EVOLUTION 105 



ency to become somatic cells. In a similar 

 way, somatic cells are never derived from cells 

 which have begun to assume the character of 

 germ cells, but are also directly derived from 

 the seg^mented ess- The two classes of cells 

 are thus fundamentally distinct, and it is not 

 inappropriate to describe the body of one of 

 the higher animals as composed of an enor- 

 mous aggregation of somatic cells in which is 

 harbored and protected a certain number of 

 germ cells. 



What has been defined as acquired charac- 

 ters, namely, those changes which are not the 

 products of inheritance, but which are the di- 

 rect results of some change in the environ- 

 ment or in the activities of an animal, are 

 necessarily alterations in its somatic cells. The 

 change in a muscle in consequence of a new 

 form of exercise is a change in somatic cells; 

 the alteration that the skin exhibits under a 

 new exposure is of the same kind ; the change 

 in the nervous mechanism by which a new 

 habit is established is also somatic. If acquired 

 characters, then, are modifications in the so- 

 matic cells, and these cells are distinct from 

 the germ cells, it is difficult to see how a change 

 which may affect a group of somatic cells, 



