24 MB. C. B. CLABKE OS INDIAN SPECIES OE CTPEBUS. 



say with much confidence that the Australian fragment (on which 

 Mr. Bentham doubtfully extended C.jlavescens to Australia) is 

 not C.jlavescens, and that G. flavescens has not yet been col- 

 lected east of Cabul. Its area, corrected by this character, 

 becomes very natural; for the South- American G. Olfersianus 

 (with its allies or subspecies from Mexico and the Andes) has 

 the outermost layer of cells of the nut subquadrate, and separates 

 off cleanly. 



These species are usually described (Boeckeler &c.) as having 

 the nut "transversely striate" or " transversely undulate-rugose.' 

 The outer cells are in this group hyaline, and more prominent at 

 the narrow ends of the cells, in many cases actually wearing away 

 from the middle of the cells; hence the nut, under a pocket 

 lens, takes the aspect in fig. 41. The "transverse striation" is, 

 however, a very deceptive character, being prominent in some nuts, 

 utterly obscure in others from the same plant (all characters de- 

 rived from degree of flaking-off are untrustworthy throughout the 

 Order, as I attempted to show for Garex in my ' List of Andover 

 Plants ' 17 years ago). The character derived from the oblong 

 (not quadrate) cells is absolute. 



In the quadrate cells there are minute differences, which are 

 mostly very constant for the same species, and eveu for the same 

 variety. In many of the smaller nuts the outermost layer of 

 cells is thin, hyaline, not detergible ; consequently it cannot be 

 detected without a microscope and a dark field. The nut is usually 

 described by cyperologists in this case as lucida. 



Sometimes the outer cells are opaque, white, permanent, as 

 fig. 35. Such a nut is seen in G. bromoides, "Willd., and is de- 

 scribed (by Boeckeler) as " transverse lineato-granulata fusco- 

 nigra nitidula." In G. unioloides and G. angulatus the outermost 

 cells of the nut are similar, but nearly always broken at the 

 centre, so that the nut is described as " porose " (fig. 38). I con- 

 sider (as does Mr. Bentham) G bromoides, G. unioloides, and G. 

 angulatus one species ; but I can nevertheless from one nut only 

 generally tell whether the specimen came from Australia, India, 

 the Cape, or South America. 



The outer cells of the nut (rarely) are somewhat inflated, lax, 

 and yet do not break up at the centre. Such nuts are described 

 as scabrous-papillose and by various names. The character 

 occurs sometimes in C. globosus (as see fig. 36), but is not uni- 

 versal for that species. 



