Mr. Woods on the Genera of European Grasses. 45 



will form a better generic character than the foriu of the spicuUe and the 

 number of fîorcts, both of which vary greatly in the ditiercnt species, and 

 even in the same. 



I am not at all clear that Cvnosurus onght not to be divided into three 

 genera. C. crlstatus seems to have but little relation to V. ec'ihiatits and 

 C. elegans, and C. aureus {Lamarekia aurea of Mœnch and Kunth) differs 

 considerably from both. 



The habit of Dactylis seems to be very distinct, while its keeled paleu- 

 separate it sufficiently from Festuea and their point from Poa. The crowded 

 position of t!ie spicuhe is also very characteristic. Tiiese characters unite to 

 a considerable degree in Poa litf oralis, which has been placed by AVilldenow 

 and Schrader in this genus. T\ni keel, however, does not continue to the 

 base of the tloret, and the outer glume is smaller, and certainly not more firm 

 than the inner. Kunth considers it as a genuine, but not a legitimate Poa ; a 

 distinction which I do not comprehend. The spicuke are strictly sessile on 

 opposite sides of an alternately flattened rachis, whose branches, however, 

 combine into a dense one-sided spike. 



Festuca is a genus comprehending two or three different appearances, whicii 

 are not easily united under one character, and which yet cannot well be sepa- 

 rated. The first tribe, the 3Iijgalurus, or Fulpia of Link, has the outer palea 

 gradually tapering into a very long seta, and the glumes in general extremely 

 unequal. The second, of which Festuca (n-ina may be considered as the typi', 

 has the glumes much more neai'ly equal, and the seta, where it exists, rising 

 much more abruptly from the palea. This passes almost insensibly into tiiose 

 which have t!ie point a little below the extremity of tiie palea, as F.prateu- 

 sis and F. elatior; and there is another group approaching in appearance to 

 Poa. Kunth has two more groups, one of which is my Sclerochloa, but also 

 including F. diraricata and Triticum maritiniuni, two plants which I leave 

 with Festuca. Gandin makes a division of those Festucas which have a point 

 not quite at the apex of the glume ; but in F. pratensis, which may be consi- 

 dered as the type of the division, the awn sometimes adheres to the palea riuite 

 to its extremity. 



The awn or seta in Bkomus is not in general a mere continuation of the 

 midrib, but is strengthened by the union of two lateral ribs, which usually 



