366 THE VARIATION OF ANIMALS IN NATURE 



mutations is not convincing, but no other equally concrete 

 explanation, supported by direct observation, can be brought 

 forward. We suggest that as far as internal relations are con- 

 cerned the organism itself may in some sense initiate new 

 steps forward. If such steps took the form of mutations as we 

 know them, the multiplication of the latter might be due to 

 a form of Natural Selection which preserved the best organ- 

 ised rather than those most specialised for any particular 

 environment. 



Specialisations, in our sense, might well be due to Natural 

 Selection of the classical type, but even here we feel that there 

 are certain difficulties. The greatest, perhaps, is the lack of 

 sufficient direct evidence for such a process. Even if its 

 efficiency had been proved, it would still be uncertain whether 

 all specialisations could be explained in this way. 



It is known in several species that each intraspecific 

 genotype has its own characteristic potentialities, e.g. viability, 

 fecundity, etc. If such genotypes are put in competition in 

 a standard environment, one type will finally replace all the 

 others. It has been held (e.g. Haldane, 1932, chapter iv) that 

 this proves that Natural Selection must be taking place con- 

 tinuously amongst such mixed assemblages in nature. No 

 doubt, if some of the types are markedly defective, this will 

 be true ; but usually the position is not so simple. The condi- 

 tions in nature, for instance, are not standardised but highly 

 variable, and many types may scarcely have any opportunity 

 to exhibit their characteristic norm. Behaviour patterns and 

 physiological attributes such as viability appear always to show 

 a considerable range of variability, even under standard 

 conditions, and in nature the selection of genotypes on 

 the basis of phenotypic performance must at the best be a 

 very slow process. As we have said in Chapter VII, selec- 

 tion between large populations, which already differ in many 

 respects, is more easy to understand than selection of indi- 

 vidual variants. We suggest that, even in specialisation, the 

 internal relations of the organism may play a not unimportant 

 part. 



Finally, many of the small characters which differentiate 

 species appear to be entirely useless, and here we believe 

 random survival, combined with isolation and occasionally 

 with hybridisation, may have played an important part. 



