Mr. Woods on the Genei-a of European Grasses. 49 



with regard to these characters, as to ahnost every other, we have here a tribe 

 in Avhicli they vary r Perliaps we may tine) in sonîe of the ti-opical genera of 

 Rottboelliaceœ an approach to the Paniceœ, reducing the arrangement of the 

 whole family to a circular order. I have only three genera to notice, con- 

 taining in all but six species. 



1. Nardiis. Spiculœ in two rows, on one side of a continuous rachis. Glumes 0. 



Outer palea keeled, tapering into a subulate point. Stigma I. 



2. Psilurus. Spicnlae on opposite sides of the cylindrical but deeply cliannelled 



rachis. Glume 1, small. Outer palea membranous, awned ; inner as long, 

 scariose. Stamen 1. 



3. Lepturus. Spiculoe imbedded in the cliannels of the cylindrical or prisma- 



tical fragile rachis. One-flowered, with an interior rudiment. Glumes 

 1 or 2, opposite to the rachis, and as long as the scariose paleae. 



Psilurus nardioides seems well separated from N'ardus. Schrader put it 

 with RotthoeUia ; Palisot de Beauvois called it 3Ionernia. I take it for granted 

 that Psilurus, the name given to it by Trinius, and adopted by Mertens and 

 Koch, as well as by Kunth, is the most ancient. The spicule are placed some- 

 what obliquely, and the glume is not exactly opposite to the rachis, the abortive 

 floret on one side appearing conspicuously from underneath it. 



Lepturus contains the 4 European species which were formerly given to 

 the genus RottboelUa. It has only one spicula at each joint of the rachis, and 

 this contains one perfect, and a superior imperfect floret or rudiment. The 

 true RotthoeUia lias 2 spicules at each joint, one of which is tabescent, and the 

 perfect spicula lias one perfect and an inferior imperfect floret or rudiment. 



Before concluding this essay I will offer an artificial arrangement of the 

 Grasses, founded chiefly on their inflorescence, M'hich seems to yield the most 

 distinct and definite characters, and is tlierefore best adapted to facilitate the 

 researches of the student, and enable him to determine to what genus any 

 plant under examination may belong. With the same object in view we may 

 observe, 



That the spikes are fingered in Cynodon, Dactylocteniuni, Digitaria, and in 

 some species oî Andropogon. 



VOL. XVIII. H 



