274 PRODUCTION OF RACES AND SPECIES IN LEPTINOTARSA. 



robust, healthy specimens, and were allowed to hibernate through the winter 

 in nature. In the following spring two males and one female emerged from 

 hibernation, the others not being able to pass through this ordeal. With this 

 material as parents, two sets of experimental cultures were started, one a 

 normal cross between m.ale and female pallida, the other between a male pal- 

 lida and a modal female decemlineata. 



The pure cultures of pallida gave in the first generation 30 adult male and 

 21 female pallida without any trace of reversion to the ancestral species. The 

 polygon of distribution as v/orked out for this generation is normal, without 

 trace of skewness, and is separated widely from that of the parent species. 

 From this generation I allowed 10 male and 14 female pallida to breed 

 together in one cage, and obtained from the culture 201 m,ale and 282 female 

 pallida, again without trace of tendency to revert to the condition of the 

 parent species. This third generation hibernated, and were unfortunately 

 almost entirely killed off by freezing, so that only 3 male and 4 female 

 pallida emerged from hibernation. I allowed these adults to breed freely 

 together as in nature, and obtained a fourth generation of 'j6 male and 81 

 female pallida, all true to type. From these 20 males and 20 females, repre- 

 senting the entire range of the variations found in pallida, were allowed to 

 breed freely, giving a fifth generation of 292 male and 306 female pallida. 

 The entire lot hibernated, and again a freeze so reduced the number that only 

 9 males and 7 females emerged in the spring. These were bred together 

 freely, giving a sixth generation of no males and 146 females true to type. 

 From these I selected 40 males and 45 females, covering the entire range of 

 variations found, as parents of the seventh generation, which gave every 

 promise of abundant and typical pallida, when the experiment was abruptly 

 terminated in July, 1904, by the accident which brought all of my experi- 

 ments to an end. 



In this series of cultures with pallida there is a striking constancy of form. 

 No selection of parents for each generation was made — only a reduction 

 because my quarters could not accommodate all. Moreover, free interbreed- 

 ing was allowed in each generation as in nature. Under these conditions 

 pallida showed no tendency to revert to the parental species nor any trace of 

 skewness of the polygons of distribution. 



In the first generation 10 male and 12 female pallida were placed in a 

 tank with 15 male and 15 female decemlineata, and all allowed to breed 

 freely. Crossings of male pallida X female decemlineata and male decem- 

 lineata X female pallida, as well as normal crossings, were observed, and the 

 proportion of normal crosses to abnormal was 7 to i. From this mix- 

 ture, allowed to breed freely as would be the case in nature, I got in the 

 second generation, L. decemlineata, 131 male and 114 female; L. pallida, 

 133 male and 162 female. These hibernated, and there emerged in the 

 following May, L. decemlineata, 10 male and 18 female; L. pallida, 9 male 



