96 ME. C. B. CLAEKE ON INDIAN SPECIES OE CYPEEIJS. 



G. usitatus, Burchell, is very close to C.jeminicus, and belongs 

 to the Corynibosi. 



0. proteinolepis, Sfceud., has woolly roots, and belongs to the 

 Sect. Galeria ; indeed, it is hardly possible to extricate it speci- 

 fically from O. glomeratus and G. curvulus. 



C. babakensis is, I think, only one of the endless varieties of 

 G. pilosus ; and, if a subspecies, must go with G.pilosus. 



G. obtusiflorus, C. margaritaceus, and G. niveus differ from my 

 type of Compressi in their stramineous or cinnamomeous tinge, 

 in their capitate inflorescence, very closely packed concave obtuse 

 glumes, and broad filaments — in a word they belong exactly to 

 my type of Arenarii. 



I take 0. compressus as the type of my Compressi, and (omitting 

 the two doubtfully placed here, C. Griffithii and G. rubicundus) 

 the half-dozen species collected under Compressi are pretty 

 clearly allied to it. But it is much more difficult to say why 

 various other species should not be brought here also. The Com- 

 pressi are one of the groups that touch the Corynibosi. Even 

 C. sphacelatus itself, with its annual root, thin umbel, and green 

 colour, seems to me exceedingly near O. compressus, much neaTer 

 to it indeed than to C. articulatus, which stands at the head of 

 the Corymbosi. The only reason for not placing C. sphacelatus 

 among the Compressi must be the wing of the rhachilla, which, 

 though narrow and inconspicuous, is oblanceolate on each notch 

 of the rhacheola in a manner more characteristic of the Corymbosi 

 than of the Compressi. Still my best reason for not placing 

 G. sphacelatus among the Compressi is that I find it among the 

 Corymbosi. 



Among the Glomerati several differ from the Compressi by 

 hardly any character except that the rays are shorter, the spikes 

 longer in proportion, giving the inflorescence a denser character. 

 C. strigosus even hardly differs in laxness of spikes, and the wing 

 of the rhachilla is very narrow ; while the small group Grlutinosi 

 differ only by trivial characters. 



The difficulty thus is not to pick out a group of 6 or 8 species 

 allied to C. compressus, but to show how such group differs from 

 all other species of Eucyperus, which difficulty I leave virtually 

 untouched. I think the stout, subequally triquetrous nut at 

 least as important as any of the characters insisted on by 

 Boeckeler. 



