20 Mr. Woods on the Genera of European Grasses. 



are barren in some genera and perfect in others. The glnines are firm, nearly 

 equal, inclosing the palese, which are thin, and often filmy, but bearing a 

 strong awn. The inflorescence is panicular, except in Andropogon, where 

 most of the species produce their spiculœ in two rows on the same side of a 

 flattened rachis ; and the spikes or racemes thus formed are frequently dis- 

 posed in pairs or digitate. Flowers in determinate sets occur among the 

 Paniceœ, which, however, are easily distinguished by the coriaceous paleae. 

 They are also found in Hordeam, and in some foreign genera among the 

 Rottboelliaceœ. E/i/mas and Lygeum liave two or three spiculte together, 

 but they off"er nothing of a systematic arrangement. 

 The genera of this tribe may be thus disposed : 



A. Spicuhe all fertile, clothed with long silky hairs. Glumes thin. 



1. Saccharion. Palea' unarmed, the inner small or wanting. Sqnamulœ 2. 



Panicle not spike-like. 



2. Imperata. Paleœ 2, unarmed. Squamuloe none. Panicle spike-like. Rachis 



abruptly flexuose, continuous. 

 ;>. Erianthus. Outer palea awued. Squamulœ 2. Panicle spreading. Rachis 

 hardly flexuose, very brittle. 



B. Lateral or stalked spiculee barren. 



4. Andropogon. Spiculae linear-lanceolate, compressed. Seeds nearly linear. 



5. Sorghum. Spiculae ovate or ovato-lanceolate, turgid. Glumes of the fer- 



tile floret coriaceous or horny. Seed roundish. 



The second and third genera were struck off"fi'om Saccliarum by Palisot de 

 Beauvois. Tlie true Saccharuin has, as it appears from this author, only one 

 palea and no awn or seta. 



Imjnrata and Erianthus differ, as stated above ; but these distinctions are 

 not sutficient to constitute genera unless supported by a marked peculiarity of 

 habit ; and where we have only one species of a genus, as in Imperata, we can 

 hardly acknowledge some difference of general appearance to decide the 

 (jucstion, since we have no proof of its having any connection with the cha- 

 racter. I may add to the foregoing description of Imperata, that in all my 

 specimens the styles are united in the lower part. The very tough and horny 



