MUTATION AND ADAPTATION 415 



took place simultaneously in flower and insect, and in the same 

 locality, and in a sufficient number of each, a supposition which, 

 to say the least of it, is wildly improbable, the delicate correla- 

 tion between the two would be thrown out of gear and the 

 individuals exhibiting the mutations would be eliminated by 

 natural selection because they were no longer sufficiently well 

 adapted to the very special conditions of their environment. 



We can only believe that the increase in length of nectary and 

 proboscis took place so slowly that their reciprocal adaptation 

 was never upset. A slight increase in the length of the nectary 

 obliged the insect to poke further into the flower for the honey 

 and thus increased the chances of fertilization. A slight increase 

 in the length of the proboscis enabled the insect to get more 

 honey and thus gave it a better chance of existence. After 

 perhaps many thousands of generations, under the influence of 

 natural selection, combined in the case of the insect with the 

 effects of use and disuse, the present enormous lengths of 

 proboscis and nectary have been attained, and no one doubts .the 

 fact that they have become blastogenic characters. 



The same argument applies to all accurate adaptations to 

 special conditions of the environment. How can we explain the 

 facts of protective resemblance and mimicry except as due to the 

 accumulation under the influence of natural selection of what 

 Charles Darwin called slow successive variations ? How, again, 

 can the theory of mutation be applied to such cases as that of 

 the flightless birds on oceanic islands ? Who can doubt that 

 the reduction of the wings and the loss of the power of flight has 

 been brought about slowly and gradually as a result of disuse ? 

 and at the same time who would venture to argue that the 

 flightless birds are not specifically distinct from their actively 

 flying ancestors ? If it be urged that mutations may be so small 

 as to be almost imperceptible, then we must ask how do they 

 differ from fluctuating variations ? and if we are told that they 

 occur very rarely and do not fluctuate, that very answer is 

 sufficient to show that they can hardly have given rise to adaptive 

 modifications. 



De Vries' theory of the origin of species -by mutation is 

 supposed to harmonize with the Mendelian principle of unit 

 characters, but we have to ask ourselves, how do new unit 

 characters arise in the first instance ? It seems at least as 

 probable that they arise by the gradual accumulation of slight 



