ASSORTIVE MATING. 24I 



I have examined this phenomenon in decemlincata, undecimlineata, and 

 dilecta, and have obtained results exactly like those just described for ob- 

 longata. Among 300 copulating pairs of decemlincata from Cold Spring 

 Harbor, Long Island, New York, ioo pairs each from Woods Holl, Massa- 

 chusetts, and Newport, Rhode Island, 200 from Yellow Springs, Ohio, and 

 300 from Chicago, Illinois, or 1,000 pairs in all, I found that extreme individ- 

 uals were to a large extent absent, and I conclude that it was because of their 

 being unable to copulate properly with the opposite sex. However, when 

 color w r as taken as a basis of selection, there was not the slightest trace of 

 preferential mating. 



It appears, therefore, that selective mating exists in these beetles, but that 

 it is not found in any character excepting size, and that even here it is due 

 solely to the inability of abmodal individuals to properly perform the sexual 

 act. We do not ordinarily realize how narrow are the limits within which 

 successful copulation can take place in insects, or how slight a variation is 

 sufficient to prevent the performance of the sexual act with such completeness 

 as to insure the leaving of progeny. The w r ork of Crampton has shown this 

 most clearly in Lepidoptera, where, fortunately, the genital organs can be 

 measured without the introduction of an enormous error. Although in Lep- 

 tinotarsa successful measurement is impracticable, it is certain that a like 

 slight variation in the size of the genitalia or in their position is enough to 

 eliminate the possessor thereof from participation in the perpetuation of the 

 species. Hence there results a most effective selective mating in which only 

 the modal individuals, or those in near-by classes, are able to leave any large 

 number of progeny, and whether the variations are heritable or not, the 

 extreme variations have little or no chance to be transmitted to succeeding 

 generations. It is noticeable in all of the seriations which have been made of 

 different characters in this genus that there is a constant high percentage of 

 the individuals in or near the modal class, nearly 75 per cent in the polygon 

 of variation being usually found in the modal class or in the classes imme- 

 diately above and below it ; and this is due largely, if not entirely, to the 

 selective mating which eliminates in reproduction the perpetuation of the 

 more extreme conditions. 



In the preferential or assortive mating in these beetles we must therefore 

 recognize a selective process which acts to keep the species as close as possible 

 to the mode or standard by the elimination of extremes. Hence it is a con- 

 servative factor, and not one that would tend toward the production of modifi- 

 cations. If, however, from any cause whatsoever there should occur a rapid or 

 slow shifting of the mode of the species, preferential mating would still keep 

 it closely aggregated about the new mode, and thus bring about a modification 

 in the entire polygon of distribution. In this way selective mating has un- 

 doubtedly been and is now a valuable aid in the formation of races and spe- 



