ROSACEAE. — ROSA 323 



Lindley's plant differs in its more acute leaflets with more acute serratures, which 

 are more hairy on the under side and in its shorter styles. In R. Davidii the 

 anther cells are markedly divergent at the base, which is not the case in the Hima- 

 layan specimens of R. macrophylla we have seen. 



Our No. 1060 has a very large much-branched inflorescence, and the venation 

 on the under side of the leaflets is very prominent; in No. 1063 the leaflets are more 

 silkily pubescent on the under side than is usual in the species. 



Rosa Davidii, var. elongata Rehder & Wilson, n. var. 



A typo recedit foliis majoribus, ad 7 cm. longis subtus glabris v. 

 pubescentibus, corymbis 3~7-floris, fructu elongato oblongo utrinque 

 plus minusve attenuato 2-2.5 cm. longo et circiter 1 cm. diam. 



Western Szech'uan: Mupin, thickets, alt. 2000-2600 m., No- 

 vember 1908 (No. 1126, type; bush 3 m.); Wa-shan thickets, alt. 1600- 

 3000 m,, June and October 1908 (Nos. 1099, 1114, 1178; bush 3-5 m., 

 flowers rose-pink, fruit scarlet to orange-red). 



This variety is distinguished from the type by its fewer flowers, its larger 

 more elongated fruit and by its usually larger leaflets, which vary from nearly 

 glabrous to silkily pubescent on the under side. In No. 1114 the leaves are pubes- 

 cent on both surfaces, but seedling plants raised from this number are normal 

 in their pubescence. 



Rosa corymbulosa Rolfe in BoL Mag, CXL. t. 8566 (1914). 



Western Hupeh: Hsing-shan Hsien, thickets, 1300-2000 ra. alt., 

 November 1907 (Nos. 630, 630% seeds only, 625; bush 1.3-2 m. tall, 

 fruit coral-red to scarlet); Patung Hsien, mountains, 1600 m. alt., 

 July 1900 (Veitch Exped. No. 1438; bush 2 m. tall, flowers pink); 

 without precise locality, A. Henry (Nos. 6491, 6714). 



This Rose is fairly common in thickets on the mountains of western Hupeh. 

 It is easily recognized by its smooth shoots, and by its ver>' membranous leaflets 

 which are gray and puberulous on the under side and turn a deep vinous-purple in 

 the autumn. The wild plants have usually much smaller corymbs than that 

 figured by Rolfe, indeed the flowers are often soUtdry. 



■Rosa setipoda Hemsley & Wilson in Kew Bull Misc, Inform, 1906, 

 158. — Willmott, Gen, Rosa^ I, 173, fig. 55 (1911). — Rolfe in BoL 

 Mag, CXL. t. 8569 (1914). 



m 



aspect from Lmdley's type. Whether or not the real R. niacrophylla Lmdley 

 occurs in China is problematical, but certainly its nearest ally is the Rose we refer to 

 R. Davidii Crepin. Several of the Chinese Roses considered by botanists to be R. 

 fnacrophylla Lindley are probably distinct species, but without having seen the 

 specimens it is impossible to refer them to their proper species. To us it is obvious 

 that in the mountainous parts of China there is a group of Roses of which the 

 Himalayan R. macrophylla Lindley may be said to represent one extreme form. 

 The members of this group possess certain distinctive characters by which they may 

 be recognized. In the present imperfect state of our knowledge it appears to us 

 best to regard them as distinct species. 



