1 82 THE MUTATION THEORY 



occurring within that immensely complex and delicately adjusted 

 machine, a higher animal. To be useful the mutation must 

 not only be an improvement in itself, but it must be correlated to 

 a thousand readjustments a thousand co-adapted mutations in 

 other parts of the machine. Otherwise it will be positively harm- 

 ful. 1 Indeed, the principal difficulty to be surmounted by the 

 supporters of the mutation hypothesis is not adaptation to the 

 environment but co-adaptation within the organism. Suppose, 

 however, we cede the point, and agree that the miracle may happen 

 and a thousand co-adapted mutations may occur in a single indi- 

 vidual. Then, either the reproduction will be blended, or it will be 

 alternative. If it be blended, the mutation will be swamped unless 

 more miracles happen and many other individuals mutate in the 

 same way and at the same time and place. If it be alternative, 

 the co-adaptation will be lost in succeeding generations through 

 mixture with the characters of individuals who have not mutated ; 

 for Mendelian observers insist, and very rightly, that the alternative 

 characters are reproduced independently of one another, and 

 that all the characters of the same individual are not necessarily 

 patent or dormant together. 



301. It must be noted, morever, that, not only are the parts of 

 a living being co-adapted, but that different species of animals 

 and plants are also co-adapted to one another. Thus the plants 

 which herbivorous animals are capable of using as food possess 

 great powers of multiplying themselves and of reacting to the 

 stimulus of injury, of regenerating lost parts. Thus also the 

 powers of pursuit possessed by carnivora are closely related to 

 the powers of escape possessed by their prey. If evolution were 

 by mutation, then a mutation, increasing the speed of one species, 

 could only be met by a mutation occurring at the same time and 

 place, increasing the speed of the other. For example, if wild 

 dogs suddenly mutated in this way, and the various animals they 

 hunt did not, the inevitable extinction of the latter would soon 

 be followed by that of the former. 



302. It is possible to account for the co-adaptation of parts 

 and species to one another on the theory that evolution results 

 from the continued selection of constantly occurring small fluc- 

 tuations from the constant and steady change of species in 

 adaptation to changes in their surroundings ; but how is it possible 

 to account for it on the theory of large and, speaking comparatively, 



1 De Vries (Darwin and Modern Science, p. 68) insists that correlated mutations 

 are common. But correlation is not the same thing as co-adaptation. 



