OTHER THEORIES OF EVOLUTION 319 



succeed or fail' (Bateson, 1909, p. 289). Nevertheless, he 

 (19 1 3) frequently implied that selection could not be operative 

 in bringing about local variation and the formation of races 

 and species, though he was aware of the necessity of explaining 

 how single mutations can multiply and spread through a 

 population. 



Recently, however, the means whereby variant indivi- 

 duals could survive and multiply without selection have been 

 formulated more definitely, and with a realisation of the diffi- 

 culties involved, by Elton (1924, 1930), Cuenot (192 1), and 

 Robson (1928). 



The prime difficulty in the way of this theory is of course 

 the theoretical one that only those mutations which are of 

 selective advantage have a chance of survival. But any theory 

 of evolution which depends on the chance survival of muta- 

 tions unaided by any directive agency is confronted by an 

 additional difficulty. The facts of evolutionary history give 

 a very decided impression that they have been influenced by 

 some directive tendency. That tendency, though not always 

 adaptive, almost invariably has some definite orientation. 

 It may not be apparent in the world of living species, which 

 appears to us very largely as meaningless and chaotic in its 

 divergences. But it is inevitably forced upon our notice in 

 any study of geological series, any morphological history and 

 in any systematic treatment of a large group. The evolutionary 

 process seen in such histories scarcely looks like one of which 

 the main tendencies have been determined by chance and 

 random survival. Evidence of such variation is, it is true, 

 seen in some of the lineages disclosed by palaeontology. But 

 the whole process is too obviously canalised and subject to 

 direction to be the product of chance. An attempt was made 

 by Morgan (1919, p. 268) to reconcile this obvious aspect of 

 the process with the operations of chance ; but we do not think 

 that his contention — viz. that a mutation in a certain direction 

 increases the likelihood of further mutations in the same direction 

 — can be sustained (cf. Robson, 1928, p. 248). 



In addition to the appearance of a directive influence in 

 evolutionary series, the development of organs of high com- 

 plexity and of 'co-adaptations' (p. 306) renders still more 

 improbable the likelihood that the chance survival of muta- 

 tions has been the only mechanism of evolutionary change. 



