132 JOURNAL, BOMBAY NATURAL HISTORY SOCIETY, Vol. XVII. 



(16) In practice "a species" is a very real unit of classification 

 presenting a very concrete shape to a naturalist's mind but is 

 difficult of definition. 



(17) The difficulty arises entirely from the exceptions which would 

 occur in classes (1) and (2) to almost any form of wording. 



(18) Nevertheless even in these the "species" is a very real 

 entity, and in class (1) can practically always be discovered 

 by selective breeding. 



(19) Some cases of geographical races (2) can also be shown to be 

 purely climatic by breeding also ; for when once the essential 

 difference between dry and wet season forms is known, the 

 extremes of each and intermediate of each can be inferred and 

 reconciled. Geographical differences of this kind should be 

 ignored systematically and the types treated as one species. 



(20) Other cases of geographical differences cannot be dealt with 

 by breeding, and each case must be judged on its own merits, 

 bearing in mind (a) magnitude of the difference, (b) its constancy, 

 (c) the relative constancy of other specific characteristics, (d) 

 the relative sharpness of the boundary of the distributive areas 

 of the several forms, and the presence or absence of intermediates 

 near the frontier. 



(21) For simplicity sake it is desirable to unite rather than divide. 



(22) A description of a "new species" from a single capture — 

 unless it shows pronounced and not purely superficial peculi- 

 arities- — is of doubtful value and most dangerous in practice. 



If these conclusions are just, then there is certainly such a thing in 

 nature as a species, and conversely a " subspecies" is an absurdity. 



There is no harm of course in describing, and defining where possible, 

 geographical races, but to give them separate names is only to add an 

 element of confusion where all should be clear and simple. 



The writer claims no finality for his views, but the subject is one of 

 such importance and general interest that it is well worth a discussion 

 in our journal. 



