216 THE VARIATION OF ANIMALS IN NATURE 



variants in natural populations. Aubertin, Ellis and Robson 

 (1931) have studied separate colonies of a land snail over three 

 years in a fairly circumscribed area, and have found a rather 

 limited degree of change in the individual colonies in the 

 period of observation. 



III. The Nature of Variation. — The causes, kinds and 

 incidence of variation are discussed elsewhere in this work 

 (Chapter II). What we have to ask here is whether our 

 present knowledge of these is consistent with a belief in the 

 efficacy of Natural Selection as the chief agency in evolution. 



As already pointed out (p. 183), Darwin took all the facts 

 of variation at their face value. In the most active period of 

 his work at least, he believed that a substantial part of variation 

 was due to environmental effects, and he was at no pains to 

 distinguish between the somatic and germinal origin of varia- 

 tion. Still less did he explicitly distinguish between what are 

 now known as gene-mutations and the variation which is due 

 to factorial combination (p. 189), though he was, in fact, 

 familiar with the variation due to crossing. There was, in 

 short, available for the action of Natural Selection a large 

 store of variation, the hereditary fate of which he did not 

 seriously consider and the potentialities of which for per- 

 manent improvement he did not explore. This vagueness was 

 in some measure clarified by de Vries. on the one hand and 

 Weismann on the other, and the evolutionary speculations of 

 the period about 1880- 1920 were based on the recognition of 

 germinal as opposed to ' fluctuating ' variation, of which the 

 former alone was held to be of evolutionary significance. 

 Furthermore, genetical investigations revealed the distinction 

 between mutation (change in the constitution of a gene or of a 

 chromosome) and the variation due to heterozygosis in the 

 parents. 



In Chapter II we have examined the evidence as to the 

 inheritance of induced modifications. We concluded that some 

 of the data suggest that this, at least, is possible in certain 

 circumstances. Although the conditions under which such a 

 process can operate appear, at present, to be rather restricted, 

 its mere possibility cannot but make the premises of all evolu- 

 tionary speculation somewhat uncertain. As we have pointed 

 out, the problem of the evolution of habit and instinct still 

 requires a solution, without which any theory deduced merely 



