Leguminosae 



Bauhinia Linnaeus 

 -, ^. . Merr. JAA 21: 368. 1340 



Bauhinia caterviflora Chen, Jour. Arnold Arb. 19: 129. 1938, Lingnan 

 Sci. Jour. 18:479. 1939. -^ ? f ' . - - 



Indo-China: Tonkin, Chapa, >e7e/o/ 2165, August 1936, ait. about 

 1600 m. Yunnan, Szechuan. 



A species new to Indo-China, the cited spécimen differing from the 

 type in being much less pubescent, but falling well within the variation 

 of the species as interpreted by Miss Chen. 



Bauhinia Pierrei Gagnep. Not. Syst. 2: 178. 1912, et in Lecomte, FI. 

 Gén. Indo-Chine 2: 136. 1913. Merr. JAA 21 : 368. 1940 

 Bauhinia Hozvii Merr. & Chun, Sunyatsenia 2:243. t. 48. 1935; Chen, 

 Linjïnan Sci. Jour. 18: 476. 1939. 



1940] MERRILL. RECORDS OF INDO-CHINESE PLANTS, II 369 



Two collections from Tonkin, which I hâve identified from Gagne- 

 pain's descriptions, Pételot 2161, 2164, the latter number more pubescent 

 ^.han indicated by him, enable me to reduce the Hainan Bauhinia Howii 

 Merr. & Chun. This species is further represented by Clemens 3158 

 'from near Hue, Annam. In no essentials do the Hainan spécimens differ 

 from the Indo-Chinese form. Annam, Tonkin, Hainan. 



There is an unfortunate error in the secondary division of Gagnepain's 

 key to the Bauhinia species which probably misled us when Bauhinia 

 Howii was proposed and described as a new species. In Lecomte, FI. 

 Gén. Indo-Chine 2: 120. 1913, under group A, stamens 2 or 3, the c'har- 

 acters of sections one and two are exactly reversed. The first 21 species 

 hâve pubescent, not glabrous ovaries while the succeeding eight hâve 

 glabrous, not pubescent ones. The key, therefore, should be read just 

 the reverse of what Gagnepain states. Again under B while the char- 

 acters are correctly given in the key on page 1 20, they are exactly re- 

 versed in the center heads § III p. 122 which should read "ovaire glabre" 

 an. § IV p. 123 which should read "ovaire velu." 



