1 2 CyperacecE. 



CXLVIIL— CYPERACE^E. 



PERENNIAL, rarely annual herbs, with the habit of grasses; 

 roots fibrous; stem terete or 3-angled, rarely branched; 1. 

 grass-like (rarely o), 3-ranked, mostly crowded at the base of 

 the stem, upper fewer, with tubular sheaths, which are more 

 or less closed, or the lower split to the base, ligule o, or a 

 short prolongation of the mouth of the sheath opposite the 

 blade; infl. of solitary, fascicled, panicled, or spicate spikelets, 

 composed of small imbricating distichous, or spirally imbri- 

 cating scales (glumes) ; fi. minute, uni- or bisexual, in the 

 axils of the glumes; perianth o, or of 6 or more hypogynous 

 bristles or scales (ovary enclosed in a utricle in Carex) ; stam. 

 1-3, fil. flattened, anth. basifixed, linear; ov. i-celled, style 

 short or long, stigmas 2-3, ovule I, basal, erect, anatropous; 

 fr. a compressed or trigonous nut; seed erect, free, embiyo 

 minute, within the base of the floury endosperm. 



In using the following key and the generic characters given of the 

 Ceylon Cyperacecs, it is absolutely necessary, for their proper understand- 

 ing, to bear in mind that the so-called spikelets of the tribe Cyperece are 

 not homologous with those of the other tribes. In the Cyperece, each 

 flowering glume of the spikelet bears one bisexual fl., and any scales or 

 bristles placed around or below the stamens are considered to be organs 

 of that fl. (disk or perianth). In the other tribes, what appear to be the 

 glumes of a spikelet are considered to be bracts of a spike, each of which 

 bracts bears in its axil a spikelet (homologous with that of Cyperece), 

 reduced to a sessile ovary, accompanied or surrounded with scale-like 

 glumes, which are not arranged upon a rhachilla, but are arranged around 

 the ovary in a certain order, which can only be ascertained by very care- 

 ful examination. In Hypolytrece, two of the scales are always external to 

 the others, placed right and left of the ov., are monandrous, concave, 

 strongly laterally compressed and keeled, the keel cihate, the other 

 scales, one or more of which are monandrous, are fiat and variously 

 placed. In Scleriece, where also what appear to be the glumes of a 

 spikelet are considered to be the bracts of a spike, the true spikelet also 

 consisting of scales, like glumes, and the male glumes are usually on a 

 separate spike. There is no trace in Scleriece of the two right and left 

 glumes of llypolytrece. Lastly, in Caricece the true spikelet is represented 

 in the male spike by stamens only, and in the fem. by an ov. enclosed in 

 a utricle, which latter is the homologue of the two right and left scales of 

 HypolytrecB. 



From the above it may be concluded that it is very difficult to apply 

 to the infl. of Cyperacecs a nomenclature at once scientific, practical, and 

 intelligible to the student without much circumlocution; I have there- 

 fore followed the simplest course of using, throughout the Order, the 

 term spikelet for both spikes and spikelets, and of scales for the glumes 

 of the true spikelets of Hypolytrece and Scleriece. — J. D. H. 



