in long, 2 in. broad ; ciliate in margin ; nerves all pinnate. 29. Populus Simonii, Carrière. North China. See p. 1839. Leaves rhombic-elliptic, cuneate and pinnately nerved at the base, contracted at the apex into a short cuspidate point, 3 in. long, i J in. broad. 30. Populus trichocarpa, Torrey and Gray. Western North America. Seep. 1836. Young stems with bark peeling off in papery shreds. Leaves ovate or ovate-deltoid, palminerved and rounded or subcordate at the base ; 5 in. long, 3 in. broad ; whitest beneath of all the balsam poplars. V. LEUCOIDES, Spach, in Ann. Sei. Nat. xv. 30 (1841). Trees with rough bark, breaking into loosely attached plates. Leaves very large, cordate, simply serrate, covered when unfolding with greyish tomentum, which speedily disappears except on the nerves beneath ; petioles rounded, much shorter than the blades, which do not flutter in the wind. Buds pubescent, viscid. These poplars are not easily propagated by cuttings. 31. Populus lasiocarpa, Oliver. Central China. Seep. 1846. Leaves ovate-cordate, longer than wide, about 9 in. long, and 6 in. broad ; green beneath, with uniform serrations, extending regularly to the sinus at the base. 32. Populus Jieterophylla, Linnaeus. United States, from the Mississippi valley eastwards to the Atlantic coast. See p. 1846, note 2. 1 This species is very distinct from the other balsam poplars, in the willow-like leaves, which are not whitish beneath. 2 This hybrid differs from the balsam poplars in the very narrow translucent border on the leaves, which are less fragrant and greenish, or only slightly whitish beneath. P. rasunurwskyana and P. petrvoiskyana are similar hybrids, of which I have not seen complete material. See pp. 1843, 1844. 1776 The Trees of Great Britain and Ireland Leaves ovate-cordate, nearly as broad as long, 4 to 7 in. long, 3 to 6 in. wide ; pale beneath, with regular serrations except at the base, where they are few and wide apart. (A. H.) THE RAISING OF POPLARS FROM SEED No good description or illustration of the germination of poplars seems to have been published in England before that of Miss Florence Woolward1 in 1907. I have never found in England a poplar grown from seed either naturally or by nurserymen ; and though Grigor3 describes the process, it seems doubtful whether he ever practised it ; and all my own attempts to raise poplars from seed were fruit less, until I followed nature as closely as possible. Having observed that poplars only germinated freely on the sandy banks of rivers, I sowed seed as soon as ripe on a pot of sand, and placed the pot in a pan of water in the full sun. Germination was then extremely rapid, according to Miss Woolward, in ten hours to two days. 1 found that after four days no more seeds came up, and that in all cases the pro portion of germinating seed was quite small. In five or six days the cotyledons are well developed, but the growth of the young plants is very slow and does not exceed 2 to 5 inches in height in the first season. Some one-year seedlings of P. nigra, sent me from the banks of the Allier in central France, were only 3 to 4 inches above ground, though their thick rather fleshy root was 6 to 10 inches long. Some seedlings of P. monilifera from the banks of the St. Lawrence in Canada, collected by Mr. Jack, were equally small ; as were some of P. nigra raised by Mr. Hankins from seed collected at Bury St. Edmunds. Miss Woolward and I also raised P. canescens from seed collected at Upcot near Colesborne, where both sexes of this tree grow on my own property ; and she raised P. marilandica from seed of a tree in Kew Gardens. In both cases the seedlings were much less vigorous than cuttings from the same trees. I also raised seeds collected under a tree, which I believe to be P. monilifera, of American origin, in the botanic garden at Padua, which was perhaps fertilized by a P. alba, the only poplar growing near it ; but none of those seedlings have shown the least sign of hybrid origin, and have grown slowly and seem tender as compared with poplars raised from cuttings. I am therefore convinced that, though we may succeed, by crossing different species, in obtaining new races of superior vigour, like P. Eugenei and P. robusta, the raising of poplars from seed is not a practice which can be recommended for general purposes. Some species of poplar are said to be difficult or impossible to strike from cuttings, among which Grigor includes P. alba and P. canescens, but I have found no difficulty in the case of the latter. In the case of species which will not strike, recourse must be had to layers, root-suckers, or grafting. (H. J. E.) 1 Journal of Botany, xlv. 417, t. 487 (1907). 2 Arboriculture, 328 (1881). Hickel, in Bull. Soc. Dend. France, No. 25, p. 88 (1912), gives directions for raising poplars from seed, which he has followed with success at Versailles. Populus 1777 POPULUS ALBA, WHITE POPLAR Populus alba, Linnœus, Sf. PI. 1034 (1753) ; Loudon, Arb. et Frut. Brit. iii. 1638 (1838) (in part); Willkomm, Forstl. Fl. 516 (1887); Mathieu, Flore Forestière, 483 (1897); Schneider, Laubhoh kunde, i. 2i (1904); Ascherson and Graebner, Syn. Mitteleurop. Fl. iv. 17 (1908); Boissier, Flora Orient, iv. 1193 (1879); Hooker, Fl. Brit. India, v. 638 (1888); Aigret, in Ann. Trav. Publ. Belg. x. 1213 (1905); Gombocz, in Math. Termes. Kbzl. xxx. 141 (1911). Populus major, Miller, Diet. ed. viii. No. 4 (i 768). Popitlus nivea, Willdenow, Berl. Baumz. 227 (1796). A tree, similar in size and bark to P. canescens. Young branchlets and buds densely white tomentose.1 Leaves (Plate 408, Fig. i), very variable in size and shape, dependent upon the position and vigour of the branchlets and on the age of the tree ; on vigorous long shoots and suckers, large, up to 4 or 5 in. in diameter, ovate, palmately three- to five-lobed ; lobes triangular, with deep sinuses ; margin ciliate with minute glandular teeth ; subcordate at the rounded or broad base, acute at the apex ; dark green and slightly tomentose above ; covered beneath with a dense snowy white tomentum ; petiole rounded, tomentose. Leaves on short shoots (and occasionally at the base of the long shoots), elongated oval, rounded at the base, with a few sinuate teeth, usually covered beneath with a thin greyish tomentum ; in some cases (var. denudata, Wesmael) glabrescent ; in some forms, broad and suborbicular like the similarly placed leaves of P, canescens, but uniformly tomentose beneath, and with fewer teeth in the margin ; petiole tomentose. Pistillate catkins similar to those of P. canescens, but more silvery in colour ; scales obovate to lanceolate, with minute teeth, and fringed with long hairs ; stigmatic lobes four, green, erect, long and linear, each of the two pairs united at the base. Staminate catkins2 (probably not perfectly developed), i inch long; axis tomentose; scales concave, spatulate, dentate, fringed with long hairs; disc pubescent; stamens six. Spread over an immense area, P. alba is very variable in the form of the foliage, and might be divided into several distinct geographical races ; but the wild specimens in herbaria are scanty and incomplete ; and only a few important varieties will be here mentioned. P. canescens, a native English tree, and P. tomentosa, wild in North China, are easily distinguishable ; and are treated here as distinct species. i. Var. nivea, Aiton, Hort. Kew. iii. 405 (1789). This is the typical form3 of the species described above. It has been long in cultivation in western Europe, where it is known in gardens as var. argentea, var. acerifolia, var. arembergiana, etc. It appears to be indigenous in eastern and south eastern Europe, in the Caucasus, in Persia, Turkestan, Afghanistan, and the Altai Mountains, where specimens have been gathered by travellers. In Kashmir, accord ing to Mr. Lovegrove, of the Forest Department, from whom I have received specimens, the typical form is fairly common in the wild state, attaining on an average 90 ft. in height and 8 ft. in girth. In cultivated trees, the leaves are extraordinarily variable in size and shape, but it 1 The suckers of P. alba are described by Du bard, in Ann. Sei. Nat. xvii. 163 (1903). ! The only male flowers of P. alba which I have been able to examine were produced by a small tree at Kew, labelled var. nivea, and did not seem to be properly developed. Most, if not all, of the white poplars in Britain are female trees. 3 Gombocz, op. cit. 148, 149, recognizes as forms of var. nivea, the slight variations, which are doubtfully constant, called by Dode, op. cit. 21, 22, P. triloba, P. Treyviana, P. Paletskyana, P. heteroloba, P. Morisetiana, and P. palmata. 1778 The Trees of Great Britain and Ireland is doubtful if these constitute distinct varieties ; on one tree I have observed ordinary branches with small leaves, varying in shape from oval and slightly toothed to trilobed, and more vigorous shoots showing large palmately lobed leaves. Old trees put forth feeble long shoots, on which the terminal white leaves are often imperfectly developed. 2. Var. subintegerrima, Lange, in Willkomm and Lange, Prod. Fl. Hisp. i. 233 (1861). Var. integrifolia, Ball, in Journ. Linn. Soc. (Bot.) xvi. 668 (1878). Populus subintegerrima, Dode, op. cit. 20 (1905). Populus monticola, Brandegee, in Zoë, \. 274 (1890); Sargent, in Garden and Forest, iv. 330, fig. 56 (1891), vi. 190 (1893), and vii. 313, fig. 51 (1894). Populus Brandegeei, Schneider, Laubholzkunde, i. 23 note, and 803 (1906). Leaves coriaceous, often sub-evergreen ; on the long shoots ovate or orbicular, sub-cordate or cuneate at the base, sub-entire or irregularly and slightly toothed, white beneath ; on the short shoots almost orbicular, slightly sinuate or quite entire, grey beneath. A native of southern Spain, Algeria,1 and Morocco, where it was gathered in the greater Atlas by Hooker. It occurs also in the Canaries and the Azores. It appears to have been introduced into Mexico by the Spaniards, and has been found, apparently naturalised, along streams in the high mountains of Lower California, where it is called guerigo by the inhabitants. According to Sargent,2 who speaks of it as a distinct native American species, its wood in this locality is quite unlike that of the other poplars, being light red, satiny, and useful for making furniture. Similar forms3 occur near Askabad, east of the Caspian Sea, and in Kashmir. 3. Var. pyramidalis,* Bunge, in Rel. Bot. Mém. Ac. St. Pétersb. vii. 498 (1851). Var. Bolleana, Lauche, ex Huttig, in Deut. Gart. 500 (1878); Masters, in Gard. Chron. x. 502 (1878) and xviii. 556 (1882). Populus Bolleana, Carrière, in Rev. Hort. lui. 40, 123 (1881), and Ixiii. 188, fig. 48 (1891); Masters, in Gard. Chron. xviii. fig. 96 (1882). Resembling the Lombardy poplar in habit. Leaves on the long shoots pal mately 3- to 5-lobed, very white tomentose beneath ; on the short shoots orbicular, with coarse triangular teeth, green beneath with traces of tomentum. The fastigiate form of the white poplar was first described by Bunge from specimens found by Lehmann in 1841, apparently wild6 on the bank of a stream on the north side of the Karatau mountain, between Bokhara and Samarkand. It appears to have been introduced6 in 1872 into Hoser's nursery at Warsaw, from a cutting sent by Col. Korolkow.7 Lauche procured it for the Horticultural Society of Potsdam from the same source in 1875. 1 Collected near Ronda in Spain, by M. P. Price, and near Affreville in Algeria by A. Henry. A specimen from Gibraltar, in the Cambridge Herbarium, has slender female catkins, 4 to 5 in. long, with a woolly axis and pedicellate flowers ; scales long, concave, irregularly toothed or lacerate ; slyles 4, spreading. 3 In Garden and Forest, vi. 190 (1893). 3 Specimens in the Kew Herbarium. 4 P. alba, var. croatica, Wesmael, in De Candolle, Prod. xvi. 2, p. 324 (1868) is an erroneous name founded on P. croatica, Waldstein and Kitaibel, in Flora, xv. 2 Beil. p. 14 (1832), which is a narrow form of P. nigra. Cf. Koch, Dendrologie, ii. pt. i. p. 489 (1872). 6 Lehmann also saw it planted at Bokhara, where he gathered a leafless branch bearing staminate flowers, on 5th March 1.842. ° Cf. 77« Garden, loth Dec. 1887, p. 543- 7 According to Carrière, Reu. Hort. Ixiii. 188 (1891), it was introduced by Col. Korolkow, under the name of P. alba pyramidalis into various places in France in 1878, as Orleans, Segrer, Angers; and was first sold by Simon-Louis in 1879- 1880. E. Morren, in Belg. Hort. 1879, p. 269, says that it was sent to Späth from Tiflis by Scharrer in 1879. Populus I779 The oldest tree in England is probably a fine specimen at Kew, growing near the large Ginkgo tree, which measured 67 ft. by 4 ft. 8 in. in 1910. Another at Terling Place, Essex, planted in 1886, measured, in 1910, 60 ft. high by 3 ft. 3 in. in girth. A good specimen at Over Bridge, near Gloucester, was 55 ft. high in 1911. The largest tree we know of in Germany is one at Späth's nursery, which in 1908 was 66 ft. by 4 ft. It was first sent out from here in 1879. 4. Mzr.globosa, Späth, Cat., No. 66, p. 3 (1886) ; Dippel, Laubholz, ii. 191 (1892). A small tree or shrub, of dense habit, oval in outline, with small slightly lobed deltoid cordate leaves, grayish beneath, and said to be pink in the young state. This has been propagated by Späth of Berlin, since 1886, when it originated in their nursery from a cutting of P. alba ; but is little known in England. 5. Var. Richardi. Leaves yellow on the upper surface, the colour lasting throughout the season. This was shown at the International Horticultural Exhibition of 1912 by Richard, nurseryman at Naarden-bussum, near Amsterdam. 6. Var. pendîda, Loudon, Arb. et Frut. Brit. iii. 1640 (1838). Branchlets pendulous. This is mentioned by Loudon as a continental variety not introduced in 1838. It has lately been noticed1 at Breslau ; but we have seen no specimens. DISTRIBUTION Both the grey and the white poplars have been so widely planted in Europe that their exact distribution in the wild state is very uncertain. According to Will komm, P. alba is certainly not a native of northern Germany, or of Scandinavia, though it thrives as a planted tree in Norway as far north as lat. 68°. Apparently P. canescens is a native of the British Isles, central and northern France, and of the Rhine valley in Baden and Alsace ; while typical P. alba is more southerly and easterly in its distribution, occurring throughout the Mediterranean region in Spain and Portugal, Italy, the Balkan States, Greece, Algeria, Morocco, Syria, and Asia Minor, and extending northwards into the Alps and Carpathian ranges, being a characteristic tree of the alluvial flats of the Danube and all its tributaries from Bavaria, throughout Austria and Hungary, to its mouth in the Black Sea. Farther east, P. alba occurs in Russia, the Caucasus, western Siberia, the north-western Himalayas, and western Tibet. It is usually a native of low-lying moist woods, especially those fringing the banks of rivers ; but it ascends in the valleys of the Alps, and in Spain and Italy to about 2000 feet elevation. The white poplar is certainly not indigenous in England, and records of it as a wild tree all probably refer to P. canescens, which often shows a considerable amount of white foliage on the long shoots on young and vigorous trees. Turner,2 in 1568, says, " As touching the whyte aspe, I remember not that I ever saw it in any place in England," and he appears to have known the tree, as he says it was plentiful in Germany and Italy. The date of the introduction of P. alba is uncertain.3 The 1 Behnsch, in Mitt. Deut. Dend. Ges. 1906, p. 212. 2 ffertall, 99 (1568). 3 Hartlib, Complète Husbandman (1659), ex Loudon, op. cit. 1641, states that some years before the time of his writing, 10,000 abeles were sent into England from Flanders, and transplanted into many counties. These trees were probably P. canescens. 1780 The Trees of Great Britain and Ireland white poplar is often called abele in England, but this is the Dutch name, abeel,1 of P. canescens. The white poplar is known in Dutch and Belgian nurseries as peuplier blanc de Hollande. The true white poplar has never become a common tree in England. Most of the large trees known as white poplars that we have seen are P. canescens, and it is impossible to separate the two species in Loudon's account of them. There are three good trees of P. alba at Bayfordbury, Hertford, the largest of which was 95 ft. by io ft. 4 in. in 1911. Mr. J. E. Little of Crofton, Hitchin, who has kindly sent specimens, tells me that between Norton Mill and Radwell Mill, near Baldock, Herts, on the bank of the river Ivel, there are four large white poplars, about 80 feet in height and 12 to 13 feet in girth. At Syston Park, Grantham, Miss F. H. Woolward, in 1905, measured a tree 105 ft. by io ft. io in. In Ireland the finest are at Adare Manor, where Mr. R. A. Phillips measured a tree 80 ft. by 9 ft. in 1910. Another at Nenagh was 60 ft. by 5 ft. in 1911. All these trees bear female flowers. Pynaert2 saw in 1882, at Troyes in France, a true white poplar which measured 140 ft. in height and 21 ft. 3 in. in girth, at six and a half feet above the ground. It divided at thirty-one feet up into three main stems, the largest of which girthed 14 ft. 9 in.; while the total spread of branches was 260 ft. in circumference. This magnificent tree, which was supposed to be about 400 years old, was destroyed3 by a storm on ist February, 1902. I saw, in 1912, a fine tree in the botanic garden at Toulouse, 90 ft. high by 8 ft. in girth. (A. H.) POPULUS CANESCENS, GREY POPLAR Populus canescens, Smith, Fl. Brit. iii. 1080 (1805), and Eng. Flora, 245 (1828) ; Smith and Sowerby, Eng. Bot. xxiii. t. 1619 (1806), and vii. 114, t. 1392 (1840); Loudon, Arb. et Frut. Brit. iii. 1639 (1838); Bromfield, in Phytologist, iii. 841 (1850); Schneider, Laubholzkunde, i. 23 (1904); Dode, in Mém. Soc. Hist. Nat. Autun, xviii. 26 (1905); Aigret, in Ann. Trav. Publ. Belg. x. 1214 (1905). Populus alba, Linnaeus, var. ß, foliis minorihis, Lamarck, FÎ. Franc, ii. 235 (1778). Populus alba, Linnaeus, var. canescens, Aiton, Hort. Kew. iii. 405 (1789). Populus alba, Willdenow, Berl. Baumz. 227 (1796) (not Linnaeus) ; Hunter, in Evelyn, Suva, i. 208, plate (1801); Smith, Eng. Bot. t. 1618 (excl. description) (1806). Populus alba, Linnaeus, var. genuina, Wesmael, in De Candolle, Prod. xvi. 2, p. 324 (1864). Populus alba, Linnaeus, var. typica, Gombocz, in Math. Termes. Kb'zl. xxx. 151 (1911). Populus megaleuce and alba, Dode, op. cit. 24, 25 (1905). Populus hybrida, Reichenbach, Icon. Fl. Germ. xi. 29, t. 615 (1849) (not Bieberstein 4). Populus Steiniana, Bornmüller, in Gartenfl. xxxvii. 173, figs. 37, 38 (1888). Populus Bachofenii, Reichenbach, Icon. Fl. Germ. xi. 29, t. 616 (1849) (not VVierzbicki5). 1 Murray, New Eng. Dût. i. 15 (1888), says that the Dutch name corresponds to an old French word, ate/, derived from late Latin, albellus, a name given to the white poplar in the twelfth century. Worlidge, Syst. Agric. 96 (1681), speaks of the abele tree as " a finer kind of white poplar." 3 Bull. Arbor. Belg. 1882, p. 190. 3 Ibid. 1902, p. 72. 4 Bieberstein, Fl. Taur. Cauc. ii. 422 (1808), describes P. hybrida, a doubtful plant which, op. cit. iii. 633 (1819), he abandons, as not being different from P. alba. A specimen, however, in the Cambridge Herbarium, labelled " P. hybrida, M. B., Caucasus," is quite distinct from P. alba, and appears to be a form of P. canescens with orbicular leaves, sinuately toothed, and glabrous beneath. It bears fruiting catkins 4 in. long. 6 P. Bachofenii, Wierzbicki, in Rochel, Banat. Reise, 77 (1838), is said by Gombocz, op. cit. 148, to be a form of P. alba, var. nivea, and identical with P. heteroloba, Dode, authentic specimens of which I cannot distinguish from typical P. alba, var. nivea. Populus 1781 A large tree, attaining 100 ft. or more in height and 15 ft. in girth. Bark on young stems thin, smooth, grey or whitish ; on older stems breaking on the surface into small roughened dark-coloured rhombic cavities, which ultimately unite together, making the bark towards the base of old trunks deep and longitudinally furrowed. Young branchlets, towards the top of the long shoots, covered with a dense whitish tomentum, which towards their base and on the short shoots diminishes in quantity, the twigs becoming dull grey or shining and glabrescent. Buds ovoid, more or less tomentose, according to their position on the branchlet. Leaves (Plate 408, Fig. 3) of two kinds—on long shoots and on suckers ovate-deltoid, cordate, acute, dark shining green and slightly tomentose above, covered beneath with a thick greyish tomentum ; margin ciliate, with a few triangular teeth, variable in size, and like the rest of the margin irregularly glandular-serrate ; petiole rounded, tomentose. Leaves on the short shoots, suborbicular or broadly ovate, subcordate, obtuse ; margin non-ciliate, with a narrow translucent border, and a few sinuate non-serrate teeth ; dark green, shining and glabrescent above ; lower surface light green, with traces of scattered grey tomentum ; petiole laterally compressed, glabrescent or slightly tomentose. Staininate catkins,1 2\ to 4 in. long ; axis tomentose ; scales obovate, toothed, yellowish brown, fringed with hairs at the apex ; pedicels pilose ; disc oblique, entire in margin, glabrous ; stamens eight to fifteen. Pistillate catkins about i in. long ; with similar scales ; disc pubescent ; ovary glabrous, with four spreading sub-sessile yellowish simple stigmatic lobes.2 Fruiting catkins, 3 to 4 in. long ; capsules glabrous, two-valved. The stigmatic lobes like those of P. alba, are simple and undivided ; whereas in P. tremula, each lobe is deeply scolloped and waved. Populus Bogueana, Dode, op. cit. 24 (1905), is a vigorous form of P. canescens, in which the leaves on the long shoots are very large, 5 in. or more in length and breadth ; and appears to be now sold by some nurserymen as P. tomentosa, the white poplar from Peking. A tree of this at Kew, obtained from Simon-Louis in 1904, is now about 25 ft. high. Another at Grayswood, obtained from Barbier in 1906, is 15 ft. high and very thriving. There are also small trees with similar large foliage in the Botanic Gardens at Edinburgh and Glasnevin. I found this form wild in the forest of Orleans, where it was growing evidently from vigorous suckers of the typical grey poplar beside it. By P. canescens? Smith meant the common grey poplar in England, which I have described above. It differs mainly from typical P. alba in the lesser amount of 1 A tree in the Cambridge Botanic Garden bore monoecious catkins in 1907, the staminate flowers at the base, the pistillate towards the apex. Similar cases are recorded by Penzig, Pß. Tératologie, ii. 321 (1894), and by Bâillon, in Bull. Soc. Linn. Paris, 1897, p. 658. 2 The eight stigmas, figured by Smith, Eng. Bot. xxiii. t. 1619 (1806), have not been observed by any other botanist, and were probably abnormal. In Eng. Bot. vii. 114, t. 1392 (1840), the number of stigmas is said to be inconstant, but I have found them invariably four. 3 The grey and white poplars were first distinguished by Lobelius, Icon. PI. seu Stirp. ii. 193 (1591). Their synonymy is much involved, but has been clearly elucidated in an interesting article by Tidestrom, in the American Midland Naturalist, i. 113 (1909), to which the reader interested in such matters is referred. I am not at all certain that Smith clearly understood the distinctions between the two species, as Sowerby's plate of P. alba, t. 1618 (1806), is the leaf of a vigorous shoot of P. canescens ; and Smith's statement that P. alba is not uncommon in moist woods (in England) is erroneous, as the latter is never seen in England, except as a planted tree. Most of his description refers to P. alba ; and his specimen in the herbarium of the Linnean Society, London, certainly belongs to this species. VII S 1782 The Trees of Great Britain and Ireland the tomentum on the branchlets and leaves, the latter having a different shape. It is adapted to the humid and mild climate of western Europe ; while P. alba, with its dense protective covering against evaporation of water, is a native of drier and more continental regions. In the grey poplar, as in the white poplar, the leaves are fundamentally of two kinds, different in shape, margin, tomentum, and petiole; but conveniently dis tinguished as "white leaves" and "green leaves." Some shoots bear only white leaves, some only green leaves ; whilst others have green leaves towards the base and white leaves towards the tip. The green leaves may be greyish beneath or nearly devoid of tomentum. The size of the white leaves -is inconstant, and depends on the vigour of the branchlets. In old trees the green leaves are preponderant, especially on the lower branches ; and such trees are popularly known as P. canescens. In younger trees, and on the upper branches of older trees, the white leaves are conspicuous ; and such trees are often erroneously called P. alba. In addition to this variation, occurring on the same tree in its different stages of growth, there are differences in the foliage of different trees, which require further study. A peculiar form, in which the leaves are thinner in texture, orbicular and not deltoid in outline, and nearly glabrous beneath, is considered by Dr. C. E. Moss to be possibly the hybrid P. canescens x P. tremula. Two pistillate trees1 with this foliage, one at Hitchin, and another growing beside a stream between Caverham and Icklingham, Suffolk, have pink stigmas, the colour of which suggests the influence of P. tremula. These two trees flower a fortnight earlier than ordinary P. canescens. Staminate trees with similar foliage have been found both in England and Ireland. Mr. R. A. Phillips, who sends specimens2 from Lorrha, Tipperary, points out that the catkins are shorter than in the typical grey poplar, about i^ in. long, with fewer (4 to 7) stamens, and narrower scales. There is also a difference in habit, the round-leaved trees have thick stiff ascending branches and short erect twigs ; while the ordinary form has spreading branches and pendulous twigs. In P. tremula* the foliage is all of one kind, similar in form to the green leaves of P. canescens, but devoid of any trace of tomentum. The theory that typical P. canescens is a hybrid4 between P. alba and P. tremula rests on this apparent similarity of foliage, and is not supported by evidence from the floral organs. Female trees of the grey poplar are, however, rare, and the pistillate flowers have only been examined in a few cases. So far as I can judge, P. canescens is a good species, being the 1 The pistillate tree with pink stigmas is P. canescens, Dode, in Mém. Soc. Hist. Nat. Autun. xviii. 26 (1908) (not Smith). 2 Mr. Phillips recently sent me another form with leaves intermediate in shape and with flowers having eight stamens. 3 The suckers of P. tremula bear leaves which, in their shape and in the presence of a slight tomentum, simulate the white leaves of P. canescens. 4 The older continental botanists, as Reiclienbach, Icon. Fl. Germ. xi. 30, t. 617 (1849), and Ilartig, Naturgesch. Forst. Culturpfl. 434 (1851), considered P. canescens, Smith, to be identical with the pubescent form of the aspen, P. tremula, var. villosa. This has silky hairs on the leaves and branchlets, very different from the tomentum of P. canescens. The occurrence of true hybrids between P. Iremula and P. alba is possible on the Continent. Radde, Pßangcnverb. Kaukas. 153 (1899), saw a group of trees near Sotschi on the eastern shore of the Black Sea, which he considered to be of this origin, as the trees strongly resembled P. tremula. Rechinger, in Verh. Zool. Bot. Ges. Wien, xlix. 284 (1899), discusses the hybrids in Austria. Adamovic, Veg. Balkanländ. 145 (1909), reports similar hybrids in eastern Roiimelia. Cf. also P. albo-tremttla, Krause, in Jahrb. Schles. Ces. 1848, p. 130. Populus 1783 western representative in Europe of the white poplar. It appears to be common in France, extending eastward to the Rhine valley ; and is known in Belgium and France as franc-picard or grisard. It often attains an immense size in France, the finest that I have seen being one in the Botanic Garden at Toulouse, which was 100 ft. in height and 12 ft. in girth in 1912. In Holland, where it is wild on the dunes near Haarlem,1 it is always called abeel, a name which is often erroneously applied to the white poplar by English writers. The grey poplar is undoubtedly a native2 of England, and is the tree referred to by old authors as the white poplar ; but has been probably always better known to woodmen and peasants as the aspe. (A. H.) Both the grey and white poplars were known to Evelyn,3 who spoke of the first as the white "to be raised in abundance by every set or slip. Fence the ground as far as any old poplar roots extend and they will furnish you with suckers innumerable, to be slipped from their mothers and transplanted the very first year, but if you cut down an old tree you shall need no other nursery." Later on, he says : " There is something a finer sort of white poplar, which the Dutch call abele, and we have late much of it transported out of Holland. They are also best propagated of slips from the roots, the least of which will take, and may in March, at three or four years' growth, be transplanted." The latter was still an uncommon tree in Plot's time, for he says :4 "Of unusual trees now cultivated in Oxfordshire is the abele tree, advantageously propagated by Sir J. Croke of Waterstock, by cutting stakes out of the more substantial parts of the wood ; which put into moist ground grew more freely than willows, coming in three or four years' time to an incredible height." French and English authors agree that the white poplar will not bear lopping like the black poplars ; and though I have no experience in this matter, yet as I have never seen a pollarded tree I presume that the grey poplar is equally liable to injury when large branches are cut. It seems able to attain its largest dimensions in poor stiff soil, and in cold situations, provided that there is sufficient moisture in summer ; and though I cannot say that the tree is equal to the black Italian poplar from an economic point of view, or equal to P. alba as an ornamental tree, yet as the illustra tion shows, it is a stately tree when well grown. The range of this tree in Great Britain is obscure, because it has been planted for a long period.6 Watson6 says, "It is given as an unquestioned native in the floras of Surrey, Essex, Herts, Suffolk, Norfolk, Cambridge. In the floras of Tyne and North Yorkshire it is reduced to the grade of denizen. And if I rightly know the distinction between alba and canescens, I should now deem the latter native, and the former planted, as seen in Surrey and elsewhere." Though the male tree is found of large size in many counties, the female is in 1 Mr. L. Springer, a distinguished landscape gardener of Haarlem, says that the true white poplar will not endure the sea wind on the coast of Holland. 2 Clement Reid, Origin of Brit. Flora, 150 (1899), states that leaves collected by Prestwich in interglacia] beds at Greys, Essex, suggest P. canescens, though they may belong to P. tremula. The latter species is only recorded from one locality, Caerwys, Flintshire, in neolithic deposits ; so that the geological evidence as to the existence of poplars in England is very scanty. 3 Sylva, 78 (1679). 4 Natural History of Oxfordshire, 175 (1705). o Loudon quotes M'Culloch, that it is the only tree found in the island of Lewis, but probably he mistook the aspen Tor it. e Topographical Botany, i. p. 357 (1873). 1784 The Trees of Great Britain and Ireland my experience very rare, the only ones I have seen being two rather stunted specimens of considerable age, which grow on my land by the side of a watercourse in a meadow on Upcot Farm, near Withington, Gloucestershire. There are male specimens close to them, and in some seasons fertile seed is produced, from which, in 1907, a few seedlings1 were raised by Miss F. Woolward. Two of these were planted at Colesborne, but one was weakly from the first and the other was killed by a fungoid disease soon after I planted it out. I have never seen a natural seedling of this or of the white poplar in England. REMARKABLE TREES The finest grey poplars that we have seen are in a group called the Grove in the park at Longleat. Mitchell,2 writing in 1827, mentions these as being then loo ft. high, with trunks 3 to 4 ft. in diameter, and 40 to 60 ft. of clear bole. In Trans. Eng. Arbor. Soc. v. 391 (1903), an illustration of one of these trees from a photograph by A. C. Forbes was given. He calls it an abele poplar, and gives the measurement of the largest as 120 ft. by 15 ft. 3 in., and the cubic contents 450 ft. Ten trees in this grove, all with straight clean boles of 50 to 80 ft., then averaged 115 ft. by 13 ft., with an average cubic content of 240 ft. When I last visited Long leat in 1909 I found the majority of them still healthy, and measured one which was 125 ft. by n ft. There are several very fine trees at Colesborne, one of which (Plate 382) has two stems rising from the base, and measuring 115 ft. by 10 ft. and 10 ft. 3 in. respectively. These trees sucker freely, and some young trees grown from the suckers, which have been pruned but not transplanted, are growing very fast —about 40 ft. in ten years. In a field about a mile from Overbury Court, near Tewkesbury, Mr. F. R. S. Balfour discovered a tree 86 ft. high, with a trunk 18 ft. long by 18 ft. 10 in. in girth and with very large spreading branches. At Kingston Lacy, Dorsetshire, a tree with many suckers round it measured, in 1906, 118 ft. by 13 ft. 9 in. At Strathfieldsaye I measured in 1905 a tree 108 ft. by 16 ft., which had a sucker coming up 80 yards from its base. In the park at Syon House there is a well-shaped tree 101 ft. by 12 ft. 5 in.8 In the home park of Windsor Castle there is a short avenue of grey poplars close to the bank of the Thames, below Victoria Bridge. Mr. Mackellar informs us that these were planted between 1840 and 1850, and measured in January, 1913, 90 to zoo ft. in height and 9 to 12 ft. in girth. At Gilbert White's old home at Sel- borne, the tallest tree is a grey poplar, which Mr. H. B. Watt* measured in 1912 as 109 ft. high by 9 ft. 5 in. in girth. At Youngsbury, Ware, Hertford, a tree measured 95 ft. by ID ft. 4 in. in 1911. In Wales I have seen none of any great size ; and I have not heard of any remarkable ones in the eastern or northern counties. In Scotland, where the tree is common, Mr. J. Renwick8 tells us of two very 1 fount. Bot. xlv. 417, t. 487 (1907). 2 Dendrologia, 51 (1827). 3 A. B. Jackson, Cat. Trees Syon, 22 (1910). 4 Seltorne Magazine, xxiii. 122 (1912). 6 Glasgow Naturalist, iii. 119, pi. ii. (1911). Populus large trees at Mauldslie Castle, Lanarkshire, the seat of Lord Newlands, which in 1911 measured too ft. by 21 ft. 3 in. at two feet from the ground, and 117 ft. by 16 ft. 5 in. at two feet nine inches from the ground. Both these trees suffered severely in the great gale of October 5, 1911, the larger tree losing one of its heavy limbs. Walker, Essays of Natural History, 49 (1812), states : " In the year 1769 there was a row of abeles at Stevenston, East Lothian, which was soon after cut down. It contained 122 trees, all about 80 ft. high and from 20 to 30 ft. of clear trunk, without a branch. The trunks were 5 to 7 ft. in circumference, and yet they stood only 7 ft. distant from each other. They were planted in a deep moist soil, were then eighty years old, and afforded a great quantity of timber. It is doubted whether or not the abele is a native of England. It certainly has the appearance of being an indigenous tree in several parts of Scotland. It was planted in many places about the end of the last and in the beginning of the present century, but it has since been neglected." The tallest recorded in the Old and Remarkable Trees of Scotland-was at Glenarbuck in Dumbartonshire, and was said to be no ft. by 12 ft. in 1867. In Ireland the grey poplar is found on the banks of the Suir, Nore, Barrow, and other rivers in the south, where it may very possibly be wild ; and Mr. R. A. Phillips has sent us specimens of several trees, one of which, growing near Birr, was 90 ft. by 12 ft. in 1910. I measured a tree at Markree Castle, Sligo, which in 1909 was 120 ft. by 13 ft. There are two fine trees at Abbeyleix, one of which was zoo ft. by 13 ft. 4 in. in 1910. In Belgium and France this tree is known as grisard, and is often mistaken for the white poplar. I have not noticed any of remarkable size.1 TIMBER I am inclined to think that the timber is of superior quality to that of the black Italian poplar ; and from what I heard in Belgium and France,2 that opinion is also held there. But this may depend on the age of the trees and the rate of growth, which, according to Crowe,3 who paid much attention to this tree in Norfolk, is slower than that of any other British poplar. It is rarely distinguished by the timber merchants in England. Messrs. Howes and Sons, Norwich, however, inform me that "it is the only poplar found to be of any service in the coach-building trade. It is light in weight, exceedingly strong, and can be cleaned up into a nice finish to receive the paint." They pay for the best quality of the wood of this tree, which they know under the name of abele, about 25. 6d. per foot, as compared with about is. 3d. for the wood of the black Italian poplar, which is much less durable and not so firm in texture. I prefer to use it at home for cart bottoms and linings, and for flooring cottage bedrooms, for which its toughness and non-inflammable nature make it valuable. (H. J. E.) 1 In Mitt. D. Dendr. Ges. 1904, p. 18, a tree was recorded at Schloss Dyck, near Dusseldorf, which, at about 95 years old, was 48 metres high by 3^ metres in girth. The height is evidently much exaggerated. 2 Mouillefert, Essences Forestières, 301 (1903), says that its fine-grained wood is one of the best of the poplars. 3 Smith, Eng. Flora, iv. 244 (1828), who obtained his information from Crowe, states that "the wood is much finer than that of any other British poplar, making as good floors as the best Norway fir in appearance, and having moreover the valuable property, that it will not, like any resinous wood, readily take fire." Cf. Gard. Chron. 1848, p. 172. 1786 The Trees of Great Britain and Ireland POPULUS TOMENTOSA Populus lomentosa, Carrière, in Rev. Hort. x. 340(1867); Wesmael, in De Candolle, Prod, x vi. 2, p. 3 2 5 (1868), and Mem. Soc. Sc. Hainaut, 228, t. 17 (1869); Schneider, Laubholzkunde, i. 21 (1904); Dode, in Mém. Soc. Hist. Nat. Autun, xviii. 25 (1905); Gombocz, in Math. Term. Kozl. 140 (1911). Populus alba, Burkill, in Journ, Linn. Soc. (Bot.) xxvi. 535 (1899) (not Linnseus). Populus alba, Linnseus, var. tomentosa, Wesmael, in Bull. Soc. Roy. Bot. Bclg. xxvi. 373 (1887); Burkill, m Journ. Linn. Soc. (Bot.) xxvi. 535 (1899). Populus alba denudata, Maximowicz, in Bull. Soc. Nat. Mosc. i. 49 (1879) (not Hartig). Populuspekinensis, L. Henry, in Rev. Hort. Ixxv. 355, fig. 142 (1903). A large tree, similar in size and bark to P. alba. Branchlets grey tomentose. Buds ovoid, slightly tomentose, chestnut brown. Leaves (Plate 408, Fig. 2) on the long shoots of old trees, 4 to 6 in. long, 3 to 5 in. broad, triangular-ovate, without lobes or lobules, subcordate or truncate at the broad base, acuminate at the apex ; margin with a few (not exceeding ten on each side) sinuate teeth ; dark shining green above, glabrescent beneath with traces of grey tomentum. Leaves on the long shoots of vigorous young trees, similar in shape, but the margin biserrate with acute glandular teeth, and the lower surface covered with a grey tomentum. Leaves on the short shoots, small, ovate or triangular, cuneate at the base, sinuately toothed, glabrous beneath. Flowers not seen. This fine poplar, which attains an enormous size in north China,1 was discovered by Simon at Siwan, north-west of Peking. His specimen, described by Carrière, is identical with those in the Kew Herbarium, collected near Peking by Sir Rutherford Alcock and by Prof. Sargent, and with another gathered in Shantung by the Rev. A. Williamson, which is preserved in the Edinburgh Herbarium. Gombocz records it also for the mountains of Shensi, and Kiaochow in Shantung. Elwes, in 1912, saw old trees in the grounds of the Summer Palace, Peking, which were about 75 ft. high and ID ft. in girth. This poplar is calledpai-yang by the Chinese.2 A young living plant sent from Peking in 1897 by Père Provost to the Museum at Paris, where it was propagated, has grown vigorously, and was in 1912 about 35 ft. high by 2 ft. 3 in. in girth. We have seen no trees in England of this species.3 Jack4 introduced in 1905 into the Arnold Arboretum cuttings from Peking, which have produced thriving and hardy young trees. He states, however, that it is more readily propagated by grafting. (A. H.) 1 According to Schneider, it was collected also by Père Giraldi, farther south in Shensi ; but the latter's specimen is identified by Diels, Flora von Central China, 274 (1901), with P. tremula, and is possibly P. ivutaica, Mayr, Frcmdland. Wald- u. Parkbäume, 494, fig. 215 (1906). 2 Cf. Bretschneider, Bot. Sinic, ii. 359 (1892), who refers to this tree as P. alba. 3 The trees often sold by French nurserymen as P. tomentosa appear to be P. Bogueaaa, Dode. Cf. aale, p. 1781. 4 Mitt. Deat. Demi. Ces. 1909, p. 281. Populus 1787 POPULUS TREMULA, ASPEN Popuhts tremula, Linnaeus, Sp. PL 1043 (1753); Loudon, Art. et Frut. Brit. iii. 1645 (1838); Wesmael, in Mém. Soc. Sc. Hainaut, iii. 229 (1869), and De Candolle, Prod. xvi. 2, p. 325 (1864); Willkomm, Forstl. Flora, 521 (1887); Mathieu, Flore Forestière, 486 (1897); Schneider, Laubholzkunde, i. 19 (1904); Dode, in Mém. Soc. Hist. Nat. Autun, xviii. 30 (1905); Ascherson and Graebner, Syn. Mitteleurop. Flora, iv. 24 (1908); Gombocz, in Math. Termes. Közl. xxx. 123 (1911). Populus australis, Tenore, Ind. Sem. Hort. Neap. 1830, p. 15. Populus grceca, Grisebach, Spie. Fl. Rum. ii. 345 (1844) (not Aiton, Hort. Kew.\ A tree, occasionally attaining in Scandinavia, Russia, France, and Germany 100 ft. in height and 6 to 8 ft. in girth, but usually much smaller, especially in the British Isles. Bark of young trees smooth, thin, greenish or whitish ; on old trunks thick, with small rhomboidal fissures, as in P. alba, and ultimately deeply furrowed. Young branchlets glabrous, rounded, shining, with orange lenticels. Buds ovoid, acute, shining brown, slightly viscid, with ciliate scales, the uppermost of which are slightly pubescent. Leaves (Plate 408, Fig. 4) suborbicular, variable in size, averaging 2 in. in diameter, thin in texture, truncate or subcordate at the base, rounded or acute at the apex ; margin with a narrow translucent border, and a few rounded or sinuate small teeth ; tomentose when young, speedily becoming glabrous on both surfaces, pale or glaucous beneath ; venation pseudo-five-palmate ; glands1 at the base two, cup-shaped, well-developed on the terminal leaves of long vigorous shoots, absent on the basal leaves and on those of the short shoots ; petioles slender, glabrous, laterally compressed, often as long as the blade. Leaves on young plants and on sucker shoots, and in rare cases on sporadic branches of adult trees, different in shape and much larger, 4 to 5 in. long, 3 to 4 in. broad ; ovate, acumi nate at the apex, truncate or cordate at the base, greyish and slightly woolly beneath, glandular-serrate, with short pubescent terete petioles. Catkins sub-sessile, densely and greyish tomentose ; axis pubescent ; scales long persistent, obovate, deeply lobed and fringed with long white hairs ; flowers dense, numerous, on very short pilose pedicels. Stamens about 10, with short filaments and purple anthers, on an oblique disc with an entire and incurved margin. Ovary glabrous ; stigmas two, reddish, each divided, forming four widely dilated curving arms ; disc funnel-shaped, oblique, glabrous ; capsule two-valved. This species in the wild state displays a considerable amount of variation in the shape, size, and colour of the leaf, and in the amount of pubescence on the branchlets and leaves. The most noteworthy2 are :— i. Var. Freyni, Hervier, in Bull. Herb. Boiss. iv. app. i. 18 (1896), and Rev. Gen. Bot. viii. 177 (1896). Leaves rhombic, cuneate at the base, ciliate, pubescent beneath when young. Central France and Prussia. 1 According to Kemer, Nat. Hist. Plants, Eng. Trans, i. 238, fig. 55 (1898), these glands exude resin and serve for absorbing water in rainy weather; but Trelease, in Bot. Cas. vi. 284 (1881), states that they contain honey at the beginning of the season, and are visited by bees and other insects. 2 Var. purpurea, Simon-Louis, ex Späth, Cat. No. IO2, p. 108 (1898-1899), with purplish young leaves, does not seem to differ from the type, as seen in cultivation at Kew. 1788 The Trees of Great Britain and Ireland 2. Var. villosa, Wesmael, in De Candolle, Prod. xvi. 2, p. 325 (1868). Populus villosa, Lange, Syll. Soc. Ratisb. i. 185 (1824), and ex Reichenbach, Fl. Germ. Excur. 173 (1830). Fopulus canescens, Reichenbach, Icon. Fl. Germ. xi. 30 t. 617 (1849) (not Smith) ; Hartig, Forstl. Kulturpfl. 434 (1851). Branchlets and leaves at first densely pubescent with long silky hairs, more or less persistent in summer. This variety1 appears to be quite as common in the wild state as the typical glabrous form. The origin of the following cultivated form is unknown :— 3. Var. pendula, Loudon, Arb. et Frut. Brit. iii. 1646 (1838). Populuspendula, Burgsdorf, Anl. Anpfl. Holzart, ii. 175 (1787). Branchlets pendulous. Usually seen as a small grafted tree. The common aspen is widely spread through Europe, northern Africa, Asia Minor, the Caucasus, and Siberia, being replaced by closely allied species in China, Japan, and the Himalayas. It occurs in every European country; but is absent from the south of Spain, Sicily, and the islands in the western Mediterranean. It is much more common in the north, where it reaches the Arctic Circle both in Europe and Asia ; and either in pure woods or mixed with birch covers extensive tracts in Scandinavia, Russia, and Siberia. Towards the south, it only occurs as a scattered tree in mixed woods ; and ascends in the Pyrenees to 6000 ft., and in the Alps to 4000 ft. In England it is not uncommon in coppiced woods; but it is of more frequent occurrence in the Highlands of Scotland, ascending to tree limit. The aspen is a short-lived tree, rarely attaining an age of over zoo years. When cut down, it produces coppice shoots of no great vigour, and usually reproduces itself in such cases by abundant suckers,2 which are given off to a considerable distance by its widely-spreading superficial roots. (A. H.) As an ornamental tree the aspen is in northern countries one of the most beautiful, on account of the splendid red and yellow tints which the leaves assume in autumn ; but in England these colours are seldom seen in the same degree, and though the bright pale green of its trembling leaves gives it a certain beauty, it is hardly worth growing in any quantity. It is not particular about soil, and may serve to clothe waste places such as old pit banks. It is one of the latest trees to come into leaf in spring. The largest trees I have noticed in England are two at Little Sodbury Manor, in Gloucestershire, which do not much exceed 60 ft. in height. Sir Hugh Beevor tells me that he has seen none larger in the eastern counties. There are four trees forming a handsome group near the river Gade at The Grove, Watford, the largest being 54 feet by 3 ft. 3 in. in 1904. These were photographed3 by Mr. Henry Irving. In Wales the tree does not seem common, but apparently attains a greater size than it usually does in England. In May 1911 I saw some large trees at an l M. Dode informed me that in the Forêt de Bondy this varieiy attains 35 metres in height.—H. J. E. 2 The suckers of P. tremula are described by Dubard, in Ann. Sc. Nat. xvii. 160 (1903). 3 Reproduced by Groom, Trees and their Life Histories, figs. 192, 193 (1907). Populus 1789 elevation of about 800 ft. at Abergwessin in North Breconshire. The Welsh name for it in this district is aethnen ; but at Hafodunos, in North Wales, W. Jones, the head gardener, told me that the local name was tafoden merched, which means "women's tongues." In the Isle of Man a name of the same meaning, ckengey-ny- mraane, was used. In the Highlands, where it ascends to the upper limit of trees—in Braemar up to at least 1600 ft.1—it seems commoner, and attains larger dimensions. I have never seen any, however, which in size equal those in the north of Norway, the largest I know of being a tree on the shores of Loch Carry, which Captain Ellice of Invergarry found in 1910 to be about 60 ft. by 7 ft.2 In the birch woods of Strath- glass, Glenaffric, and Guisachan the aspen occurs in clumps which appear to have grown from suckers, but old trees are scarce. The largest I saw was a fallen tree above the falls in Glenaffric, which, when standing, was about 50 ft. high, with a trunk 7 ft. in girth. The belief3 of the Highlanders, who call the aspen crithean or critheac, that the Cross of Christ was made of this tree still exists both among Catholics and Protestants in this district. I was assured by reliable persons that it is looked on as an accursed tree, and that no Highlander will use the wood for any purpose, even for fuel.4 Notwithstanding this belief, I saw on the croft of Peter Macdonald at Balnaith, near the head of Glen Urquhart, a group of well-shaped aspen about 40 ft. high, which, as he told me, he had trained up from suckers, and were about forty years old. In Ireland, the aspen is recorded from almost every county, but is by no means a common tree. Mr. R. A. Phillips informs us that it is native on mountain cliffs, rocky lake shores, the banks of rivers, and in old hedgerows in uncultivated bogland districts, and also on the islands off the west coast. In the mountains of Kerry and Antrim it is a mere bush, but in the lowlands is a small tree, rarely over 40 ft. in height. The finest which he has seen grows on the edge of the river Nore near Durrow, and measured 65 ft. by 5 ft. in 1908. According to Schiibeler it extends in Scandinavia as far north as Alten (lat. 70°), where it attains 60 ft. in height, and in the south ascends to 3500 ft. The tallest aspens that he mentions in Norway grew some miles east of the farm of Viken, in Niderven, and were 90 to 100 ft. high ; whilst at Femrade, in Sogndal (lat. 61°), there was a very old aspen 58 ft. high, with a girth of 16 ft. at four feet from the ground. I saw myself in Junkersdal, in lat. 67°, trees of 80 feet high and 6 to 7 ft. in girth, which were finer than any I know in Great Britain. In France, Mouillefert says, that though common in the damper parts of the forest on sandy or gravelly soils, it is rare on dry or calcareous formations, and that owing to the freedom with which it produces suckers it tends to supersede other trees in suitable places. 1 White, Flora of Perthshire, p. 268, says that it ascends to 2100 ft. in Athole, and 1400 ft. in Rannoch ; and tliat both the forms—villosa, Lange, and glatra, Syme,—occur in the county. 2 Loudon mentions trees of much greater size at various places in England and Scotland, but there is little doubt that they were P. canescens. 3 Cf. Loudon, op. cit. 1648, and Cameron, Gaelic Names of Plants, 70 (1883). 4 Cf. Carmichael, Carmina Gadelica, 104 (1900), who states that in Uist the hateful aspen is banned. If it still exists in Uist, it is now an uncommon tree, as I saw none in North or South Uist in 1910. VII . T 179° The Trees of Great Britain and Ireland TIMBER Though usually looked on as a forest weed, it is possible that when grown naturally from suckers thickly enough to clean the stems from branches, it may have some value for making matches. In Sweden it is largely employed for this purpose, and according to Schiibeler 1,400,000 cubic ft. were thus used in 1882, and valued at 70 ore (about 8d.) per cubic foot. Cargoes of poplar timber, which I believe to be mainly aspen, are now imported from the Baltic for making matches, and cost delivered in Gloucester from 385. to 485. per load of 50 cubic ft. caliper measure. It is also largely used for pulp-making, for which it is very suitable ; but it could scarcely be produced here in sufficient quantities at a price that would compete with the produce of northern Europe and America. On this subject, however, a valuable paper1 by Weigle and Frothingham on the American aspens should be consulted, as these species resemble our native aspen in life-history, characteristics, and uses. The conclusions of these authors may be summarised as follows : 1. No other trees have so wide a distribution in Europe and Asia as the aspens, P. tremula covering 140° of longitude and 35° of latitude, whilst P. trenmloides ranges over 112° and 41° respectively. 2. They are both pre-eminently cold- and moisture-loving trees, requiring a very short season of growth and thriving—as P. trem^^la does at Colesborne—where frosts may occur during every month of the year. 3. For their best growth they require deep fresh or moist, porous and well- drained soils ; but they will grow on thin dry soil and in poorly drained situations. 4. They are strikingly intolerant of shade ; and this applies even more strongly to the suckers by which they are commonly reproduced, and which are often mistaken for seedlings. The latter are rarely seen in England. 5. Their growth is rapid during the first twenty to thirty years, and though they may attain considerable size, yet they are short-lived, usually decaying before 100 years of age, and often much sooner ; and root-suckers do not produce such large or well-shaped trees as seedlings. 6. They are best managed as a pure crop under a short rotation ; and on account of their extreme light-demanding character require timely thinning. 7. The wood produces the best and whitest pulp, which can be produced and manufactured, more cheaply than other species, into paper which is peculiarly suitable for books and magazines. As the fibre is too short to make good paper alone, it is mixed with a proportion (usually about 40 per cent) of sulphite spruce pulp which adds strength. The finished paper is tough, white, and easily sized, and though inferior to rag paper for the finest uses, is much cheaper. Waste land suitable for profitable planting of P. tremula (and possibly also P. canescens) might be found in some parts of Scotland and Ireland ; and experiments in this direction are advisable. In France, according to Mouillefert, the wood is valued for charcoal, but as firewood it burns out very quickly. (H. J. E.) 1 U.S. Forest Service, Bull. No. 93, The Aspens: their Gimvth and Management (1911). Populus 1791 POPULUS TREMULOIDES, AMERICAN ASPEN Pofulus tremuloides, Michaux, FI. Bor. Amer. ii. 243 (1803) ; Sargent, Silva N. Amer. ix. 158, t. 487 (1896), and Trees N. Amer. 154 (1905); Schneider, Laubhohkunde, i. 19 (1904); Dode, in Mém. Soc. Hist. Nat. Autun, xviii. 33 (1905). Populus trépida, Willdenow, Sp. PL ii. 803 (1805); Loudon, Arb. et Frut. Brit. iii. 1649 (1838). Populus tremuliformis, Emerson, Trees Massachusetts, 243 (1846). Populus atheniensis,1 Ludwig, Neue Wilde Baumz. 35 (1783), ex Koch, Dendrologie, ii. 486 (1872); Koehne, Deut. Dend. 80 (1893). Populus grata,1 Loudon, Arb. et Frut. Brit. iii. 1651 (1838) (not Aiton) ; Lauche, Deut. Dend. (1883). A tree, attaining in America 100 ft. in height and 9 ft. in girth. Bark like that of P. tremula. Young branchlets slender, glabrous, shining reddish brown, with orange lenticels. Buds ovoid, sharp-pointed, shining brownish, slightly viscid, with glabrous scales. Leaves (Plate 408, Fig. 5) ovate to nearly orbicular, i^ to 2 in. in diameter, thin in texture ; truncate, rounded or cuneate at the base ; shortly cuspidate at the apex ; glabrous on both surfaces, pale beneath ; margin with a narrow translucent border, ciliate especially on the leaves of the short shoots, finely glandular- serrate ; pseudo-three- to five-palminerved at the base, where the glands in occurrence and appearance are like those of P. trennda ; petiole slender, glabrous, laterally compressed, variable in length, often as long as the blade. Leaves on sucker shoots, similar to those of P. tremula, but glabrous on both surfaces, with ciliated margins. Flowers scarcely distinguishable from those of P. tremula, but with more slender catkins, smaller in all the parts of the flowers ; disc of the pistillate flower crenate. 1. Var. pendula. A weeping form, with pendulous branches, always grafted. This is generally known as the Parasol de St. Jidien, which is said 2 to have been first propagated by Messrs. Baltet, who found in 1865, on the bank of a canal at St. Julien, near Troyes, a tree with weeping branches, which they crown-grafted on the white poplar, and considered to be a weeping form of P. trenmla. It is said by Koch3 to have been much more common in commerce in 1872 than the weeping variety of the common aspen, and it is possible that the preceding history is applicable rather to a weeping variety of P. tremula. 2. P. cercidiphylla, Britton, N. Amer. Trees, 180 (1908), seems to be a form with small entire or undulate leaves, which was found in Wyoming by Dr. C. C. Curtis in 1900. (A. H.) 1 Loudon describes this poplar under both the names F. trépida and P. graca, and states in Gard. A fag. 1840, p. 231, and Trees and Shrubs, 823 (1842), that P. graca was " named after the village called Athens, on the banks of the Mississippi, where the tree grows abundantly." P. atheniensis is said by Koehne to derive its name from the town of Athens in New York State, whence it was introduced. P. graca, Aiton, Hort. Keiu. iii. 407 (1789), was insufficiently described, and said to be a native of the Greek archipelago. 1 Cayeux, in The Garden, 1886, p. 2. The pendulous variety of P. trenmloidts is called Parasol de St. Julien, in Simon-Louis's catalogue, 1899-190», and appears to be now always sold in France under this name. Späth, Catalogue, No. 57, p. 61 (1883), identifies the Parasol de St. Julien with P. canescens pendula; and the latter name, now no longer employed, would seem to show that Späth's tree was rather a weeping P. tremula than a pendulous form of P. tremuloides. 3 Dendrologie, ii. pt. i. 487 (1872). 1792 The Trees of Great Britain and Ireland The American aspen is similar in its habits to the common species, and is widely spread throughout North America, from southern Labrador, the eastern shores of Hudson's Bay, the mouth of the Mackenzie river, and the Yukon valley in Alaska, southwards to Pennsylvania and Nebraska, and through all the mountain regions of the west to central California, the San Pedro mountain in Lower California, northern Arizona, New Mexico, and the state of Chihuahua in Mexico. It is common in the east on moist sandy soil, and often borders the western prairies with a wide belt ; and to the northward is often mixed with spruce and birch. In the mountain regions of the western and Pacific states, it ascends to 10,000 ft. above the sea. Sargent says that on account of its remarkable power of germinating on burnt soil, and rapidly covering mountain sides which have been devastated by fire, it has had a greater influence than any other tree on the composition and distribution of the subalpine and boreal forests of North America. Macoun says that in the North west, Athabasca, and Mackenzie districts it is everywhere common on dry soil, but not on alluvial flats, and that it reproduces freely after a forest fire by root-suckers, but not from seed. It is said to have been introduced1 in 1779 by Hugh, Duke of Northumberland ; but we have seen no trees of considerable size, and the weeping variety appears to be now the only kind usually cultivated. Loudon says that a tree in the Chiswick Garden was 12 ft. high eight years after planting ; and on account probably of its northern habitat it produced leaves so early that on the 2Oth April 1835 they were cut by frost. It is so like the aspen of the old world in appearance that even if it would grow, it is hardly worth cultivation except in botanic gardens. (H. J. E.) POPULUS GRANDIDENTATA Populus grandidentata, Michaux, Fl. Bor. Am. ii. 243 (1803); Loudon, Arb. et Frut. Brit. iii. 1650 (1838); Sargent, Suva N. Amer. ix. 161, t. 488 (1896), and Trees N. Amer. 155 (1905); Schneider, Laubholzkunde, i. 17 (1904); Dode, in Mém. Soc. Hist. Nat. Autun, xviii. 28 (1905). Gombocz, in Math. Termes. Közl. xxx. 138 (1911). A tree, attaining in America 70 ft. in height and 6 ft. in girth ; bark like that of the common aspen. Young branchlets covered at first with a greyish tomentum, persistent more or less during summer. Buds ovoid, acute, grey tomentose. Leaves (Plate 408, Fig. 7) on the long shoots, 3 to 4 in. long, 2 to 3 in. broad, ovate- deltoid ; truncate, rounded, or cuneate at the base ; acuminate at the apex ; glabrescent and dark green above; lower surface pale or glaucous green, glabrescent or with traces of the grey tomentum, which is dense at the time of unfolding of the leaves ; margin with a narrow translucent border, and with a few sinuate triangular teeth, but entire near the base ; petioles slender, laterally compressed, glabrescent. Leaves on the short shoots, oval, with sharper teeth, and often with two glands at the summit of the petiole. 1 Loudon, under P. gnzca, p. 1651. Populus Catkins with deep narrowly lobed scales, fringed with long hairs ; differing chiefly from the other aspens in the pubescent disc and ovary ; stamens six to twelve ; style divided into four long filiform lobes ; capsule two-valved. This species is much less widely spread in North America than the other aspen (P. tremuloides), apparently requiring a moister soil, and mainly growing in deep sand on the banks of rivers and swamps. It occurs in Ontario, southern Quebec, Nova Scotia, and New Brunswick, extending southward in the United States to north Delaware on the coast, and along the Alleghany Mountains to North Carolina, and westwards to Minnesota, Wisconsin, Illinois, Indiana, Kentucky, and Tennessee. According to Dame and Brooks,1 it is best distinguished from P. tremu loides in early spring, by the colour of the unfolding leaves, which are cottony white, whilst those of P. tremiiloides appear yellowish green. The leaves when open are much larger and more coarsely toothed, and the buds divergent, dull, and dusty - looking ; whilst those of P. tremidoides are mostly appressed and highly polished with a resinous lustre. In Canada it generally grows on sandy soil, mixed with pines, and is often mistaken for aspen. It is never a large tree, though usually larger than P. tremuloides, and as Elwes saw it, near Ottawa, is a straggling, ill- shaped tree of 40 to 50 ft. high, liable to be broken by the wind, and of little or no value either for use or ornament. Though introduced, according to Loudon,2 in 1772, it has always been a scarce tree in England, and the only specimen at Kew died about a year ago. At Grays- wood, Haslemere, a tree obtained from Meehan in 1887 is only 16 ft. high, and apparently this species does not thrive in our climate. The weeping grafted tree, commonly cultivated under the name P. grandidentata, van pendula, differs from that species in flowers and other characters,8 and may be distinguished as follows :— Popuhis pseudo-grandidentata, Dode, in Mém. Soc. Hist. Nat. A^lt^ln, xviii. 31 (1905). Populus tremula, Linnaeus, var. pseudo-grandidentata, Ascherson and Graebner, Syn. Mittcleurop. Flora, iv. 26 (1908). Young branchlets stout, dark reddish, with orange lenticels, covered with whitish tomentum in spring, which persists in summer at the base of the shoots. Buds viscid, tomentose near the top of the branchlet, glabrescent elsewhere. Leaves (Plate 408, Fig. 8) similar in shape and dentation to P. tremula, but larger, 3 to 4 in. in diameter, and thicker in texture ; margin with a translucent border, ciliate in spring. Staminate catkins, 2 in. long, with a slender pubescent axis ; pedicels glabrous ; stamens five, on an oblique glabrous shallow spatulate disc, which is entire in margin ; filaments slender, white ; anthers red. The origin of this plant is unknown, but it is probably a hybrid ; and if it came 1 Trees of New England, 32 (1902). 2 Trees and Shrubs, 823 (1842). It was introduced earlier into France, as it is well figured as P. tremula, ampliori folio, by Duhamel, Traité des Arbres, \\. 178, pi. 38, fig. 8 (1755)- 3 In P. grandidmtata the leaves are ovate, long acuminate, with fewer and larger teeth than in the weeping tree ; stamens more numerous, with short filaments ; pedicels and disc pubescent. 1794 The Trees of Great Britain and Ireland from America may be a cross between P. tremuloides and P. grandidentata} It appears to have been first mentioned by Simon-Louis2 in 1869. Koch3 speaks of it as a pendulous tree existing in England in 1872. There are good specimens at Abbotsbury and at Glasnevin. (A. H.) POPULUS SIEBOLDII Pofulus Sieboldii, Miquel, in Ann. Mus. Bot. Lugd. Bat. (excl. pi. masc.4) iii. 29 (1867); Wesmael, in De Candolle, Prod. xvi. 2, p. 327 (in part) (1864) ; Schneider, Laubholzkunde, i. 17 (1904); Dode, in Mem. Soc. Hist. Nat. Autun, xviii. 32 (1905); Gombocz, in Math. Termes. KSzl. xxx. 131 (1911). Populus tremula, Linnaeus, var. villosa, Franchet and Sa varier, Enum. PI. Jap. i. 465 (1875) (not Wesmael); Maximowicz,4 in Bull. Soc. Nat. Mosc. liv. 49 (1879); Shirasawa, Icon. Ess. For Japon, i. text 37, t. 18, figs, i-io (1900). Populus rotundifolia, Simon-Louis, ex Dippel, Laubholzkunde, ii. 192 (1892). A tree, attaining in Japan 60 ft. in height. Young branchlets stout, covered in spring with a dense white tomentum, persistent in part during summer. Buds more or less tomentose, not viscid. Leaves (Plate 408, Fig. 6) thicker in texture than those of P. tremula, densely tomentose and ciliate when young, glabrescent in summer, dark shining green above and yellowish or pale beneath, about 3 in. long and 2 in. broad, ovate, rounded or cuneate at the base, abruptly contracted into a glandular short acuminate apex ; margin with a translucent border, minutely (varying even on the same leaf) sinuately toothed or glandular serrate ; basal glands usually well developed ; petiole slender, pubescent, laterally compressed. Flowers similar to those of P. tremula, but with the disc slightly pubescent. This species, which is very distinct in appearance from the common aspen, appears to be confined to Japan, where it was collected at Aomori by Elwes. It appears to have been introduced by Simon-Louis about 1887, but the only tree which I have seen is a grafted one at Glasnevin, planted in that year, which has only attained a height of 15 ft. and is not very thriving. (A. H.) POPULUS FREMONTII Populus Fremontii, Watson, in Proc. Am. Acad. x. 350 (1875); Sargent, Suva N. Amer. ix. 183, t. 496 (1896), and Trees N. Amer. 164 (1905); Dode, in Mém. Soc. Hist. Nat. Autun, xviii. 40 (1905); Gombocz, in Math. Termes. KSzl. xxx. 76 (1911). Populus monilifera, Torrey, in Sitgreave's Rep. 172 (1853) (not Aiton). A tree, attaining 100 ft. in height and 15 ft. in girth ; bark at first smooth and thin, ultimately becoming on old trunks deeply fissured into broad rounded scaly 1 Nuttall, Gen. PI. ii. 239 (1818), describes as P. grandidentata, var. pendilla, a tree "with pendulous branches, as in the weeping ash, on the Alleghany Ridge, Pennsylvania, rare." Loudon, of. cit. 1651, states that there was a tree bearing this name in the Horticultural Society's garden in 1838 ; but that its branchlets were not pendulous. There is, there fore, no evidence that Nuttall's tree, seen only in the wild state, was ever introduced into cultivation. 2 Cat. Général, 1869, p. 73, where it is called P. grandidentata, var. pendula, and is described as a weeping form with large teeth to the leaves. 3 Dendrologie, ii. pt. i. 488 (1872). 4 The staminate specimen described by Miquel, which is preserved in the Leyden Herbarium, is a species of Carpinus, according to Maximowicz. ridges. Young branchlets glabrous. Buds small, viscid. Leaves on old trees deltoid, about 2^ in. broad, truncate at the base, and abruptly contracted at the apex into broad short entire points: on young cultivated trees (Plate 409, Fig. 13), reniform or rhombic, with a cuneate base and a similar apex ; serrations few, coarse, with incurved points ; margin with dense minute cilia, discernible with a good lens ; glands absent at the base ; petiole glabrous. Staminate catkins, 2 in. long ; axis glabrous ; disc broad, oblique, entire in margin ; stamens sixty, with dark red anthers. Pistillate catkins, 2 in. long ; pedicels short ; disc crenate, cup-shaped ; stigmas three, irregularly and crenately lobed. Fruiting catkins, 4 to 5 in. long ; capsule thick-walled, three- to four-valved. This species grows on the banks of streams in California, Lower California, Nevada, southern Utah, southern Colorado, and western Texas. It1 has only lately been introduced into cultivation, and small specimens may be seen at Kew and Glasnevin. Var. Wislizeni, Watson, in Amer. Journ. Sei. xv. 136 (1878). Populus Wislizeni, Sargent, Silva N. Amer. xiv. 71,1. 732 (1902), and Trees N. Amer. 165 (1905); Dode, in Mém. Soc. Hist. Nat. Autun, xviii. 39 (1905); Gombocz, in Math. Termes. Közl. xxx. 78(1911). This appears to have similar foliage, and is mainly distinguishable by the long pedicels of the flowers. It is the common poplar in the Rio Grande valley of western Texas and New Mexico, and the adjacent parts of Mexico. (A. H.) POPULUS NIGRA, BLACK POPLAR Populus nigra, Linnaeus,2 Sp. PI. 1034 (1753); Loudon, Arb. et Frut. Brit. iii. 1652 (1838); Wesmael, in De Candolle, Prod. xvi. 2, p. 327 (1868), and in Mém. Soc. Se. Hainaut, iii. 258 (1869); \\r\\\Vioram, Forstliche Flora, 527 (1887); Mathieu,/""/eweForestière, 491 (1897); Schneider, Laubholzkunde, i. 5 (1904); Dode, in Mém. Soc. Hist. Nat. Autun, xviii. ("groupe nigra") 37 (1905); Ascherson and Graebner, Syn. Mitteleurop. Fl. iv. 36 (1908); Gombocz, in Math. Termes. Közl. xxx. 85 (1911). A tree, attaining above 100 ft. in height and 20 ft. in girth, usually with a straight and single stem, but occasionally dividing near the base into several limbs ; with wide- spreading stout and irregular branches, not slender and regularly ascending as in many of the hybrids. Bark deeply furrowed on old trunks, and often covered with large burrs. In all its forms this species is readily distinguishable from the American species and the hybrids by the leaves, non-ciliate on the margin, without glands at the base, and when well-developed gradually tapering from the middle of the blade to a long acuminate apex. The black poplar, and apparently all the poplars of the same section, rarely if ever produce suckers while the trees are living, but if one is cut down suckers are 1 According to Dode, the trees introduced are var. Wislizeni; but until they flower their identification is uncertain. However, Späth, Cat. No. 95, p. 89 (1895-1896), states that the young plants first introduced in 1894 were obtained from Colorado, where only the typical form of the species exists. 2 By this name Linnseus meant the black poplar inhabiting temperate Europe ; though he quotes a Virginian poplar ex Herb. Gronov. It is most convenient to assume, as the typical form of the species, the tree planted by Liiinseus at Upsala, which is still living, and from which I gathered specimens in 1908. 1796 The Trees of Great Britain and Ireland produced abundantly from the roots. Root cuttings are as readily propagated as ordinary cuttings. Dubard, in Ann. Sc. Nat. xvii. 147 (1903), gives an elaborate account of the peculiarities of the suckers of this tree. This species comprises two distinct forms, one glabrous in all its parts, the other more or less pubescent, described in detail as follows : — 1. Var. typica, Schneider, op. cit. 5 ; Ascherson and Graebner, op. cit. 39. Con tinental Black Poplar. Young branchlets rounded, glabrous, ashy grey in the second year. Buds reddish, viscid, glabrous, closely appressed to the twig at their base, with a sharp apex curving outwards. Leaves on the long shoots (Plate 409, Fig. n), about 3 in. long, and 2 in. broad, cuneate at the base, gradually tapering from about the lower third, where they are widest, towards the long acuminate apex ; glabrous ; dark green above, light green below ; margin with a narrow translucent border, non-ciliate, finely and crenately glandular-serrate ; petiole glabrous, laterally com pressed. On the short shoots the leaves are smaller, broader at the base, which is often less cuneate, and truncate or rounded ; and similar leaves often occur on old trees even on the long shoots. Catkins about \\ in. long, with early deciduous scales, which are broadly obovate, and divided into numerous irregular, linear entire or lanceolate toothed lobes ; axis glabrous. Staminate flowers sub-sessile ; stamens twenty to thirty on an oblique concave non-ciliate glabrous disc, which is slightly waved and upturned in margin ; filaments white, thread-like, as long as the deep red anthers. Pistillate flowers shortly stalked ; ovary glabrous, globose, in a cup-like glabrous disc ; stigmas two, dilated, crenate in margin, closely appressed at first to the sides of the upper part of the ovary. Capsules two-valved, glabrous, on long pedicels. 2. Var. betulifolia, Torrey, Fl. New York, ii. 216 (1843) ; Skan, in Bot. Mag. t. 8298 (in part1) (1910) ; Schneider, Laubholzkunde, ii. 870 (1912). English Black Poplar. Var. virtdis? Lindley, ex Louden, Arb. et Frut. Brit. iii. 1652 (1838). Var. betulœfolia, Wesmael, in De Candolle, Prod. xvi. 2, p. 328 (1868). Var. hudsonica, Schneider, Laubholzkunde, i. 5 (1904); Ascherson and Graebner, Syn. Mitteleurop. Fl. iv. 39 (1908). Populus nigra, Michaux, Fl. Bor. Am. ii. 244 (1803) (not Linnaeus). Populus hudsonica, Michaux f., Hist. Arb. Amer. iii. 293, t. 10, fig. I (1813), and N. Amer. Sylva, ii. 114, t. 96, fig. i (1819). Populus betulifolia, Pursh, Fl. Amer. Sept. ii. 619 (1814); Loudon, Arb. et Frut. Brit. iii. 1656 (1838); Dode, in Mem. See. Hist. Nat. Autun, xviii. 48 (1905). Populus nigra, Vaillantiana, and Muelleriana, Dode, in Mém. Soc. Hist. Nat. Autun, xviii. 48 Young branchlets rounded, covered with a dense short pubescence, orange or yellow in the second year. Buds greenish, tinged with brown, viscid, otherwise as 1 The female catkins figured show ovaries with three stigmas, and were taken from a hybrid tree (P. Lloydii) growing near Turnham Green station. Cf. p. 1831, note I. 2 Lindley's specimen in the Cambridge Herbarium is a vigorous branch from a young tree of P. nigra, var. betulifolia. This is confirmed by a tree labelled var. viridis in the Cambridge Botanic Garden. Loudon's description of this variety, " leaves of a brighter green than the species," is inadequate ; and the name var. viridis, though older than that of var. betulifolia, cannot be used. Mackie, in Loudon, Card. Mag. xiii. 230 (1837), says that var. viridis was discovered at Beatings, near Woodbridge, and had been grown in his nursery at Norwich for twenty years. Populus 1797 invar, typica. Leaves1 (Plate 409, Fig. 12) similar in shape, colour, size, and margin to those of var. typica, but slightly pubescent when young ; petioles pubescent. Catkins i^ to 2 in. long, as in the typical variety, but with a pubescent axis ; stamens in the specimens examined, fewer, about twelve to fifteen ; scales, ovary, and stigmas, identical. Fruiting catkins—on the Bury St. Edmunds tree, which was probably fertilised by staminate trees of the same variety close beside it—about 4 in. long, with ovoid capsules about \ in. in length, glabrous and tuberculate on the outer surface ; seed oblanceolate, yellowish, about ^ in. long, covered with dense cottony hairs enveloping the whole catkin after the dehiscence of the capsules. Forms, in which glabrous catkins are associated with pubescent leaves and branchlets, occur ; and on this account I have refrained from making var. betulifolia a distinct species.2 The pubescent variety of the black poplar is the only form occurring in England in the wild state ; and it is also a native of the greater part of France, from Normandy and Picardy to the foot of the Pyrenees. In 1912 I saw it apparently wild in many places, as in hedges on hills not far from Argentan, where it grows in a small and stunted form. It is most common, however, as a fine tree, often with a burry trunk, along the banks of the great rivers, as on the Seine at Mantes, on the Loire near Tours, on the Garonne in the vicinity both of Bordeaux and of Toulouse, on the Adour between Bayonne and Dax, and on the Gave de Pau, where there are two good trees in a meadow opposite the shrine at Lourdes. This poplar is also frequently planted in botanic gardens, as at Le Mans, Tours, and Montauban, a fine tree in the latter place measuring 90 ft. by 13 ft. There is a specimen in the Montpellier Herbarium, gathered at Ganges on an island in the river Hérault ; but I have seen no specimens from Provence.8 It is remarkable how this variety has escaped the notice of British botanists, though it has been collected from early times, as there are specimens in the British Museum4 gathered by Plukenet and Buddle towards the end of the seventeenth century. This tree was first distinguished by the younger Michaux, who found it growing on the banks of the Hudson river above Albany, and mentions large specimens planted in New York city; but adds that he never saw it in the forest. Sargent5 in 1896 stated that it was growing then on an island in the Delaware river near Easton, Pennsylvania; but in a letter to Kew, dated 3151 July 1902, he mentions only a single specimen known to him, an old tree near Boston ;6 and adds that it is 1 On very old trees the leaves are smaller, truncate or occasionally subcordate at the base, and with a shorter acumen at the apex. These appear to be P. Muelleriana, Dode. 2 P. nigra, var. pubescent, Parlatore, Fl. Ital. iv. 289 (1867), was described from trees growing in moist valleys at S. Martino, Palermo; and is recorded in Thessaly by Halacsy, Consp. Fl. Grcec. iii. 136 (1904). A specimen in the Cambridge Herbarium has branchlets, leaves, petioles, and female catkins covered with long white hairs ; and is much more pubescent than trees growing in England. P. hispida, Haussknecht, which I have not seen, is probably the same as var. pubescens. 3 Fardé, in Bull. Soc. Dend. France, 1911, p. 255, states that P. nigra is pretty common in Provence; but this is probably var. typica. 4 Herb. Sloan, 83, fol. 8, and 126, fol. 6. 6 Silva N. Amer. ix. 153, note (1896). In the Montpellier Herbarium there is a specimen of this tree labelled " Populus, New York, growing planted opposite Dr. Hosack's door in Broadway, May 7, 1807." Another specimen named P. hudsonica, Michaux, was taken from a tree growing at Versailles in 1808. 8 This is no doubt the tree which I saw growing on the shore of Jamaica Pond, when staying with Prof. Sargent in 1904, and recognised at once as the English black poplar by its burry trunk and foliage.—H. J. E. VII u 1798 The Trees of Great Britain and Ireland sold in the United States by Ellwanger and Barry of Rochester under the name of P. elegans? There is little doubt that this poplar was introduced into the United States in the eighteenth century from England. 3. Var. italica, Du Roi, Harbk. Barmiz. ii. 141 (1772). Lombardy Poplar. Var. pyramidalis, Spach, in Ann. Sc. Nat. xv. 31 (1841). Populus italica, Moench, Bäume Weissenstein, 79 (1785). Populus dilatata, Aiton, Hort. Kew. iii. 406 (1789). Populuspyramidata, Moench, Meth. 339 (1794)- Populuspyramidalis, Borkhausen, Forstbot, i. 541 (1800). Populus fastigiata, Poiret, in Lamarck, Encycl. v. 235 (1804); Loudon, Arb. et Frut. Brit. iii. 1660 (1838). Branches directed nearly vertically upwards, forming a narrow fastigiate tree. A sport of the typical glabrous variety of P. nigra, differing in no respect except in habit.2 The leaves are variable, many being the same as those of the ordinary form ; but others are often broader than long, truncate or subcordate at the base, with a short acuminate apex, due to increased vigour, as is usual in this species. The common Lombardy poplar is a staminate tree, always reproduced by cuttings ; and for aught we know, all the numerous individuals planted throughout the world may have originated from a single tree, as happened, without any doubt, in the case of the upright form of the common yew. No instances of a second origin have been recorded. A few treess of similar habit, though with branches not quite so vertically inclined, have been observed bearing pistillate flowers.* Plate 383, reproduced from a photograph sent to us by the late Prof. W. Blasius, shows a remarkable female tree at the village of Greene, near Kreiensen, in the Duchy of Brunswick, which has ascending and not erect branches, and differs considerably in habit from the ordinary Lombardy poplar. There is a good specimen at Kew with nearly erect branches, about 50 feet high, which was covered with woolly catkins in 1908. It produced flowers in the spring of 1910, which did not, however, ripen into fruit 1 This is referred to as a variety of P. nigra, commonly sold by nurserymen in the United States, by L. H. Bailey, in Cornell Univ. Bull. Agric. No. 68, p. 227 (1894). 2 I carefully compared in 1908, in Servia, the branchlets, foliage, and buds of a Lombardy poplar with those of some wild common black poplars growing near it, and did not detect the slightest difference. The bark of some trees in this region, and also in Algeria, is remarkably whitish ; while the colour of their third year and older twigs is peculiarly greyish. This form has been named P. thevestina, Dode, in Mem. See. Hist. Nat. Auiun, xviii. 52 (1905). Siehe, in Mitt. jDeut. jDend. Ges. 1912, p. 123, describes the remarkable pale bark of the Lombardy poplar in Asia Minor. Cuttings were obtained in Algeria by Mr. A. W. Hill in 1910, which are now growing in the nursery at Kew. Vigorous shoots from near the base of old Lombardy poplars at Cambridge show the same coloration. 3 Spenner, Fl. Friburg, i. 274 (1825), mentions a female tree near the Carthusian monastery at Friburg in Germany. Another was reported to have been noticed in the University Botanic Garden at Göttingen in 1828 (cf. Denson, in Loudon, Gard. Mag. vi. 419 (1830)). Loudon, Derby Arboretum, 57 (1840), states that the female tree was introduced in 1840 into the Horticultural Society's garden from Monza near Milan. There is a specimen in the Kew Herbarium, sent from Carlsruhe by A. Braun in 1845, and others undated, which are labelled Frankfort-on-Oder and Switzerland. Mr. W. L. Wood also noticed, in 1910, two smaller trees with pistillate catkins growing in the Walpole Road, Twickenham. 4 P. fannonica, Kitaibel, ex Besser, PI. Enum. Volhynice, 38 (1822), and Reichenbach, Icon. Fl. Germ. xi. 30, t. 619 (1845) (figured with rhombic acuminate leaves), is possibly the correct name of the female Lombardy poplar. Besser, in Flora, 1832, ii. Suppl. 14, states that P. croatica, which was published at the same time as P. fannonica, is the name that should be applied to the supposed cross between typical /'. nigra and the Lombardy poplar, which grew in the Theresa garden at Vienna, the sex of which is not mentioned. It was supposed to occur wild on the Dnieper. Zawadski, Enum. PI. Galic. 117 (1835), saw fine specimens of this on the Dniester in Podolia. Petzold and Kirchner, Arb. Muse. 593 (1864), state that they received P. fastigiata, var. pannonica, from many sources, but never were able to discover any distinguishing characters. Cf. also P. nigra, var. fannonica, Dippel, Laubhohkunde, ii. 198 (1892). Populus 1799 in this season, though there are staminate trees at no great distance. It opens its flowers at the same time as the Lombardy poplar. The latter is about three weeks earlier than the native black poplar, an indication of its southern origin.1 The history of these pistillate trees is quite unknown ; but they may have arisen as the result of hybridisation between the staminate Lombardy and the ordinary poplars. The staminate Lombardy poplar appears to have originated on the banks of the river Po in northern Italy, probably in the beginning of the eighteenth century, as it was unknown to classical writers 2 and is not mentioned by mediaeval Italian authors.8 Moreover, it was not noticed by Ray and other English travellers in Italy in the seventeenth century. Séguier,4 an old writer, states that it was known anciently in Lombardy, and mentions a superb avenue, which he saw in 1763 at Colorno, the residence of the Duke of Parma. It was apparently carried5 by the Genoese to the Levant ; and there are no grounds for supposing that it originated in Asia Minor or Afghanistan, as Royle,6 who first made this statement, simply relied on the fact that it bore a native Persian name.7 W. G. Browne,8 who travelled in Asia Minor in 1798, makes the first reference to its occurrence in western Asia, where he states that it abounds all over the plain of Damascus, and when old becomes rugged and uncouth, as usual in other regions. It was introduced from Lombardy into France in 1749 ; and is usually stated to have been brought into England and planted at St. Osyth's in Essex, in 1758, by the Earl of Rochford, who was ambassador in Turin at the time. It was possibly, how ever, first planted at Whitton some years earlier by Archibald, Duke of Argyll, who died in 1761, as the tree still growing there in 1838 was much larger than any of the others recorded by Loudon, being 115 ft. high and 19 ft. 8 in. in girth at 2 feet from the ground.9 1 Sargent, Suva N. Amer. ix. 154, note (1896), says that the fact that the Lombardy poplar does not suffer in the cold of the Canadian winter, shows that it originated in a climate much more severe than that of northern Italy. (The winter in the plain of the Po, it may be stated, is very cold, the mean temperature being below 35° Fahr. ; and in Milan the thermo meter sometimes sinks below zero.) Prof. Budd, quoted by L. II. Bailey, Cornell Univ. Bull. Agric. No. 68, p. 228 (1894), however, explains that the Lombardy poplar, grown in Canada, was imported from Voronej in central Russia, where it has become acclimatised, and is perfectly hardy. The Russian botanists assured him that its hardiness depended on the region from whence it was obtained. Bailey, Survival of the Unlike, 297 (1896), in an interesting chapter on acclimatisation of trees, states that cuttings of the white poplar, taken from trees at Montpellier and at Geneva, which were planted at the latter place, differed as much as twenty-five days in their time of coming into leaf ; and similar results were obtained at Ithaca (New York) with cuttings of the Lombardy poplar. 2 The often quoted lines of Ovid, Met. ii. 345-360, and of Virgil, ^.neid, x. 190, do not refer to this tree, as has been supposed. The poplars depicted by Perugino (1446-1524), as in a picture in the London National Gallery, are slender, but not with vertical branches, and are probably Populus alba. Cf. Rosen, Die Natur in der Kunst, 293 (1903). 3 Rostafinski, in Verhandl. K. K. Zool. Bot. Gesell. Wien, xxii. 170 (1872), states that it was introduced from Italy into Poland by King Sobieski, who reigned from 1624 to 1696, and that the original trees are still standing in the garden of the Wilanow Castle, near Warsaw. Miss Ivanovska, at my request, examined the old poplars there, which proved to be all of the ordinary wide-spreading form ; and Prof. Rostafinski in a letter acknowledges that he made a mistake. 4 Hist. Plant. Nat. Envir. Vérone, ii. 267 (1745). 0 Fougeroux de Bondaroy, in Mini, f Agric., Paris, 1786, p. 84. 6 Illust. Bot. Him. i. 344 (1839). Griffith's statement that it is wild near Kabul, at 7000 ft. altitude, is not confirmed. Aitchison, in Trans. Bot. Soc. Edin. xviii. 162 (1891), says: "I only met with this tree cultivated in orchards or near houses in Afghanistan and north-east Persia." 7 Boissier, Fl. Orient, iv. 1194 (1879), doubts its existence in the wild state in western Asia; and his reference to it being perhaps wild in the Karatau mountain in Turkestan is an error, as the wild fastigiate poplar in this locality is P. alba, var. fyramidalis. Cf. p. 1778. 8 Travels in Africa, Egypt, and Syria, 397, 408 (1799). Siehe, in Mitt. Deut. Dend. Ges. 1912, p. 123, states that in Asia Minor the Lombardy poplar is extensively cultivated and is a most useful tree, producing after twenty or thirty years' growth, long slender but tough beams, which are much used in house-building. 8 Loudon, Arb. et Frut. Brit. i. 58 (1838). i8oo The Trees of Great Britain and Ireland According to Sargent,1 it was brought to America in 1784 by W. Hamilton, who planted it in his garden at Woodlands near Philadelphia. The Lombardy poplar appears to be a short-lived tree, and is said to be dying out in Germany. As it does not now apparently attain the immense size recorded in former years, there may be some truth in the opinion advanced by Focke 2 that as all the trees have been raised by cuttings since the origin of the first sport, they may now be dying of old age. (A. H.) The most reliable account we have of the introduction of this tree is that given by Aiton, Hortus Kewensis, iii. 406 (1789), who states that it was brought by Lord Rochford from Turin, where he was ambassador about 1758. He planted cuttings at St. Osyth's Priory in Essex, where two trees now much decayed still survive. I am indebted to Mr. J. Edge, the gardener there, for photographs and measure ments of the larger of these trees, which show a large hollow stump divided into two trunks about 20 feet high and measuring 18 ft. in girth at 3 feet from the ground. Living branches have sprung from different places in the trunk, two of which attain a height of about 50 feet. Loudon records a large tree of the same age as the last, which was blown down at Canterbury in 1836 ; but the tallest tree mentioned by him was at Great Tew in Oxfordshire, said to have been 125 feet high when only fifty years old. A tree was recorded by Thomas Hogg, forester at Hampton Court, Herefordshire,3 as growing at Wharton Court farm near that place, which in 1879 was said to measure no less than 160 feet. I visited this place in 1905 and found no trace of its remains ; but if the height was correct, which from the other measurements given of trees on the estate seems probable, it was much taller than any that I have measured in England or France. Sir Hugh Beevor tells me of a tree at Pitchford, Shropshire, which in 1907 was 120 ft. by 14 ft. 8 in. ; and I have measured many of 100 to 115 feet, but none which can be said to stand out from the average of mature trees. Henry measured a tree at Shiplake House, near Henley, which was 105 ft. by 10 ft. 10 in. in 1905; and another at Alderbourn Manor, Gerard's Cross, which was 100 ft. by 12 ft. in 1912, and visible for many miles around. An old tree in Lensfield Road, Cambridge, 90 ft. high in 1904, of which a photograph was sent me by Mr. Lynch, was removed in 1912. J. Smith4 recorded a tree growing at Fox Mills near Romsey as 125 ft. by 13 ft. 2 in., but when I was there in 1900 I could not find it; and another at Greatbridge House, near Romsey, which was 130 ft. by 13 ft. 9 in. These died in 1881, no doubt from the effects of the inclemency of the weather in 1879-80. The seasons of 1879-81 appear to have killed a very large number of Lombardy poplars in the eastern and midland counties, not perhaps so much by their excessive 1 Sargent, op. cit. 154, note (1896). 2 In Cart. Zeit. September 1883, quoted in Card. Chron. xx. 571 (1883). Cf. also Rev. des Eaux et Forêts, xxiv. 277 (1885). Manetti's letter quoted by Loudon, Card. Mag. xii. 450 (1836), is rather obscure, and his statement that plants were raised in Italy from seed, which preserved the characters of their parents, is extremely doubtful. In the Cambridge Herbarium, however, there is a specimen, with female flowers and fruit, of a poplar sent by Manetti, which he considered to be the female Lombardy poplar. 3 Trans. Scot. Art. Soc. ix. 151 (1879). « Ibid. xi. 534 (1887). Populus 1801 cold, as because of the two unusually cold and wet summers. A paper on the subject by Mr. H. D. Geldart1 gives many interesting particulars ; and all his correspondents seem to agree that the damage was much greater than in the colder winter of 1860-61, when the thermometer at Audley End near Saffron Waiden went down on Christmas morning to — n°. Mr. Geldart quotes the replies received by Mr. Southwell to his inquiries as to the death of Lombardy poplars in other parts of England as follows. Twenty miles round York, death or severe injury was almost universal ; in Wilts, they had suffered very much ; in west Dorset, most of them were killed or seriously injured ; near Doncaster, all were more or less killed ; but at Oxford among the College gardens it was the exception to find a damaged tree ; and Mr. C. B. Plowright wrote that in the west of England the Lombardy poplars did not seem to be injured at all. In Scotland it was, according to Dr. Walker, introduced at New Posso in Tweeddale as early as 1765 from cuttings sent by the Earl of Hertford, and was extensively distributed some years later by Lord Gardenstone ; but the climate of most parts of Scotland is evidently too cold or too wet to suit this tree, as none of those mentioned by Loudon are still alive so far as I can learn, except one at Brahan Castle, Ross-shire, that was 70 ft. by 6 ft. in 1838. When I visited this place in 1907 I found a fine tree 98 ft. by about 9 ft., which may be the same. Six miles north of Inverness, on the high-road to Beauly, I also saw four well-shaped trees, of which the largest was 90 ft. by 11^ ft., showing the excellent climate of that district. Mr. Renwick tells us of one near Braidwood, Lanarkshire, which was 93 ft. by 10^ ft. in 1910; this was blown down on 5th November 1911. In France the Lombardy poplar is common, though now on account of its inferior growth often replaced by the hybrid poplars. It commonly attains no ft. to 120 ft. in height, but I have seen none approaching the trees near Rouen mentioned by Loudon2 which, according to M. Dubreuil, were then 150 ft. high. A tree8 at the Trianon, Versailles, was 17^ ft. in girth at four feet from the ground in 1888. The Lombardy poplar has been planted largely in the irrigated districts of Utah, and, according to F. C. Sears,4 rows of tall Lombardy poplars, marking the irriga tion canals, are a feature in the landscape. (H. J. E.) In Chile,5 especially about Valparaiso, the Lombardy poplar is largely planted, both in gardens and on the margins of the irrigation canals, where it grows so rapidly as to be ready for felling in fifteen years. Its timber is used for indoor work in houses. Dode states6 that in Chile and Argentina there is a form of the Lombardy poplar which keeps its leaves evergreen. Mr. Lovegrove has sent us specimens of the Lombardy poplar from Kashmir, where it is planted along roads, and often attains 100 ft. in height and 7 ft. in girth. 1 Trans. Nor/, and Norm. Nat. Soc. iii. 354-366 (1880-1884). Cf. Gard. Chron. xv. 764, 798, and xvi. 246 (1881), where instances are given of the death of many trees also in the north of France and in Belgium. 2 Art. et Frut. Brit. iii. 1670 (1838). 3 Garden and Forest, i. 174 (1888). * Ibid. x. 357 (1897). 6 Dr. W. Balfour Gourlay, in Trans. Bot. Soc. Edin. xxiv. 74, plate 7 (1910). It is much attacked in Chile by the quintal, Loranthus tetrandus. e Bull. Soc. Dend. France, 1908, p. 29, and 1909, p. 152, where this form is named P. fyramidalis, var. Thaysiana, Dode. 1802 The Trees of Great Britain and Ireland Its timber is used in house-building, and lasts well when protected from rain. A tree 5 ft. in girth sells for 12 to 26 shillings. 4. Var. plantierensis, Schneider, Laiibholzkunde, i. 803 (1906). Populus fastigiataplantierensis, Simon-Louis, Cat. 1884-1885, p. 51. Populus plantierensis, Dode, in Mém. Soc. Hist. Nat. Autun, xviii. 43 (1905). A fastigiate form of var. betulifolia, similar to the Lombardy poplar in habit and foliage, but with reddish pubescent petioles, and shortly pubescent branchlets. It originated in Simon-Louis's nursery at Plantières, near Metz, whence it derives its name, and is said to occur in both sexes, and to be the result of a cross between the Lombardy poplar and var. betiilifolia ; but this origin is unlikely. The original tree is a male, and when measured by Elwes in 1908 was 74 ft. by 5 ft. It is claimed for it that it is more vigorous than the ordinary fastigiate poplar, and not liable to die off at the top, as is frequent in the latter. There are specimens at Kew about 25 ft. high. 5. Var. viadri, Ascherson and Graebner, Syn. Mitteleurop. Flora, iv. 40 (1908). Populus viadri, Rüdiger, in Abhand. Naturw. Ver. Reg. Bez. Frankfurt, viii. Mon. Mitt. 12 (1891); Koehne, in Verh. Bot. Ver. Brandenb. xxxvii. p. xxviii (1895), and in Gartenßora, xxxix. 447 (1890). A narrow pyramidal tree, with ascending branches, which in branchlets, buds, and foliage resembles the typical glabrous form of the species.1 It occurs along the banks of the Oder, near Frankfort, whence it derives its name ( Viadrus being the Latin name for this river). It is said to produce pistillate flowers identical with those of P. nigra; but a tree2 at Kew, about 25 ft. high, produced staminate flowers in 1910. These differ slightly from the type in occasionally having peculiar scales deeply bilobed at the summit. (A. H.) DISTRIBUTION The distribution of the black poplar is very wide in Europe, but difficult to define accurately, as it has been much planted in former times. In Norway and Sweden, Schübeler only knew it as a planted tree, and figured a very large one, which from its burry trunk has the appearance of the English tree. This grew on the banks of a river at Ronneby in Sweden ; and when measured in 1882 by Prof. Wittrock, was, at four feet from the base, 34 ft. 8 in. in girth, dividing into two main trunks a little higher up ; and this is the largest girth of which I have any record. In the Botanic Garden at St. Petersburg a tree, supposed to have been planted by Peter the Great, measured in 1908 about 90 ft. high by 17 ft. in girth, forking low down, and with a large burr on its trunk. In north Russia it extends to 57° N., according to Von Herder; but in the St. Petersburg Herbarium I found a note by Kusnetsov stating that it was found at 1 Koehne, who mentions trees of both sexes in Deut. Detid. 84 (1893), describes this peculiar poplar as a hybrid, P. candicans x nigra, but he afterwards withdrew this very unlikely hypothesis. The leaves are more cuspidate at the apex than in ordinary P. nigra, and it is possible that P. viadri is a hybrid, but I have seen no pistillate flowers. P. viadri appears to have been introduced into cultivation by Späth, as it is mentioned as a novelty in his Catalogtie, No. 91, p. 96 (1893-1894). 2 Another tree at Kew, labelled P. viadri, also obtained from Rüdiger, is different in habit, having spreading and not ascending branches. It is pistillate, and appears to differ in no respect from P. nigra, var. typica. Populus 1803 Emetskoie on the Dwina in lat. 63° 30'. It is found on the Volga1 as far south as Astrachan, and on the Ufa river, where, according to Loffiewsky, it attains too ft. high. I also found specimens in the herbarium under the name of P. nigra from Zlataoust on the southern Ural, from Tobolsk and Barnaoul, and from the Yenesei river in lat. 66° N., collected by Brenner, with very small leaves. In France I saw a large female tree at Chenonceaux, near Tours, which measured 100 ft. by 10 ft. 10 in. in 1908. At Dijon there is in the Botanic Garden a very large tree2 of this species, probably the oldest in Europe. It is said from historical documents to be over five centuries old, and bears an inscription to the effect that in 1866 it was at the ground 12 metres, and at two metres high 8 metres in girth, and contained about 40 cubic metres_about 1400 cubic feet of timber. It is quite hollow at the base, the shell being only a foot or so in thickness, and has deeply ridged bark, 4 to 6 in. in thickness. As nearly as I could estimate, its present height is about 125 ft., but having lost several branches has probably been taller. Its girth at 5 ft. is 26 ft. 7 in. The leaves though small are typical in shape, but the trunk is not as burry as in English trees. I could not learn whether it is male or female. M. Mathey, inspector of forests, told me that the black poplar which was formerly common around Dijon is now growing scarce, being superseded, as in England, by P. serotina. In Spain and Portugal I only saw small and stunted trees ; and in Italy I have seen none which looked like true P. nigra. In Austria it is common in the valley of the Danube; and in the Prater at Vienna there are many good-sized trees. In Greece, Heldreich records it from Mt. Pelion. In Morocco, Maw and Ball collected specimens in the Atlas at 3000 to 4030 feet. In Great Britain it has been occasionally confused by local botanists with P. serotina; and I can find nothing in the older works, such as those of Evelyn, Miller, and Boutcher, to show that these authors knew the tree from personal observation. Though rare in most parts of England it seems to be a native of the counties on the Welsh border, where it is still fairly common ; and it is probably indigenous in Norfolk, Suffolk, Cambridgeshire, and Essex. It is so very distinct in trunk, foliage, and time of leafing from P. serotina, that it is extraordinary th