OF ARTS AND SCIENCES. 335 



compressis usque ad apicem acutum adnatis (in sicco subulatis) anthera 

 brevioribus ; polliniis ovalibus juxta medium stipiti brevi flexuoso af- 

 fixis adsceudentibus. — Ovolau, Feejee Islands. Stigma depressed. 

 Immature follicles smooth, short, acuminate-rostrate. Probably this is 

 a congener of Endlicher's Hyhanthera higlandidosa, the pollen-masses 

 of which are probably not so pendulous as is represented. The struc- 

 ture of the andrcecium is very similar, but the coronal appendages are 

 transversely dilated at the base, thence gradually tapering to an acute 

 summit, the whole perfectly adnate to the back of the anther. In Dr. 

 Wight's Iphisia {T. Iphisia and T. Govanii, Decaisne) I find the 

 same structure, the fleshy appendages equally adnate and laterally 

 compressed. 



Gtmnema subundum (sp. nov.) : volubile, undique glabellum ; fo- 

 liis membranaceis ovato-lanceolatis seu ovato-oblongis basi rotundatis 

 vel subcordatis ; pedunculis petiolum adasquantibus ; umbella saepius 

 bifida ; corolla rotata 5-partita imberbi squamulis fere obsoletis sinubus 

 instructa ; gynostegio brevissimo. — Mountains of Muthuata, Feejee 

 Islands. — To Gymnema both Gongronema and Bidaria must doubt- 

 less be restored. The estivation of the corolla, said by Blume to be 

 valvate, is convolute, as described by Decaisne, in all the species I have 

 examined, but in most of them the margins so slightly overlap that the 

 EBstivation might readily be taken for valvate. 



Gymnema stenophyllum (sp.nov.) : fruticosum,erectum (3-6-ped.), 

 ramosissimum, fere glabrum ; foliis coriaceis linearibus basi attenuatis 

 marginibus revolutis, costa subtus pilosula ; pedunculis axillaribus bre- 

 vissimis ; corolla rotata alte 5-fida inappendiculata, lobis extus glabris 

 intus tenuiter barbatis ; gynostegio brevissimo ; poUiniarum stipitibus 

 gracilibus spiraliter contortis. — Feejee Islands, on the barren upland 

 of Muthuata. This has recently been collected by Dr. Seemann (no. 

 322), who has obtained it with young follicles. These are slender, 

 almost as much so as the leaves, and smooth. The pollinia accord 

 with the character of Sarcolobus, R. Br., but they are not "apice late- 

 raliter pellucidte," as Miquel has it. Dr. Seemann takes this for a 

 new genus, and indeed, as the genera are arranged by Decaisne, it 

 does not accord throughout with either Bidaria, Gongronema, or Gym- 

 nema proper, while the erect habit is also peculiar. But if the two 

 former genera be restored to Gymnema, the present plant could not 

 well be excluded. 



HoYA bicarinata (sp. nov.) : scandens ; foliis glabellis subcarnosis 



