REDUCTION OF VARIABILITY. 127 



comes soaking wet, it is not adapted to life in the open. It is 

 very probable that the difference in hair is primary and the 

 habit of living either indoors or outdoors is the result of this 

 difference in texture of the hair. 



There are cases in which an organism, when transported into 

 a new environment, or employed for other uses finds itself very 

 well adapted. In other cases some factor in theenvironment may 

 change, and the animal may find itself able to profit by the 

 change, to adapt itself to it. The case of the Australian Kea, a 

 parrot which is killing sheep is a very striking example. Attempts 

 to introduce new species of animals or plants often fail because 

 local parasites find themselves beautifully adapted to a lite in 

 which the prc sence of the new introduction is an important factor. 



Of course it is impossible to rule out of court the possibility 

 that isolation may in some instances be the cause of speciali- 

 zation and adaptation. However, it will be seen that isolation 

 must be of a complete nature, and the initial adaptation must 

 be rather great, to keep the species in the environment to 

 which it is not yet wholly adapted, and in the second place to 

 prevent it from dying out. 



If, by a heightening of the potential variability of a species, as 

 the result of some cross, individuals are produced, which are so 

 constituted that at the time of reproduction they prefer an 

 environment different from that in which the other individuals 

 live, this fact will provide a means of isolation as effective as 

 colonization on an island. Such a group of individuals (toads 

 which stay on land to pair) will speedily lose its potential varia- 

 bility. They must at the outset have a somewhat high variability 

 for the cross which was the primary cause of the production of 

 their new character cannot have failed to have heightened the 

 potential variability all around. Therefore the chances are that 

 the group will finally have a genotype different in several genes 

 from that of the parent-species. And during this reduction of its 

 potential variability natural selection could play a role. Causes 

 heightening the potential variability of the parent-species would 

 not affect it. 



