on the Hortus Malaharicus, Fart III. 107 



carnosiuscula. Petiolus linearis, anceps, brevis, glaber, non 

 stipulaceus. 



Flores magnitudine nucis moschatae pedunculati, cernui, her- 

 bacei, glabri : latérales axillares solitarii, terminales sub- 

 terni. Fedimculi folio dimidio breviores, penduli, unitlori, 

 nudi, glabri, apicem versus incrassati, angulati, articulis 

 1 — 3 divisi. 



Calyx inferus, monophyllus, coriaceus, crassus, campanulatus, 

 ultra medium quadritidus laciniis ovatis patulis acutis, fundo 

 tectus disco crasso integro staminifero ultra calycis divi- 

 sionem producto. Filamenta plurima, indetinita, linearia, 

 longitudine calycis incurva, ad disci marginem inserta. Aii- 

 therœ parva?, cordataî. Germai turbinatum. Sti/liis teres, 

 staminibus longior, medium versus angulo duplice Hexuosus. 

 Stigma maximum, supra convexum, subtus concavum, pilei- 

 forme. 



Fomum? orbiculatum, depressum, calyce patente ad basin cinc- 

 tum, stylo persistente mucronatum, septis tenuibus carnosis 

 in loculos circiter octo, putamine lignoso extra cinctos, ob- 

 solete divisum. Semina angulata, in pulpo nidulantia. 



Panitsjika maram, ^9. 45. tab.-M. 



Commeline in his annexed note considers this as the same 

 with the Janipaba of Piso, a plant of Brazil, which Linnicus 

 called Genipa, but by Willdenow joined to the Gardenia. Whe- 

 ther or not this last arrangement be proper I need not here in- 

 quire, because the Janipaba is no doubt a plant of the order of 

 Rubiacece, with opposite leaves, while the leaves of the Panits- 

 jika are alternate. Plukenet, although not aware of Commeline's 

 error, judged more soundly concerning the affinities of the P«- 

 nitsjika; for in treating of the Fishamin of Virginia {Aim. 180.), 

 the Diospi/ros virginiana of Linnœus, he says {Mant. 99-), "Inqui- 



p 2 rendu m 



