MB. C. B. CLABKE ON INDIAN SPECIE8 OF CTPEBUS. 8 



to produce stolons or seen with a woody rhizome ; though it is 

 not improbable that if prevented flowering, by a jungle-fire or 

 otherwise, just when shooting for bloom, they might take on a 

 different habit, and live to flower in the next season. 



At each node of the rhizome there is normally a scale, often 

 1-2 centim. long, usually slate-coloured or black-brown, almost 

 scarious in texture, lanceolate or sometimes almost ovate, some- 

 times almost linear. In the figures which I give I have removed 

 these, partly because the essential features of the rhizome can be 

 more distinctly shown, partly because I am not artist enough to 

 draw them properly, and can only attempt pictures virtually 

 u diagrammatic." 



Figs. 20 and 21 represent two states of the rhizome of C. Icevi- 

 gatus, Linn., which may be seen in one plant at Suez : on the dry 

 sandy bank the nodes of the rhizome are very short, very tough, 

 the plant rigid, with white spikelets usually (fig. 20) ; but directly 

 the same rhizome touches the brackish pools adjacent, it com- 

 mences to run freely (as in fig. 21), the rhizome becomes supple, 

 the spikes become generally chestnut-purple. When the same 

 rhizome floats out on the sweet-water canal, the culms become 

 dwarfed, with often only 1 or 2-3 spikelets. 



Boeckeler has founded two new species (G. viridulus and G. rep- 

 tans) on Peruvian plants which I have seen, and which differ 

 from G. Icevigatus only in habit, and in habit scarcely more than 

 C. Icevigatus varies in itself. 1 doubt whether they may not be 

 merely geographic forms of G. Icevigatus. At all events, both 

 from the example of G. Icevigatus and from many other well- 

 known species, I think it is to be inferred that a difference in 

 rhizome alone (supposing the plant not annual) is insufficient 

 ground for specific separation. In G. umbellatus, Benth., in the 

 type form alone, we have caespitose fibrous roots, or a short thick 

 woody rhizome, or long slender stolons becoming ultimately 

 tough black slender rhizomes. 



I propose to separate specifically all the annuals from the 

 others. The only difiiculty arising is in the case of C. Easpan. 

 This has typically an elongate rhizome, the result of stolons 

 (fig. 23) ; but we frequently have, as Boeckeler says, the rhizome 

 exceedingly short (fig. 24); and beyond this we have flowering and 

 fruiting examples, apparently annuals, with no trace of a rhizome. 

 I must allow these, though I separate specifically as C. flavidus, 

 Betz., a large portion of Boeckeler's G. Easpan, which indeed he 



b2 



