3 o6 THE VARIATION OF ANIMALS IN NATURE 



conditions provided by the habitat may be present, but we 

 certainly cannot yet show that it exists. The highly successful 

 introduction of species from one country into another, or, as 

 in the case of many pests, from one habitat to another, does 

 not suggest that species so introduced were originally adapted 

 to a very close range of conditions. And where several appar- 

 ently closely allied species occur in one habitat and yet differ 

 from one another in structure in much the same way as other 

 species which live in different habitats, we find it difficult 

 to believe that every expansion or restriction of the habitat 

 range of a species necessarily implies a closer adaptation to 

 the new conditions. 



(4) Complex organs and ' co-adaptations. ' 



Though most organs are complex and probably all adapta- 

 tions are ' co-adaptations,' both have been supposed by many 

 authors to present a special problem, and we think a brief 

 consideration of them may be of some value. 



The difficulty of explaining the origin of complex organs 

 by means of the selection of small variations is well set out 

 by Darwin in his ' Origin of Species ' (1884, pp. 143-9). ^ n 

 one respect, however, Darwin's argument has been weakened, 

 inasmuch as Fisher (1930, pp. 73-83) has demonstrated that, 

 if his premises are admitted, new characters which are not 

 directly or indirectly (i.e. by correlation) adaptive are very 

 unlikely to spread through a population. Now Darwin 

 throughout his book supposes that some part of the origin of 

 complex characters may be due to the persistence of characters 

 not positively harmful, and this is helpful in accounting for 

 the early stages of various evolutionary processes. This 

 supposition, however, cannot be made if variants are supposed 

 to arise through rare mutations which have to spread through 

 the population and have little chance of persistence without 

 the aid of selection. A somewhat heavier burden is thrown, 

 therefore, on Natural Selection, which has to play the 

 dominating part throughout the evolution of any structure. 



The essential feature of any complex organ such as the 

 mammalian eye or kidney is the co-ordination into one working 

 whole of a number of separate structures and tissues. The 

 difficulty of obtaining such co-ordination by the selection of 

 random variations in the various parts is sufficiently obvious. 



