358 WILSON EXPEDITION TO CHINA 



stigmate sessili, ovulis plerumque 4 fere sessilibus instructum. Fructus 

 matunis non visus. 



Western Szech'uan: Mupin, thickets, alt. 2000 m., June 1908 

 (No. 2873). 



This species seems most closely related to Berheris verrucidosa Hemsley & 

 Wilson which is easily distinguished by the terete, densely verruculose branches, 

 by the more coriaceous leaves, with more and stronger spiny teeth and by the 

 shorter pedicels. 



At the request of Mr. Wilson I have named this species for Dr. Asmy, an emi- 

 nent surgeon in charge of the German Hospital, Chungking in recognition of 

 valued service rendered to him in the years 1908 and 1911. 



Berberis Gagnepainii Schneider in Bull. Herb. Boissier, s6t. 2, 

 VIII. 196 (1908); III. Handb. Laubholzk. II. 914 (1912). — Hemsley 

 in Gard. Chron., ser. 3, XLVL 226, fig. 96 (1909). 



Berberis acuminata Stapf in Bot. Mag. CXXXIV. t. 8185 (non Franchet) 

 (1908). 



Western Szech'uan: Wa-shan, thickets, alt. 2500-3000 m., 

 September and November 1908 (No. 1344; bush 1.30-2 m. high, 

 fruit blue-black); Mupin, thickets, alt. 2000-2500 m., June and 

 November 1908 (No. 1137; evergreen bush, 1.30-2 m. high, flowers 

 yellow, fruit blue-black) ; southeast of Tachien-lu, alt. 2300-3000 m., 

 July and October 1908 (No. 1137"; bush 0.3-1 m. tall, flowers yellow, 

 fruit black); Ching-chi Hsien, near Wa-wu-shan, woodlands, alt. 

 2600 m., October 16, 1908 (No. 2874; shrub 0.5-0.75 m. tall); cliffs, 

 July 1903, alt. 4000 m. (Veitch Exped. Nos. 3148, type, 3148=^). 



The type (No. 3148) is distinguished by its finely pustulate branchlets, in this 

 somewhat resembling B. verruculosa. In No. 1344 the branchlets also are some- 

 what pustulate. This number bears fruits, which measure about 10:6 mm. and 

 have sessile stigmas. No. 1137 is distinguished by more numerous fascicles and 

 somewhat smaller fruits, but the ovaries and fruits contain 4 ovules and 4 seeds. 

 The leaves are up to 7: 1.5 cm. long, and the petioles are 1-4 mm. (not cm. as 

 misprinted in my diagnosis) long. I think all these numbers belong to B. 

 Gagnepainii, but No. 1137 may represent a somewhat different variety, of which I 

 do not know the young branchlets. No. 1137" also has 5 ovules, but the flowers 

 are borne on very long (3-4 cm.) pedicels, and the serrature of the leaves much 

 resembles that of the leaves of B.Vdtchii Schneider. No. 2814 differs m its rather 

 large leaves (to 7: 2 cm.), some of which have a slight bloom on the lower surface. 

 The unripe fruits contain four ovules. This number may represent a distinct 

 variety. 



Berberis triacanthophora Fedde in Bot. Jahrb. XXXVI. Beibl. 82, 

 43 (1905). 



