516 Bicknell : Further Notes on the Agrimonies 



know nothing may well have been available to him at that time to 

 influence his determination, at any rate any amount of argument 

 as to what might, could, would or should have been done can be 

 of no avail as against what actually was done, and on matters of 

 that kind we may well hesitate to attack the wisdom of Dr. Gray. 



In attempting to correlate an actual specimen of a plant 

 with a written description there is always a very natural tendency 

 to read into the words of the description the particular characters 

 of the specimen seeking determination. Descriptive terms must 

 necessarily be more or less elastic owing to the great variability 

 in the characters of organic beings and it is well, therefore, al- 

 ways to guard against a too formal understanding of such terms 

 or to attach to them an adequacy which nature itself is certain to 

 repudiate. It follows that a very brief description of a plant 

 which may be quite sufficient in the case of a monotypic species is 

 very apt to be found to be little more than generic if the species 

 finally proves to be only one of a closely related group. The 

 description of A. striata Michx. is, as already shown, a case in 

 point. Contrariwise, in the case of the plant A. glabra (Muhl.) 

 which forms part of the type of A. striata Michx., it will not do to 

 insist too strongly on the character of " globular fruit," a term 

 which wholly fails to cover the great variability in the fruit of this 

 species. At full maturity its fruit is often perfectly graduated to 

 the base — fairly turbinate on a not too narrow understanding of 

 that term — which may on may not be somewhat curved, the very 

 character from which Wallroth derived the specific name rostellata. 

 So, too, the fruit becomes definitely reflexed, if less conspicuously 

 so than in A. Brittoniana, and in some of its variations may well 

 be described as sulcate and crowned by the bristles just as is 

 called for by Michaux's description. It may be said further that 

 A. Brittoniana is by no means always so sharply distinguished in its 

 characters as to be readily separable from specimens of A. hirsuta 

 or A. mollis, and specimens occur which need careful discrimina- 

 tion in order to be confidently distinguished. 



In discussing A. Brittoniana Dr. Robinson has more than once 

 referred to the " first mentioned type," a phrase absolutely without 

 meaning when applied to a species of which an exclusive type has 

 been set apart. It so happens that the plant thus referred to was 



