XXI 



sion was a subject of no little consideration on the part of the au- 

 thor ; but, himself unsatisfied with any existing systems of arrange- 

 ment, which are all so artificial and uncertain that scarcely any 

 two, among the more intelligent class of writers, seem agreed as to 

 the limits of groups or the allotment of genera, it was determined 

 to avoid interruption to the series of the latter in the body of the 

 work, and rectify such omission by a tabular arrangement of them 

 in the Introduction. The decision may, on the one hand, have its 

 disadvantages; but, on the other, it has tended to render the 

 book less fearfully technical than it would have been, had the con- 

 tinuity of the subject been disturbed by sectional paragraphs in 

 itahcs, which, however useful to the persevering student, contri- 

 bute much to perplex and even disgust the uninitiated reader. 



The attempts to classify the Grasses, by associating the genera in 

 natural groups, have been numerous; and some of the highest 

 names in the annals of natural science are guarantees for the in- 

 genuity and judgment exercised on the systems to which they have 

 given existence. That of Endlicher may eventually become the 

 basis of one less exceptionable than it is itself at present. Under 

 it, the British Grasses, wild and cultivated, are distributed in the 

 following tribes, the leading characters of which are given, followed 

 by the names of the genera appertaining to each, Such characters 

 will be more easily understood by reference to the magnified 

 figures of the flowers on our Plates. 



Tribe 1. OryzejB. Rice family. 

 Spikelets laterally compressed, one-flowered. Glumes absent. 

 Paleffi two, enclosing, but not adhering to, the ripe fruit. Stamens 



three or six. 



Leersia. 



In Orysa itself, the Rice-plant, there are two glumes to the 

 spikelet and the stamens are always six. Zizania and others 

 furnish valuable farinaceous food in their seeds. 



