268 W. J. CROZIER 



'17 c). Given the conditions imposed upon the nudibranchs in 

 the experiments just described, it is legitimate to conceive that 

 here, if ever, the animals would have conjugated according to 

 the chance dictates of random contact. The magnitude of the 

 correlation index makes untenable the idea of random pairing; 

 moreover, close study of the behavior of C. zebra in copulation 

 fully substantiates the proposition that true selective conjuga- 

 tion does indeed occur. 



VI THE BA§IS OF HOMOGAMY 



1. Homogamy in Paramecium and in man. In human mar- 

 riages there occurs, according to Pearson (Pearson and Lee, '03) 

 and others, an appreciable degree of positive correlation between 

 the two members of a mating pair as regards their stature and 

 certain other characters. The evidence available with refer- 

 ence to this matter has been reviewed in an interesting essay by 

 Harris ('12 a). The cause of assortive mating in man is in 

 most cases far from obvious; its significance, its consequences, 

 cannot be followed in detail. 



For Paramecium the case is clearer. It was proved by Pearl 

 ('07) that in Paramecium there exists a rather high degree of 

 correlation between the lengths of the components of conjugat- 

 ing pairs. Pearl held this correlation to be the result of assortive 

 mating, and he was inclined to accept as its explanation a rela- 

 tively simple mechanical condition concerned with the manner 

 in which two conjugants become adjusted to each other in a 

 successful mating. According to Pearl, if two individuals, in 

 the proper physiological condition for conjugation, are by a 

 chance mutual approach induced to draw together, 'Hheir oral 

 surfaces adhere in whole or in part. The extreme anterior ends 

 of the oral grooves firmly adhere to one another first. If the two 

 individuals are so nearly the same size that the mouths ap- 

 proximately coincide when the anterior ends are together, firm 

 union occiu-s at the mouth regions and definite conjugation fol- 

 lows" (Pearl, '07, p. 267). If the two individuals do not 'fit,' 

 they separate, or die. ''The homogamic correlations arise, then, 

 as a result of the necessity for the mouths of the two individuals 



