HITCHCOCK GRASSES OF CUBA. 185 



present paper the new species published in Sauvalle's article are 

 credited to Wright. The original set upon which Sauvalle's list is 

 based is at the Gray Herbarium, and a fairly complete duplicate set 

 is in the Sauvalle Herbarium. 



The sets of Wright's plants were made up at the Gray Herbarium 

 and given herbarium distribution numbers. Each number included 

 such collections as were thought to be of the same species. Thus it 

 often occurs that different specimens of the same distribution num- 

 ber may have been collected in different localities or may even bel< >ng 

 to different species. The data found upon the field labels in various 

 herbaria are mentioned under each species in the present list. There 

 is also appended a list of the species of grasses included in Sauvalle's 

 Flora Cubana, with references to their identification, and a list of the 

 Wright numbers in sequence with their identification. 



The plan followed in the present paper is to give under specimens 

 cited a list of the specimens found in the herbarium of the Estacion 

 Central Agronomica, including the Sauvalle Herbarium, and in the 

 National Herbarium, without statement as to the herbarium in which 

 they are deposited. To these are added specimens found in the Gray 

 Herbarium which do not occur in the herbaria just mentioned, and 

 finalty, specimens in the Herbarium of the New York Botanical 

 Garden (Herb. N. Y. Bot. Gard.), including the herbarium of Colum- 

 bia University, of which the Torrey Herbarium forms a part, which 

 are not found in the others mentioned. The specimens collected by 

 the staff of the botanical department of the Cuba Experiment Station 

 are numbered in a single series and are indicated in this list by the 

 letters HC (Herbarium Cubae). The data for the Wright specimens, 

 given in the paragraph devoted to the enumeration of specimens, are 

 understood to be found with the specimens in the Sauvalle Her- 

 barium. Additional data, found with specimens in other herbaria, 

 are quoted in the succeeding paragraph devoted to notes. 



Grisebach enumerated 154 species of grasses in his catalogue. 

 Sauvalle's Flora Cubana includes 170 species. The present list 

 includes 228 species or well-marked subspecies. 



KEY TO THE GENERA. 



Series Paniceae. Spikelets 1-flowered, rarely 2-flowered; when 2-flowercd the 

 terminal floret perfect, the lower staminate or neutral (except in Isachne), no apparent 

 internode between them; rachilla articulated below the glumes, the spikelets falling 

 from the pedicels entire, singly, in groups, or together with joints of an articulate 

 rachis; spikelets not laterally compressed (except in Lithachne). 

 Lemma and palea (the latter sometimes wanting in Andropogoneae) hyaline; glumes 



more or less indurated, the first largest ; sterile lemma like fertile lemma in texture 



(except in Alloteropsis). 



In this key the tribal characters are given with reference to the Cuban genera 

 only, and in some cases would not hold good for the entire tribe. 



