ME. C. B. CLARKE ON INDIAN SPECIES OP CYPEBFS. 17 



ripe, by a transverse line ; and in a limited number of species the 

 wing, at nearly the same early period, separates from the rhachilla 

 on which it is decurrent, and thus drops off freely : i. e., in the 

 words of cyperologists, the wing is soluble, not persistent. The 

 soluble wing is often discoloured, yellow or reddish, and thus pro- 

 minent, and is the character which separates the Papyri of Kunth 

 and Boeckeler from the Exaltati. In the Exaltati, however, I 

 find sometimes a very small linear (but discoloured unmistakable) 

 soluble wing. Also in C. tegetum, and some neighbouring species, 

 the solubility of the wing appears to vary greatly in the same 

 species, and the wing differs altogether in degree and character 

 in species placed close together by authors. The solubility of the 

 wing of the rhachilla is a useful character, but one concerning 

 which I can draw no hard-and-fast line ; it connects the Papyri, 

 a series of species naturally connected by many other characters ; 

 it only throws doubt on the next group, the Corymbosi, as to 

 whether they have been rightly placed together. 



The form of the rhachilla (much compressed or quadrangular) 

 is not of sectional value, i. e. a subquadrangular rhachilla occurs 

 in species from various sections of the genus, as in G. globosus, 

 G. Monti, &c. ; but it is very useful in separating some jumbled 

 species, e. g. the quadrangular rhachilla of G. alopecuroides, Rottb., 

 distinguishes it from G. alopecuroides of Roxb. and Boeckeler, and 

 the shape and markings of the notches in 0. alopecuroides, Rottb., 

 support its otherwise evident affinity to G. Monti. The shape of 

 the notches of the rhachilla, which vary from quadrangular (or 

 even broader than long) up to narrow-oblong and linear-obovate, 

 is a character rarely used, for it is (in the main) involved m the 

 character "glumes crowded" or "glumes remote ;" but, in the 

 case of critical species, it may be well used, not as a mere repeti- 

 tion, but as capable of much greater defiuiteness and of more 

 additional detail than the ordinary " glumes subremote." 



The later groups of Gyperus — viz.Leptostachyi,Jlforw0«*, Dicli- 

 dium—diffeT from all the rest, not so much in the number of the 

 flowers or the persistence of the wing of the rhachilla, as in the per- 

 sistence of the glumes, which do not fall off by a clean disarticula- 

 tion from their decurrent bases. In the Leptostachyi the spike- 

 lets fall entire from the papilliform disks on which they are seated 

 before the glumes fall from the rhachilla, usually leaving the two 

 lowest barren glumes still attached, or leaving only the papilliform 

 disk bare ; while the rhachilla is very narrow. Mariscus is nearly 



LINK. JOUBN. — BOTANY, VOL. XXI. C 



