DIADELPIIIA. 337 



and have only to ascertain whether any papilionaceous 

 plant we may have to examine has ten stamens, all 

 alike separate and distinct, in which case it belongs to 

 the tenth Class, or whether they are in any way com- 

 bined, which refers it to the seventeenth. 



** Stigma downy, without the character of the pre- 

 ceding section, for this and all the following are truly 

 diadelphous. Very nice, but accurate, marks distin- 

 guish the genera, which are sufficiently natural. The 

 style and stigma afford the discriminative characteristics 

 ofOrobus, t. 1 153 ; Pisum> t. 1046; Lathyrus, t. 670, 

 1108; Ficia, t. 334, 481483; and no less deci- 

 sively in Erviim y t. 970, 1223, which last genus, not- 

 withstanding the remark in Jussieu 360, " stigma non 

 barbatum" (taken probably from no genuine species,) 

 most evidently belongs to this section, as was first re- 

 marked in the Flora Britannica ; and it is clearly 

 distinguished from all the other genera of the section 

 by the capitate stigma hairy all over-, nor is any 

 genus in the whole Class more natural, when the 

 hitherto mistaken species are removed to their proper 

 places. See Fl. Brit. 



Legume imperfectly divided into two cells, 

 always, as in all the following, without the character 

 of the preceding sections. This is composed of the 

 singular Biserrula, known by its doubly serrated fruit, 

 of which there is only one species ; the Phaca, Jacq. 

 Ic. Rar. t. 151 ; and the vast genus of Astragalus, 



7. 



