Mr. Woods on the Genera of European Grasses. 19 



vations on Glyceria, seems disposed to place some reliance on the simple or 

 branched filaments which compose the stigma; but from Kunth's work, as 

 well as from what little I have been able to examine myself, I should doubt if 

 this could be trusted as a generic distinction. 



The seed in some cases is entirely detached and of itself falls out of the 

 paleœ ; in others it is inclosed by, and falls off vvith them, but without ad- 

 hering to either ; sometimes it is attached to tlie upper, and sometimes to both 

 paleœ. It is sometimes furrowed, and sometimes has an even surface ; is 

 sometimes crested, and sometimes hairless ; sometimes linear and thin, some- 

 times thick and turgid. All these particidars are useful in the formation of 

 genera, yet perhaps hardly any so constant as to prove at once a difference of 

 genus. 



The tribes I propose are the following: 



In discriminating these tribes, and the genera which they contain, I have 

 endeavoured to give such characters as will apply to the plants of Em-ope. 

 My knowledge is too imperfect to enable me to hope to seize correctly the 

 distinguishing marks of those of more distant regions. Where a character 

 occurs only in one genus of a tribe, or in one tribe among the Grasses, I mark it 

 with a note of admiration : this plan renders it unnecessary to mention under 

 the other tribes or other genera that such a particular is wanting. 



Andropogone^. 



The proper characters of this tribe seem to be, that the spiculœ are in pairs, 

 one sessile and one stalked, or, at the extremity of each branch, in threes, of 

 which two are stalked. The sessile spiculee are always perfect, containing 

 one perfect floret and the palea-like rudiment of another. The stalked spiculee 



D 2 



