PREFACE. V 



a subject, he has fully and gracefully acknowledged 

 his indebtedness. One important feature of the work, 

 which is quite new, and which alone could make a 

 production of this kind of value to a student, is the 

 method of grouping. This is done in such a way 

 that should a first test fail to identify a grass in one 

 group, it is certain to be discovered by referring to 

 another group which also contains the grass. The 

 keystone of success, however, lies in the simplification 

 of the whole system, by the ease with which the five 

 most common grasses in pastures Ryegrass, Fescue, 

 Dogstail, Holcus, and Foxtail can with the greatest 

 certainty be eliminated from the mass of herbage, and 

 separated from one another, by the colours of the 

 bases of their leaf sheaths. This, I believe, is what 

 no one who before endeavoured to formulate a 

 scheme for the identification of grasses by their 

 foliage, has ever noticed, and no doubt is the main 

 reason why all attempts have hitherto failed from a 

 practical and useful point of view. The great bulk 

 of the work of separation can be carried through by 

 the aid of the naked eye, and simply at a glance. All 

 grasses that remain, after the five coloured ones are 

 removed, are but a trifle to treat critically, as com- 

 pared with the whole bulk. 



I believe this work is one which will find its way 

 into the hands of all teachers and students who are 

 interested in the details of the composition of natural 

 A 



