SECONDARY' SEXUAL CHARACTERS 125 



which are recurrent in each generation. The germ- 

 plasm is homogeneous for all members of the pure line, 

 while in a mixed population the germ-plasm is not the 

 same for all individuals. 



Darwin himself saw this to some extent, for he has 

 repeatedly pointed out that selection depends on the 

 materials offered to it by variation ; that in itself it 

 can produce nothing. Yet from Darwin to Johannsen 

 the teaching of the post-Darwinians has been such as to 

 lead most people to believe that selection is a causative 

 or creative principle that will explain the progressive 

 development of animals and plants. 



DISCONTINUOUS VARIATION OR MUTATION AS A BASIS 



FOR SELECTION 



The second great movement since Darwin has been 

 to show that hereditary variations do not give a con- 

 tinuous series but a discontinuous one. Bateson and 

 De Vries brought forward some twelve years ago evi- 

 dence, in favor of this view, that has gone on increasing 

 in volume at an amazing rate. 



I cannot attempt to discuss this evidence here, but 

 I may point out the bearing of the new point of view 

 on the meaning of secondary sexual characters. 



In a number of butterflies there occur two or three 

 or even more different kinds of females. One of the most 

 remarkable cases of the kind is that of Papilio polytes 

 that lives in India and Ceylon. It has a single male 

 type and three types of females (Fig. 64). 



Wallace, who first observed that the three types of 

 female belong to one male type, argued that two 

 of these three types owe their origin to their resem- 



