IQQ PHANEROGAMIA. 



as in the genus. Capsule conoidal, with a somewhat truncate and 

 umbilicate apex, about 3 lines high and 3 lines broad at the base, 

 where it is girt by a saucer-shaped, or at length flat and shield-shaped, 

 membranaceous dish (larger than the calyx), to which the carpels are 

 adherent, and from which they tardily separate at maturity. The 

 scarious circular disk, which thus persists after the carpels have 

 fallen, is entire, or very obscurely crenulate, often margined with 12 

 to 16 short setaceous processes, which are a portion of a dorsal nerve 

 of each carpel, torn away at the separation : in the centre it bears the 

 columnar persistent axis, which is moderately dilated and umbilicate 

 at the apex. Carpels 12 to 16, separable, and at length falling away 

 entire, flatly compressed, broadly semiovate (the straight side ex- 

 ternal), membranaceous in texture; the sides reticulated, acutely two- 

 margined all round; the ventral edge bilamellar and scarious; the 

 back moderately sulcate; the apex wingless, barely pointed, rather 

 than appendaged, with two very small and obtuse flat processes or 

 teeth : the dehiscence is from the apex between the teeth, extending 

 downward first along the ventral, then along the dorsal suture, at 

 length two-valved. Seed resupinate-suspended, triangular-obovate, 

 compressed, conformed to the cell, which it fills. Radicle superior, 

 straight : cotyledons incumbent and folded. 



The crest, or pair of apical wings, from which the genus Gristaria 

 takes its name, varies greatly in size, compared with the body of the 

 carpel, in different species. In the present plant it is reduced to a 

 minimum, or is, in fact, obsolete. Yet I do not hesitate to refer it to 

 this genus, with which it otherwise accords in structure, and espe- 

 cially in the remarkable scarious disk. 



5. GAYA, KB. K 

 1. Gata subtriloba, K B. K. 



Gaya subtriloba, H. B. K. Nov. Gen. & Spec. 5, p. 270, t. 476. 



Hab. Obrajillo, Peru. 



Our specimen, a miserable one, accords well with No. 3236 of Mat- 



