BENTHAM ON MIMOSE^. 331 



Tribe, II. Eumimose^e. 



Calyx et corolla aestivatione valvata. Stamina definita, 

 numero petalorum aequalia v. diipla, una cum petalis hypo- 

 gyna. Filamenta libera v. basi brevissime inter se et cum 

 petalis connata. Pollinis granulse numerosissimas, distinct£e. 



Subtribe, I. ADENANTHERiE. 



Antherse glandula stipitata caduca terminatae. 



Since the old genus Mimosa was first broken up by Willde- 

 now, the characters of the genera adopted by modern botan- 

 ists have been chiefly derived from the form and consistence 

 of the pod : this organ is however so seldom to be seen in 

 herbaria, and, when seen, often so doubtful as to both of 

 those characters, as to have produced much confusion in the 

 suborder. In seeking for some characters which should pos- 

 sess more fixity, it appeared to me that the stamina would 

 afford a primary division both constant and definite, and at 

 the same time easy to observe. The definite stamina have 

 always, in the species I have examined, (excepting perhaps 

 the anomalous genus Parkki.) granular pollen, and the inde- 

 finite stamens agglomerate pollen, and this division does not 

 appear to break up any genera otherwise natural. The sub- 

 division of Euminwsea', upon the presence or absence of the 

 curious stipitate gland on the anther, is perhaps more artifi- 

 cial, but is always constant in the same genus, excepting per- 

 haps one species o^ Prosopis, where I have not found it, as 

 in all others of that genus. 



With regard to the ultimate reduction to genera, I have 

 been obliged, in many cases, to follow previous authors, in re- 

 lying on the pod alone, as otherwise it would have been ne- 

 cessary to combine old Linnsean genera, which are not un- 

 natural. It must, indeed, be admitted, that in some cases 

 these distinctions are troublesome, and far from satisfactory ; 

 but too many species are as yet unsufficiently known, to enable 

 nie to propose any better system. 



