154 MR. C. B. CLARKE ON INDIAN SPECIES OP CTPERUS. 



Radix (in speciebus gerontogeis typicis) perennis, rhizomate 

 repente, stolonibus saepe additis ; in speciebus araericanis (hoc 

 loco dubie locatis) annua aut iibrosa. Culrai trigoni vel teretes. 

 Folia evoluta, aut in paucis abbreviata vel subnulla. Umbella 

 saepius laxa aut aperta, radiis longis aut conspicuis ; spicul33 in 

 pluribus numerosas, spicatae (in paucis americanis globoso-fasci- 

 culataV), compress83, multiflorae, saepius acutae. B-hachillae alaa 

 saepius conspicuae, persistentes aut solubiles. Grlumae naviculares 

 saepius muticae, rarius breviter mucronatae. Stamina 3. Styli 

 rami 3, saepius elongati. Nux saepius ellipsoidea, trigona, cum 

 g-| parte glumae aequilonga, nee minima nee maxima. 



This group here contaius all the species of Eucyperus which I 

 have not been able to get into the preceding sections. The Old- 

 World species form a tolerably natural group, except C. stoloni- 

 ferus and G. inacer, which perhaps should each have a section to 

 itself. The American species, though some of them brought here 

 by all authors, appear to me to differ considerably ; C. sphacelatus 

 with its annual root and narrowly-winged rhachilla is very near 

 C. compressus. I have brought here C. strigosus and C. congestus, 

 two common species placed by Boeckeler in a section called Fas- 

 ciculati, which from its character differs from the Corymbosi in 

 " spiculae saepissime disco impositae ; " but neither C. strigosus 

 nor C. congestus is said (in its own description) to have such 

 discs. 



I hold that two very different things are included as discs by 

 Boeckeler. In many species of Cyperus where the spikelet is at 

 all oblique in its axis, in the dried state the rhachilla contracts 

 very near the base below the lowest glume: here it seems to sit 

 on a small cushion, but it is very firmly attached, and does not 

 disarticulate at this point. This appearance (often only a result 

 of drying) is not uncommon throughout the genus, as in C. poly- 

 stachyus, where it is often prominent ; and I estimate it as of no 

 classificatory importance. 



But in a number of species of Cyperus (Mariscus for me) above 

 the two lowest barren glumes, the rhachilla of the spikelet dis- 

 articulates by a clean cut, leaving a small disc or cushion 

 within the two barren glumes : this character may, I think, be 

 made of the greatest use in a natural classification of the 

 species. 



I have divided this large section into three " series " for conve- 

 nience ; but the series pass into one another at various points. 



