338 THE VARIATION OF ANIMALS IN NATURE 



heterogony of the Roe-deer has arisen because ' it was for 

 some reason biologically desirable for the Roe-deer to have 

 small antlers.' According to this view the ultimate causes 

 of quantitative differences have to be sought in various 

 circumstances of adaptation. Huxley's formulae give us only 

 the expression of particular relationships. If we interpret 

 Huxley's meaning correctly, we might say that while, e.g., it 

 might be functionally desirable to have a large appendage, the 

 precise size is determined by the absolute size of the body. It 

 is, indeed, by no means clear to what extent increase of total 

 bodily size alone is held to be causal. Huxley (I.e. p. 227) 

 suggests that the increase of the male chela in Uca is due to the 

 increase of absolute size ' owing to the specific growth-intensity 

 of the organ, which in its turn is presumably due to a specific 

 growth-promoting substance.' Huxley claims (pp. 218-19) 

 that the principle of heterogony enables us to dispense with an 

 appeal to orthogenesis (in the sense of determinate evolution) , 

 e.g., in explaining the large size of the horns of the Titano- 

 theria. ' Granted (a) that there existed in the germ-plasm of 

 the ancestor of the four lines of descent the hereditary basis of 

 growth-mechanism for a frontal horn, and (b) that increase of 

 size up to a certain limit was advantageous for Titanotheres in 

 general, as would seem inherently probable, then the results 

 follow without any need for invoking orthogenesis. Natural 

 Selection would account for the increase of absolute size, and 

 increase of absolute size would evoke the latent potentialities 

 of the horns' growth-mechanism.' The value of this explana- 

 tion is, of course, entirely dependent on the validity of 

 Huxley's assumption that increase of body size is produced by 

 selection. 



(3) The theory by which Haldane has sought to explain 

 certain types of orthogenetic phenomena in terms of Fisher's 

 work on the effect of selection on metrical characters deter- 

 mined by many genes, has been already discussed (p. 327). 

 It was, no doubt, intended by its author to explain excessive 

 size of parts (of the Machaerodus type) as well as other examples 

 of useless orthogenesis.' As we pointed out (I.e.), the theory 

 is an ad hoc construction and its premises have to be accepted 

 on trust. 



(4) From this review of theories as to the cause of excessive 

 growth, which are based on some form of selection and on 



