148 



The Mechanism of Evolutiok in Leptinotaksa 



From pairs A and D matings were made and began breeding early in July, 

 giving adult progeny that had all emerged by September 12, and then hiber- 

 nated. These F3 fraternities gave 416 larvae and 318 adults in all respects like 

 the populations in F^ and Fj, and one line was continued to Fg without any 

 indication of change either ia the characters or reactions of the line. 



Efforts were made to cross this line with L. diversa and with L. signaticollis 

 as well as L. undecimlineata without success, but it crossed feebly with L. decem- 

 lineata. This cross, made in the effort to break up the combination, showed 

 not the breaking up of the combination but its dominance in F^ and the separa- 

 tion in Fj of pure decemlmeata, and the pure combination, which could be 

 easily distinguished and tested by their rate of development and the limitation 

 of food. In no test did I succeed in breaking up the combination that had been 

 formed by the cross. 



Table 18. 



Both of the series that have been observed behaved alike, were in all discovered 

 respects the same, and neither yielded to any of the efforts to disrupt the com- 

 bination. That it was a combination there is no doubt. The virginity of the 

 females was certain, the fact of the cross equally sure and proven by the 

 presence in the combination of characteristics that could only have come, under 

 the conditions of experiment, from the male line, and in all respects the product 

 of the cross was distinctly a combination of the lines, even though the form and 

 usual taxonomic characters were of the female type entirely. The length of 

 ontogeny, the food limitation, and the many peculiarities of action that came in 

 with the male line showed fully the existence of a combination of the parental 

 lines into a complex that did not dissociate in the production of Fj or subse- 

 quent generations. A new and stable type had arisen by this combination that 

 was stable from the start, existed as a group from the start, that with its slow 

 progression in ontogeny and limitation in food would have, in nature, effectually 

 isolated it from possible intermingling with the female-parent species, while it 

 was imable to cross with the male-parent stock. 



It is not difficult to imagine the advent of a female L. decemlineata into 

 the habitat of L. diversa, and perchance if the crossing took place there would 

 be produced in the location a new species, if its origin were not known, that 

 would feed with diversa upon the same food plant, but not cross with it, at least 

 with any ease; that would be by food-relations, growth-rates, and other agents 



