508 MR. G. BENTHAM ON CLASSIFICATION AND 



The double enclosure of the caryopsis in the flowering glume 

 and upper empty one occurs also in the small or perhaps mono- 

 typic allied genera Courtoisia and Semirea, which, however, are 

 sufficiently distinct from Kyllinga in the shape and texture of the 

 glumes as well as in habit and inflorescence. 



Within the glumes of some genera of Cyperaceae, on the secon- 

 dary or floral axis, and outside or below the stamens or ovary, are 

 certain scales or sometimes setae, very much diversified in number, 

 shape, and relative position, the terminology of which has been very 

 much confused, and requires settlement according to fixed rules, 

 in so far as their probable homology can be fairly exhibited, and 

 more especially so on the present occasion, with a view to com- 

 parison with that of G-ramineae. 



In the majority of the species of Hypolytrum (Plate VIII. fig. 6) 

 there are at the base of the floral axis two distinct apparently 

 opposite concave or complicate keeled scales, placed laterally with 

 regard to the axis of the spikelet and to the subtending glume, 

 and no others. These, from their position and other characters, 

 may fairly be considered as bracteoles, analogous to those of 

 Anarthria and Lepyrodia in Eestiaceae, or of Xyris and of several 

 genera of Coronariae. In Hypolytrum punyens these two brac- 

 teoles are united by their margins next the axis of the spikelet 

 nearly to the apex into a single flat two-keeled scale, open on the 

 anterior face only (next the glume) for the emission of the stamens 

 and pistil. In Platylepis (Plate VIII. fig. 7) the union is the 

 same, but more complete at the apex, the whole scale being pro- 

 duced into a short, solid, but flat point. The keels are lateral as 

 in Hypolytrum, but broadly winged ; and the bracts at their very 

 base are rather obliquely inserted. In Ascolepis, a genus closely 

 allied to the Platylepis capensis, the scale is more oblique, and 

 consists chiefly of a long solid point, with a short fissure at 



the base for the stamens and pistil. In He 



is exceedingly reduced; it is flat, or slightly concave, entire, 

 or 2-toothed at the apex, very thin and hyaline, often very 

 difficult to detect, and sometimes disappears altogether. That it 

 is, however, homologous to the scale of Platylepis, and not, as 

 suggested by Boeckeler, a staminode, is shown by the frequent 

 bifurcation of its apex ; and occasionally (as in a flower examined 



is below one of them. Whethe 



