230 ON THE ASCLEPIADE^. 



has been more difficult to determine than this : it is, how- 

 ever, perfectly natural and distinct, and is evidently the 

 66] connecting link between the true Asclepiadese, and the 

 Periplocese, which follow. I have examined five species, of 

 which two are pubhshed plants, viz. Periploca secamone 

 Linn, and emetica of Betz. and Willdenoto . The third is a 

 climber, a native of India, discovered by Dr. Roxburgh ; 

 and two are erect shrubs, natives of the tropical part of 

 New Holland. 



III. PeHIPLOCEtE. 



Massa pollinis 5-20, granulosse, (granulis e sphserulis 4 

 compositis), solitariee, usque quaternatim apici dilatato 

 corpusculi singuli stigmatis affixse. 



Filamenta partim v. omnino libera. 



HEMIDESMUS. [Periploca sp. Linn. 



Char. Corolla rotata, squamis 5, sub sinubus, obtusis. 



Filamenta basi connata, supra distincta. AntliercB co- 

 hserentes, a stigmati liberse, imberbes, apice simplices. 



Masses pollinis 20. 



Stigma muticum. 



Folliculi cylindracei, divaricatissimi, teves. Semina 

 comosa. 



Habitus. Frutices volubiles, glabri. Folia opposita, 

 nitida. Flores interpetiolares, congesti, parvi. 



Patkia. India Orientalis. 



57] Obs. This genus, whose name is derived from the 

 partial connection of the stamina, is composed of Periploca 

 Indica, and two very nearly related unpublished species. 

 Its strict affinity to Periploca is unquestionable, but the 

 differences in the structure of its flower appear to me suffi- 

 cient to justify its separation from that genus. 



