( 284 ) 
is impossible to make out exactly which form they had inview. In 
order to find anything definite about the two forms of the capsule 
we must take de Candolle’s names based on Wallich’s plants. But 
one can imagine how such names would only increase the con- 
fusion. Therefore we propose the following systematic arrange- 
ment. 
Var. campylocarpa Hochr. nom. nov. 
Fructus lobi valde incurvati. 
AL. paniculata Wall. ex DC. loc. ett. ; Hook, Ic. ¢. 828; Progel 
M. petiolata Torr. et Gr. loc. c¢t.; Benth. in Journ. Lin. Soc. 
re 
Cnr petiolatum Gmel. loc. cit. 
Brasil. Prov. Goyaz (Gardner no. 3897) in iieeb. . Y. Bot. 
ard. 
Var. orthocarpa Hochr. nom. nov. 
Fructus lobi valde divergentes. 
AL. oldenlandioides Wall. ex DC. loc. czt.; Benth. loc. crt. 
AL. petiolata Progel in Mart. doc. czt., non Tort: et Gr. 
(AL. taconspicua Zoll. et Mor. Verz. Pl. Jav. 55, fide Benth.) 
Var. intermedia Hochr. nom. nov. 
Fructus lobi partim incurvati, i. e. forma intermedia. 
Ophiorrhiza Jfitreola L. sensu stricto, quoad specimen Linneanum. 
Except for the forms of the capsule, I could not find any other 
constant character separating the varieties of Jtreola. 
In regard to distribution, the var. campylocarpa is more fre- 
quent in Brazil, but both of the extreme forms may be met with in 
India as well as in America. 
OLEACEAE 
Osmanthus vaccinioides Hochr. comb. nov. 
Notelaea vaccintotdes Schlechter in Engl. Jahrb. 39: 230. 1906. 
Cap oo Noy. Caledon. (Deplanche herb. no. 323), in 
Herb. N. Y. Bot. Gard. and Kew. 
After examining the figure given by Schlechter I do not hesitate 
to identify our specimen with it, but I must object that the plant 
has a quite gamopetalous corolla, the lobes of which are distinctly 
imbricate. That is why I feel bound to refer this species to the 
