TERMINOLOGY IN MONOCOTYLEDONS. 503 



venation, or of the greater or less development or total absence of 

 the axillary flower, and to no others. 



These glumes are universally present in the three orders, and 



with very few exceptions (as in Anarthria and Lepyrodia) closely 

 imbricate. Generally speaking, one or two, or sometimes more 

 of the lowest, and occasionally one or more at the end of the spi- 

 kelet, are empty and usually smaller than the others. Sometimes, 

 however, and especially in some uniflorous spikelets, the upper- 

 most or all of the lower empty glumes is larger or differently- 

 shaped from the others. In androgynous, unisexual, or polygamous 

 spikelets, the glumes frequently differ considerably in form, size, 

 or venation, according as they enclose hermaphrodite male or 

 female flowers or are quite empty ; but their alternate imbricate 

 insertion (whether distichous or all round) on the main axis of 

 the spikelet is always constant. 



The special terms conveniently applied to the peculiar scales 

 observable in some genera or orders within the glumes on the 

 secondary or floral axis will best be considered under the several 

 orders, on which I shall now offer a few observations. These ob- 

 servations will not be so much directed to the several distinctive 

 characters of the orders, which have been too clearly explained by 

 various botanists to require further comment, as to the various 

 modifications which the spikelets with their glumes and other 

 scale-like organs undergo, and which have given rise to the con- 

 fused terminology it is my object to expose. 



Eriocauleae have on each scape or peduncle a single terminal 

 spike with closely imbricate bracts, each enclosing a single flower, 

 or the outer ones empty, thus answering to the general character 

 of the spikelet and its glumes ; but as the axis of the spike or 

 receptacle is a direct prolongation of the peduncle or scape without 

 any distinctly subtending bract, as the spike itself takes the 

 globular or depressed form usually designated by the term head, 

 very rarely lengthening to a cylinder, and as the imbricate bracts 

 have often a somewhat herbaceous or flaccid consistence not 

 directly associated with the idea of a glume, there appears to be 

 no inconvenience in retaining in this order, as is usually done, 

 the more general terms of flower-head (capitulum) with its im- 

 bricate bracts, bearing always in mind their strict homology with 

 the spikelet and its glumes of the larger glumal orders. 



The flowers of Eriocauleae are unisexual, arranged in androgy- 

 nous or rarely unisexual heads ; and although concealed under the 



