1 921.] and theh' part in Evolution. 721 



subspecies, and were designated in our " system " by a tri- 

 nomial. In my own opinion the binomial species name is the 

 only one that should appear as a heading in books and papers, 

 and the subspecies should be given as a sub-heading— this, 

 by the way, as my present object is not to deal with trivial 

 details, but to get at what subspecies are and the part they 

 play in evolution . 



It is generally agreed, though the rule is frequently 

 honoured in the breach, that trinomials should only be given 

 to geographic forms — that is to say, to variations due solely 

 to the influences of the climatic conditions of their habitat, 

 and it is reasonably stretched to include island varieties; but 

 it must not be forgotten that in an island race, "isolation' 

 forms an additional factor which has influenced the separa- 

 tion ; and in a closely-lying group of islands where each 

 island has a distinct form though the climate is similar, 

 " isolation " has probably been the chief factor in separating 

 the race"^. I will now consider Messrs. Lowe & Praed's 

 letter (' Ibis,' p. 344), in which at the beginning they strike 

 the right note by asking " of what scientific value are these 

 variations ?" I will not deal at length with the rest of this 

 interesting letter, but merely note two points — they imply 

 that subspecies are of two kinds: (1) due to "discontinuous" 

 variation, (2) due to environment. I had always understood 

 that a true subspecies was always supposed to be restricted 

 to the latter cause, and certainly think it should be so. 

 Secondly, they imply that variations (subspecies) caused liy 

 environment cannot be inherited ; and hence it would follow 

 that a true subspecies could have had no influence on 

 evolution, and could never develop into a new species. 

 Unfortunately, neither of these authors can have had any 

 practical experience of breeding birds or animals, or they 

 would realize how very small variations can be intensified 

 and fixed from a purely selective, sis apart from any creative, 

 agency. They are quite right in carefully distinguishing 

 between continuous and discontinuous variations, but they are 



* The same conditions occur not uufrequontly on land where races are 

 separated by tracts of country unsuitable to them. 



