372 WILSON EXPEDITION TO CHINA 



This species appears to be most nearly related to B. Silva-Taroucana Schneider, 

 which chiefly differs in the pruinose not shining under surface of the leaves, in its 

 more irregularly racemose inflorescences and in its longer pedicels. Further in- 

 vestigations are necessary, however, to decide whether B. Mouillacana must be 

 considered as a variety of B. Silva-Taroucana. 



At the request of Mr. Wilson I have named this species for Dr. Mouillac, a dis- 

 tinguished French army surgeon, at one time in charge of the Ecole de Medecine & 

 R. C. Hospital, Chengtu, in appreciation of valued service rendered to him during 

 the autumn of 1910. 



The following description and notes are based on material not collected in the 

 Arnold Arboretum Expeditions. 



Berberis Poiretii Schneider in Mitt. Deutsch. Dendr. Ges. XV. 180 (1906). 

 Berberis sinensis Schneider in Bull. Herb. Boissier, s4r. 2, V. 655 (non Poiret, 



nee Desfontaines) (1905). 

 Berberis sinensis, /3. angustifolia Regel in Act. Hort. Petrop. II. 2, 416 (1873). 



Manchuria: Sheng-king, Mukden, May 28, 1906, F. N. Meyer, (No. 121). 



Berberis Poiretii, f. weichangensis Schneider, n. forma. 



A typo recedit spinis inferioribus 3-fidis, bracteis plerumque paullo brevioribus. 



Northern Chili: east Weichang, 1909, W. Purdorn (Nos. 2, type, and 35, 

 in part). 



I cannot find any other difference between the type and this form of northern 

 Chih, but there is another flowering branch on the same sheet with No. 35, of 

 which all the leaves are distinctly serrate, while the flowers are quite like those of 

 the other branch, which cannot be separated in any way from No. 2. The leaves 

 of the serrate form resemble perfectly those of the following species, which possibly 

 may be only a variety of B. Poiretii. 



Berberis Purdomii Schneider, n. sp. 



Frutex metralis, ut videtur habitu B. Poiretii; ramuU glabri, leviter sulcati v. 

 fere teretes, paullo divaricati, juniores fuscescentes, vetustiores cinerascentes; 

 intemodia 1.5-2 cm. longa; spinae 1-3-fidae, graciles v. majores, mediae ad 1.5 

 cm. longae, flavescentes, acutae, subtus sulcatae (in No. 345 spinae robustiores, 

 ad 2 cm. longae). Folia ad 12 fasciculata, lanceolata, apice acuta, basim versus 

 sensim in petiolum brevem decurrentia, 1:0.4 ad 4:0.8 cm. magna (in No. 345 

 folia majora, obovato-lanceolata, ad 4:1.3 cm. magna), utrinque fere concoloria 

 et pari modo satis laxe sed distincte reticulata, margine graciliter spinuloso-den- 

 tata, dentibus utrinque 0.5-1.5 mm. longis 5-15 instructa. Inflorescentiae valde 

 juveniles, cum pedunculo nudo 1.5 cm. longo ad 3 cm. longae, glabrae, flores nondum 

 evoluti bracteis lineari-lanceolatis 2 mm. longis acuminatis stipitati; ovaria ovulis 

 ut videtur 2 sessihbus instructa. 



Shensi: south Yenan Fu, 1910, W. Purdom (No. 3); same locality (No. 345; 

 3 ft. tall). 



Without having seen fully developed flowers it is impossible to say, whether the 

 branch of Purdom's No. 35 of North Chili, Weichang, mentioned under B. Poiretii, 

 var. weichangensis Schneider may not belong to B. Purdomii. This species cer- 

 tainly is very closely related to B. Poiretii, but the leaves of B. Purdomii are 

 serrate and show a more distinct nervation on both sides. 



Berberis Vernae Schneider, n. sp. 



Frutex 0.75-1.5 m. altus; ramuli glabri, satis sulcati, annotini biennesque pur- 

 purascentes, deinde cinerascentes, lenticeUis nigris plus minusve conspersi; inter- 



