442 CORRECTIONS AND ADDITIONS TO VOLUME I 



Wilson's and Purdom's plants belong to the same species, but Fedde's description 

 agrees well enough with the specimens before me. It was drawn from a flowering 

 plant only with probably rather young flowers because the author says: " flores 

 circiter 3 mm. diamet." The living plants in this Arboretum look somewhat dif- 

 ferent from those of the preceding species, forming a more spreading, loosely branched 

 shrub. The leaves are mostly narrowly elUptic and acute, but those of the vigorous 

 young shoots are much more obtuse and broader, and more distinctly whitish 

 beneath. The inflorescences are variable, being shorter and almost fasciculate- 

 racemose at the top of the branches and longer and more or less paniculate at the 

 base. The fruits have a very short, sometimes very indistinct style. It needs further 

 investigation to determine if B. Dielsiana represents a good species or only a 

 variety of B. Henryana. Wilson's plant from western Szech'uan when fully known 

 may even represent another distinct variety. 



Bekberis Feddeana Schneider in Bull. Herb. Boiss. ser. 2, V. 665 (1905). 



fBerberis vulgaris Fedde in Bot. Jahrb. XXIX. 341 (1900). 

 fBerberis Henryana Schneider, 1. c. VIII. 262 (1908), quoad specim. Wilsonii 

 No. 3151. 



Western Szech'uan: northeast of Tachien-lu, thickets, alt. 2800-3100 m., 

 July 9, 1908 (No. 2862; bush 1.8-2.4 m. tall; flowers yellow). 



Wilson's specimen consists of old branches with very straight, apparently up- 

 right, 4-13 cm. long inflorescences which in the somewhat verticillate arrangement 

 of the rather small flowers are similar to those of B. dasystachya Maximowicz. 

 The leaves are almost entire or very indistinctly serrate. The type, A. von Ros- 

 thorn's No. 2044, was coUected in Nanchuan, southeastern Szech'uan, and has the 

 same kind of inflorescences the flowers of which had fallen, and the fruits are un- 

 known. The leaves are similar in shape but distinctly serrulate. I described the 

 racemes as pendent, but probably they are upright. The flowers of Wilson's speci- 

 men are about 5-7 mm. wide and similar to those of B. dasystachya, and I sup- 

 pose that B. Feddeana may rather belong to the following than to the section 

 Vulgar es. 



Here may be added the following species representing in my opinion a distinct 

 section: 



Sect. Dasystachyae Schneider, sect. nov. 



Berberis, sect. Vulgares, subsect. Dasystachyae Schneider in Mitt. Deutsch. 

 Dendr. Ges. XIV. 1905, 118 (1906), sensu stricto. 



Plantae glaberrimae. Inflorescentiae iis sectionis Brachypodae similes, sed erec- 

 tae, magis laxiflorae, floribus minoribus, pedicellis longioribus. 



Berberis dasystachya Maximowicz, in Bull. Acad. Sci. St. Petersbourg, XXIII. 

 308 (1877); in Mel. Biol. IX. 711 (1877); Fl. Tan.gut. 30, t. 7, fig. 1-7 (1889).— 

 Hemsley in Jour. Linn. Soc. XXXIII. 3 (1886). — Schneider in Bidl. Herb. Boiss. 

 s6t. 2, V. 664 (1905); 1. c. VIII. 262 (1908). — Fedde in Bot. Jahrb. XXXVI. 

 Beibl. 82, 43 (1905). 



fBerberis dolichobotrys Fedde, 1. c. 41 (1905). 



Western Hupeh: Hsing-shan Hsien, woods, alt. 2100 m., June and Septem- 

 ber 1907 (No. 307; bush 0.9-1.5 m. tall, flowers yellow, fruits coral red); without 

 exact locality, A. Henry (No. 6816). Shensi: Tai-pei-shan, 1910, W. Purdom 

 (Nos. 1, 5 and 9). 



The type of B. dolichobotrys was collected on the Tai-pei-shan, and I cannot 

 detect in Fedde's description any character sufficient to separate his species except 



