96 JOURNAL OF THE [October, 



abundantly supplied with stomata, and thus are the physiologi- 

 cal equivalent of the comparatively greater leaf-surface of other 

 plants. 



The flowers are borne in a several or many-flowered spike at 

 the summit of the stem ; this spike consists of scales or glumes 

 more or less densely imbricated, and behind each scale is a 

 perfect flower, with the occasional exception that the lowermost 

 scales are empty ; the flower consists of a perianth composed of 

 from three to nine bristles, which are generally barbed down- 

 wards, and differ greatly in length in the different species, in 

 some being entirely absent and the flower thus quite incomplete ; 

 in most species there are normally three stamens, but this is a 

 variable character, and of but little value in classification, 

 for there may be fewer in a very large number of species ; these 

 stamens have flattened filaments as a general rule, a feature 

 hared by many other Cyperacese ; there is a single pistil with a 

 simple style which divides near its apex into either two or three 

 stigmatic lobes, and this is so good a character as to be taken as 

 the mark of primary division of the genus into the subgenera 

 Eleogenus with two-cleft style, and Eueleocharis with three- 

 cleft style ; it is further found that almost invariably the former 

 have flattened or lenticular achenia, while those of the latter are 

 triangular ; the surface of the achenium is smooth in some 

 species, ribbed in others and again cancellate in others and 

 these markings are quite persistent ; the thickened style-base or 

 tubercle, is of extremely differing forms and many species can 

 be determined at a glaace from an examination of this alone. 



On these morphological characters I have entirely depended 

 in the grouping of the species here presented. I am not 

 unaware that histological details have been invoked in the 

 classification of this natural order, and I have been particularly 

 impressed by the extremely minute and laborious researches of 

 Palla as published in Engler's Bot. Jahrb. x. 293, but as the 

 results reached by him appear to me to destroy natural alliances 

 rather than to ascertain them, I have not used the arrange- 

 ment of the fibro-vascular bundles of the stem as proposed by 

 him. nor, indeed, have I found it necessary to invoke it. 



In point of geographical distribution the species are most 

 abundant in our southeastern states. Of the ^6 species 24 

 occur in Florida, and Texas is almost as well supplied. Several 



