256 WILSON EXPEDITION TO CHINA 



winter-buds of Japanese plants of U. laciniata are spindle-shaped and rather acute, 

 resembling those of U. laevis Pallas, but in plants from ChiU grown in the Arnold 

 Arboretum the buds are more ovate and the deep red-brown scales are very hairy 

 on the margin, while in Japanese plants which were brought by J. G. Jack from 

 Chuzenji the scales are only very finely ciUate and of a lighter color in the lower part. 

 Whether or not the plants from northern China may represent a distinct variety 

 needs further investigation. 



This species occurs sparingly in moist woods near Lake Yumoto, in the Nikko 

 region, and in slightly increasing numbers northward, but it is nowhere plentiful in 

 Hondo. In Hokkaido it is fairly common, more especially in the central parts of 

 this island, and always is associating with mixed trees. As I saw it this Elm is in 

 Japan always a small, slender and unimportant tree, not more than 12 m. tall, with 

 a trunk scarcely 1 m. in girth and covered with gray, fairly smooth, fibrous bark. 

 My observations as to the size of this tree agree with those of Maximowicz and Sar- 

 gent. Mayr, who says it grows 30 m. tall and is an important constituent of the 

 forests in central Hokkaido, confuses it with U. japonica Sargent. On the mainland 

 of eastern Asia it would appear to be a larger tree than in Japan, for Sargent noted 

 it as a " very large tree on the mountains twelve hours east of Harbin by railroad." 



A picture will be found under No. x365 of the collection of my Japanese photo- 

 graphs. E. H. W. 



Subsect. 2. FOLIACEAE Schneider in Oester. Bot. Zeitschr. LXVI. (1916). 

 This subsection contains the most difficult forms, especially those of the following 

 series. 



Series a. Nitentes Moss, Cambridge Brit. Fl. II. 89 (1914). — Schneider in 

 Oester. Bot. Zeitschr. LXVI. (1916). 



The species of this series are very closely related, and it is difficult to distinguish 

 some of them without carefully collected material with flowers, fruits and matvu-e 

 leaves. In Europe they are connected with the species of ser. Glabrae by many 

 natural hybrids, but we do not yet know if there are any such hybrids in eastern 

 Asia or in India. 



9. Ulmus castaneifolia Hemsley in Jour. Linn. Soc. XXVI. 446, t. 10 (1894). — 

 E. Pritzel in Bot. Jahrb. XXIX. 296 (1900). — Schneider, III. Handb. Laubholzk. 

 II. 904, fig. 565 c-d, 566 e-f (1912). 



CHINA. Hupeh: Changyang Hsien, A. Henry (No. 7780, type; with ripe 

 fruits on leafless branchlets) ; same locality, mountains, summer 1900, E. H. Wil- 

 son (Veitch Exped. No. 677 in part; sterile); Chienshi Hsien, May 1900, E. H. 

 Wilson (Veitch Exped. No. 545; with ripe fruits and very young leaves). Eastern 

 Szech'uan: Wushan Hsien, A. Henry (No. 5498, co-type; sterile, young and old 

 leaves). 



The young flowers of this species which seems to have a very hmited distribu- 

 tion in central China are not yet known. The species is well characterized by 

 the shape of its leaves and fruits, as described in the key. The largest leaf I have 

 seen is about 14 cm. long and 5 cm. broad, wthout the rather thick petiole, which 

 measures 10 mm. in length. The surfaces of the leaves are more or less rough, but 

 otherwise glabrous with the exception of a few hairs on the impressed rib of the 

 upper surface and the clusters of gray hairs in the axils of the prominent lateral 

 veins beneath. 



10. Ulmus Wilsoniana Schneider. See p. 238. 



Ulmus Wilsoniana Schneider, var. psilophylla Schneider. See p. 239. 



