24 Mr. Woods on the Genera of European Grasses. 



glume and of the outer palea of the abortive floret meet each in a point, 

 which in many cases is prolonged into a seta. O. Crusgalli has, I believe, 

 always a filmy inner palea to the neutral floret, but it does not appear that 

 this is constant through the genus. A comparison of Kunth's descriptions of 

 the genera Panicum and Oplismenus leaves no distinguishing mark between 

 them, except the "glumae et paleae muticse" of the former, and "palese mu- 

 cronatse et glumes plerumque aristatse" of the latter. A mark of this sort is 

 not often to be depended on among Grasses, and with the qualifying adverb 

 it seems to be a very slight distinction. This learned author is amazingly 

 careless in his divisions, whether of tribes, of genera, or of species, as to 

 whether the description contain, or do not contain, a differentia : he seems in 

 all cases to trust rather to the general habit of the plant and to his own tact, 

 than to any decided or describable character. 



Setaria. Sir J. E. Smith says, that the bristle-like involucrum is not suf- 

 ficient to form a generic character. I am glad to seize anything which sepa- 

 rates a group of plants which have a peculiar appearance. The smaller spikes, 

 where the whole is compound, have something of a one-sided appearance, but 

 I think rather from their position than from their structure. The whole in- 

 florescence is, I believe, invariably alike all round. Since there is a marked 

 habit, I am not willing to throw this genus back again into Panicum : how far 

 it may be rightly separated from Pennisetum I will not venture to decide, but 

 I have high authority for keeping them distinct. In the European species 

 the leaves are rough and the sheaths smooth, bearded at the mouth. 



I follow Kunth, rather than my own judgement, in placing Lappago among 

 the Panicece, but I do not know where else to put it. The outer glume is 

 broad-based, but small, and so thin that it is often difficult to detect it. The 

 inner is hard and prickly. This difference of structure might give some ground 

 to consider the latter as an abortive floret ; but as the thin valve is towards 

 the rachis of the spike, and therefore probably the inferior glume, analogy 

 does not support us in this view of the subject. The inflorescence is in small 

 spikes, not quite sessile, on a five-channeled rachis, each spike springing from 

 an angle and having a channel above it. A difference in the direction of the 

 parts gives to the whole the appearance of a one-sided raceme. The spiculse 

 grow all round the rachis of these short spikes ; there are two or more of them. 



