12 MB. C. B. CLABKE ON INDIAN SPECIES OF CYPEEFS. 



culm produces 1-5 spikelets only. The number of spikelets on 

 one culm of G.platystylis varies from 1300 down to 5 in the series 

 of examples in the Calcutta herbarium. In G. pilosus the umbel 

 varies from a capitulura 3 centira. in diameter up to a long-rayed 

 umbel 3 decim. in diameter, and the spikes themselves vary so 

 greatly in density that numerous species have been founded there- 

 on. The degree of development of the umbel would therefore 

 be a character of little value, but that there is a great number of 

 species in which the umbel is apparently simple, and never (or 

 most rarely) in these varies so as to be clearly compound. Yet 

 in the species best defined as to their umbels variation occurs, so 

 that both Kunth and Boeckeler avoid as much as possible " umbel 

 simple," " umbel compound ; " though as a description of the 

 general appearance and average character of the inflorescence 

 such terms are short and convenient. Thus in G. pumilus, Linn, 

 (i. e. 0. nit ens, Vahl) the umbel has nearly always the definite 

 appearance of being simple, with rays carrying heads at their 

 extremities (or, as in all such inflorescence, the umbel may be 

 reduced to a capitulutn) . In numerous examples the simplicity 

 of the umbel appears absolute ; but in those having heads of 

 numerous spikelets it is easily seen that these heads are not 

 simple spikes, but have their axis divided. In very fine speci- 

 mens a green bract 1 centim. long will sometimes be found under 

 the heads ; the presence of such a bract (homologous with the 

 main bracts of the umbel) is here (as I believe in all species) 

 a sure indication that the apparent head is really an umbel. In 

 a specimen I have now before me, 0. pumilus has distinctly a 

 compound umbel, the secondary rays attaining 15 millim. in 

 length. 



In this genus the distinction in the number of subdivisions 

 of the axis before we come to the flower, a distinction held of 

 such high value by Eichler &c. in Kobresia and Cares, is, from the 

 nature of the inflorescence, utterly illusory. In G.polystachyus, 

 var. ferruginea (fig. 27), the umbellule is laxly corymbose, the 

 lower branches manifest, bearing 4-2-1 spikelets ; and from the 

 picture it is evident that the upper solitary spikelets may or 

 may not at some arbitrary point be considered as spikes of one 

 spikelet. Other primary rays on this same plant carry apparently 

 simple spikes which are clearly homologous, being = fig. 27 with 

 the two lower branches suppressed. These are treated as simple 

 spikes in ordinary descriptions ; and indeed in ordinary descrip- 



