RHAMNACEAE. — RHAMNUS 251 



circiter 6 mm. crassus; pyrenae 2, an dehiscentes? ; semina 5-6 mm. longa, olivacea, 

 laevia, nitida, obovato-oblonga, dorso convexa, lateraliter sulco angusto fere ad 

 apicem extenso aperto v. clauso, cartilagineo-marginato instructa. 



Yunnan: Mi-16, mountain forest, A. Hmry (No. 1002i; type); Mengtze, l^ 

 mountain forests, alt. 2200 m., A. Henry (No. 10021*). 



This variety may represent a good species. But I have not yet seen shoots 

 with fully developed leaves, which may have a different shape. The t>T)e has 

 rather broadly ovate leaves on such shoots, and its seeds show a broader furrow. 

 No. 10021^ is much more glabrous than the type of the variety but agrees otherwise 



exactly with it. 



There is another specimen collected by Henry near Szemao, W. mountains, 

 2000 m. (No. 11890; shrub 2.5 m., yellow flowers). It bears young leaves and 

 pistillate flowers, which have rather narrow and acute sepals, small lanceolate 

 petals and reduced stamens. The ovaries show a deeply two-cleft style. 



Another of Wilson's specimens from western Hupeh, May and October 1907 

 (No. 410) resembles var. milensis. It consists of flowering and fruiting branches. 

 The male flowers agree well with those of Henry's No. 10021, but the fruits have 

 a shorter furrow, opened only at the very base. All these forms need further 

 study. 



32. Rhamnus virgatus Roxburgh, FL Ind. II. 351 (1824). — D. Don, Prodr. 

 Fl Nepal. 190 (1825). — Schneider, III. Handb. Lauhholzk. II. 285, fig. f-h (1909). 



Rhamnus dahuricus, Lawson in Hooker f., Fl, Brit. Ind. I. 639 (1875), ex- 



cludendis synonym et varietate. 



British India: western Himalaya and the western peninsula. 

 This seems to me a good species, which may be nearer related to E. iieinophyllus 

 Schneider than to any other Chinese species. 



33. Rhamnus hirsutus Wight & Amott, Prodr. Fl. Ind. I. 165 (1834). 



Rhamnus dahuricus, var. hirsutiis Lawson in Hooker f., Fl. Brit. Ind. I. 639 

 (1875). 



British India: western peninsula. 



This species is certainly most closely allied to R. virgatus Roxburgh, and may 

 possibly represent only a variety of it. It differs in its more hirsute pubescence, 

 but the name is somewhat misleading, because the pubescence seems to be slight 

 and soon disappears. I have not seen a type specimen. 



34. Rhamnus japoaicus Maximowicz in Mim. Acad. Set. St. Petersbourg, s6r. 

 7, IV. No. XI. 11, fig. 52-64 (Rhamn, Or.-As.) (1866). — Schneider, III Handb. 

 Lauhholzk. II. 285, fig. 196, t-u\ 197 e-k (1909). 



Rhamnus Yoshinoi Makino in Tokyo Bot. Mag. XVIII. 97 (1904). 



Japan: Hondo, " basi Ontake, in fruticetis," July 1905, U, Faurie (No. 

 6873). 



Of R. Yoshinoi I have not yet seen a specimen, but accordmg to the descrip- 

 tion it may be the same as R. japonicus, which Makino does not mention. The 

 seeds have a closed groove. This is not the case in R. Buergeri Miquel (in Ann. 

 Mus. Lugd.-Bat. III. 31 [1867]) which I only know from the description. 



35. Rhamnus davuricus Pallas, Reise Russ. Reich. III. append. 721 (1776). 

 Schneider, III Handb. Lauhholzk. II. 287, fig. 192 m-p (1909). 



Rhamnus catharticu^, y. davuricus Maximowicz in MSm. Acad. Sci. St. Piti 

 boura, s^r. 7. IV. No. XI. 9 (Rhamn. Or.-As.) (1866). 



