262 WILSON EXPEDITION TO CHINA 



branchlets and buds is not unlike in both the species as far as I can judge by the 

 material before me. U. Davidiana certainly needs further observation. See also 

 my remarks under U. macrocarpa Hance (p. 252). 



Series d. Pumilae Schneider in Oester. Bot. Zeitschr. LXVI. (1916). 



This series could probably be raised to the rank of a subsection equal to subsect. 

 Foliaceae, but there are some forms, especially of U. foliacea GiUbert in Europe and 

 western Asia, as also of other species in India and northeastern Asia, which are not 

 yet sufficiently known to decide whether or not U. pumila Linnaeus is really closely 

 alUed to any species of ser. Nitentes. 



15. Ulmus pumila Linnaeus. See p. 242. 



Ulmus pumila, var. arborea Litwinow in Sched. Herb. Fl. Ross. VI. 166, 



t. 3 (1908). 

 Ulmus campestris, y pumila Kegel in Act. Hort. Petrop. VI. 478 (1879). 

 Ulmus parvifolia Lauche, Deutsche Dendr. 348 (non Jacquin) (1880), exclud. 



synon. — ?Hooker f., Fl. Brit. Ind. V. 481 (1881). 

 ? Ulmus turcestanica i Regel in Gartenflora, XXXIII. 28 (nomen nudum) 



(1884). 

 Ulmus pinnato-ramosa Dieck ex Koehne in Fedde, Rep. Spec. Nov. VIII. 74 



(1910); in Mitt. Deutsch. Ges. 1910, 92 (1911). 

 Ulmus pumila, var. pinnato-ramosa Henry in Elwes & Henry, Trees Great Brit. 



& Irel. VII. 1926 (1913). 



TURKESTAN. 



INDIA. ?Kashmir. 



As far as I can judge by the cultivated specimens before me and by those col- 

 lected by P. Sintenis, "regio transcaspica," Krasnowodok, March 22, 1900 (No. 

 29^; with flowers), and Kisil-Arwat, Karakala, July 5, 1901 (No. 2016; sterile), 

 and distributed as U. glabra Miller, this variety is most closely related to the 

 typical U. pumila. The characters indicated by Koehne as distinguishing his 

 U. pinnato-ramosa can also be seen in specimens from northeastern Asia. 



There is an Elm from Kashmir: " Western Tibet, Nubra, alt. 10,000 ft." col- 

 lected by T. Thomson, which was distributed in Herb. Ind. Or. as U. pumila, and 

 later was referred by Hooker f., to U. parvifolia Jacquin. But Hooker says: 

 " The Tibetan plant may be a small-leaved form of U. Wallichiana." Unfortu- 

 nately the specimen is only a sterile one, and it may belong to U. pumila, var. 

 pinnato-ramosa as far as I can judge by the leaves and branchlets. 



In many respects very similar to U. pumila is also the type of U. campes- 

 tris, var. Aitchison in Trans. Linn. Soc. ser. 2, III. 108 (1880) from Khorasan, 

 July 11, 1885 (No. 711; a cultivated tree; sterile), which may be described as fol- 

 lows: Arbor; ramuli hornotini graciles, glabri, fiavescentes, deinde cinerascentes ; 

 gemmae (nondum satis evolutae) parvae, obtuso-ovatae, perulis paucis fere ni- 

 grescentibus extus vix v. paulo puberulis sed margine distincte longe fulvo-fimbriatis 

 obtectae. Folia parva, ovata v. ovato-elUptica, basi rotundata, valde asymmetrica, 

 apice sensim acuta, supra laevia, tantum costa parcissime pilosa, ut videtur laete 



1 According to Litwinow in Sched. Herb. Fl. Ross. VI. 164, t. 1-2 (1908), U. 

 turcestanica Kegel is a synonym of U. densa Litwinow, 1. c. 163 (1908). Unfor- 

 tunately I have not seen any material, but according to Litwinow the sterile 

 branchlets must be very similar to those of U. pumila Linnaeus, while the fruits are 

 up to 2 cm. long and to 1.2 cm. broad, and like those of U. foliacea Gilibert. Lit- 

 winow adds as a synonym U. campestris, var. laevis Regel in Act. Hort. Petrop. 

 VI. 477 (1879), quoad plant. Turcestanicam. 



