112 Mr. G. C. Robson on 



be, ranging from the crude concept of the individual 

 taxonomist^s ^\flair^^ or " eye for a species " to a more serious 

 attempt to synthesize a number of correlated characters after 

 a careful biometrical study of variation in the group. 



The latter method seems to offer at least a means of 

 bridging the gap between the taxonomist and the geneticist. 

 We owe to Morgan (5) the concept of the species as a gene- 

 complex. At present this is only a suggestion unelaborated 

 and undeveloped ; but it seems to be precisely parallel to the 

 view of certain taxonomists who lay stress on the importance 

 of a number of closely correlated characters. 



When animals representative of a "species'^ are studied 

 intensively by the taxonomi.>«t there is revealed a large 

 complex of structural features which exhibit the effects of 

 "specific " differentiation. According to my own experience 

 it is hard to find a system which is not capable of showing 

 the latter. Even glandular structures may be found to 

 exhibit differentiation in two closely related forms, as 

 Kleiner (3) has shown. 



It is the composition and fate of the character-complex 

 that has been most neglected in discussions on the nature of 

 species. In practice it has been recognized, but its 

 implications have not been followed up by intensive work on 

 genetic and morphological lines. 



We have to ask ourselves, if a species is a complex of con- 

 stantly associated characters, how does the complex behave in 

 heredity ? Are the various characters affected by linkage * ? 

 If so, can w^e ascertain the degree of linkage and thus obtain 

 a standardization of the species ? If there is no linkage, 

 what is the nature of the association of characteis? Is it 

 purely fortuitous and susceptible of dissociation by crossing ? 

 Have we, finally, an advance from unstable character- 

 associations to systems exhibiting linkage in various degrees ? 



We are certainly not in the position to answer these 

 questions yet ; and to suggest lines of research and hint at 

 problems to be solved is perhaps a poor sort of policy. But 

 in the present case any sort of criticism of method is 

 desirable, especially if it affords a liaison between two types 

 of research. The writer is constantly impressed in his 

 laxonomic studies of the Mollusca with the instability of the 

 cliaracter-complexes conventionally known as '^ species '^ and 



* The term linkage is used here in the general sense of a tendency on 

 the part of certain characters " to keep together rather than to assort 

 freely," Morgan (5, p. 80). 



