438 CORRECTIONS AND ADDITIONS TO VOLUME I 



Berberis Julianae (p. 360). As stated above, Purdom's No. 7 should be re- 

 moved from the enumeration of specimens to which should be added Wilson's No. 

 2878, mentioned on p. 362 under B. Bergmanniae Schneider (see below), and 



A. Henry's No. 1458 from Hupeh. B. Julianae seems to be the hardiest of the ever- 

 green Chinese Barberries growing in the Arnold Arboretum. The remark under 



B. Sargentiana Schneider on p. 360 should be transferred to B. Julianae. The true 

 B. Sargentiana is also cultivated in this Arboretum and may prove as hardy and 

 useful as the other species. 



Berberis Bergmanniae (p. 362). As already stated, Wilson's specimen No. 2878 

 from Hupeh has to be transferred to B. Julianae Schneider. The differences men- 

 tioned by me regarding the glands of the petals seem to be due to a condition caused 

 by the process of drying. I have made a new investigation and I find that other- 

 wise No. 2878 is very much Uke the type-number 417 of B. Julianae. B. Berg- 

 manniae, of which No. 2877 should be regarded as the type, may be distinguished 

 by its coarser more distant serration and the glossy under surface of its leaves and 

 by its somewhat shorter and thicker fruits, which are without the style about 

 8-9 mm. long and 6-7 mm. thick, while in B. Julianae they are more elliptic-oblong 

 and about half as thick as long. This species is confined to western Szech'uan. B. 

 Bergmanniae, var. acanthophylla Schneider may perhaps better be referred as a 

 sjmonym to the type. The leaves of young plants of this form have a strong 

 distant and sinuate serrature. B. Bergmanniae is closely related to B. Julianae 

 and the other species with thick leaves and with pruinose fruits bearing a short 

 style, and has apparently little connection with B. pruinosa Franchet, as I 

 remarked on p. 362. 



Berberis Veitchii (p. 363). The quotation should read "acuminata J. H. 

 Veitch, Hortus Veitchii, 391 (non Franchet) (1906)." To the description add: 

 Flores extus rubescentes; ovula pleraque 3-4. Fructus late elhptici, nigri sed 

 pruinosi, ad 9 mm. longi et 7 mm. crassi, stigmatibus sessiUbus coronati. This species 

 may be most closely related to B. Gagnepainii Schneider, from which it is easily 

 distinguished by its bronzy yellow flowers, its reddish and smooth young branch- 

 lets and by its different habit, forming a loose shrub with gracefully arching 

 branches, while B. Gagnepainii forms a densely branched shrub. 



Berberis Griffithiana (p. 364). Remove from enumeration of specimens Wil- 

 son's No. 535 in Herb. Hofmuseum, Vienna, which is the same as the specimen in 

 Herb. New York Bot. Garden and belongs to B. Julianae Schneider, B. Griffithiana 

 being entirely absent from China. 



Berberis Delavayi (p. 364). Wilson's No. 2879 which I doubtfully referred to 

 this species represents a new species which may be described as follows: 



Berberis silvicola Schneider, n. sp. 



Frutex semiprostratus, 0.3-0.6 m. altus; ramuli juniores ignoti, vetustiores 

 cinerei, striato-angulati v. teretes; internodia 3.5-5 cm. longa; spinae valde 

 reductae v. trifidae, flavae, graciles, mediae ad 1 cm. longae, subteretes. Folia 

 sempervirentia, tenuiter coriacea, 3-5 fasciculata, elUptica v. elUptico-oblonga, 

 basi in petiolum vix ad 3 mm. longum satis subito attenuata, apice breviter acutata, 

 mucronulata, margine distincte et satis anguste graciliter spinoso-serrata dentibus 

 subpatentibus saepe inaequilongis 0.5-1.5 mm. longis 4-5 pro 1 cm., 2-4.5 cm. 

 longa, 1-1.8 cm. lata, utrinque fere concoloria, viridia, subnitentia (?), supeme 

 fere enervata, subtus subdistincte laxe reticulata. Flores 1-5 fasciculata, lutei 

 (an extus pubescentes?), circiter 1 cm. diam. pedicelUs 5-7 mm. longis graciUbus 



