132 



The Mechanism of Evolution in Leptinotaesa 



to 11 months, showing a considerable lengthening of the developmental period. 

 This result is attained regardless of the direction of the cross, and uniformly so 

 long as the conditions in the materials combined and in the medium are those 

 given. 



This cross and the results that come out are shown in table 16, in which are 

 given the parental conditions, the F^ results, and the breeding of undecimlineata 

 types in Fj or beyond. No purpose is served by presenting the breeding-test of 

 the mid-types ; enough has already been given to show the characteristic reac- 

 tions they present. 



The Fi undecimlineata types have been kept in the laboratory for gener- 

 ations ; they have been sent to Tuscon ; placed in isolated locations in Mexico, 

 and planted upon some of the Florida Keys. Nowhere have they shown any 

 change in the length of life-cycle or in the character of the form, or any 



Table 16. 



The values given are for the length of ontogeny, egg to adult, and the number observed. 

 Average length of ontogeny has been 90 days = A^. 

 Average elapsed time between generations, 316 days. 



other discovered change or return to the parental rates of development; nor at 

 any point in the series did there appear any trace of the signaticolUs type 

 coming out of the line. In Mexico, where they were introduced into a location 

 north of Jalapa in 1908, they persisted for at least two generations without any 

 change that I could discover; and material taken from the location in 1910 and 

 tested showed that the same condition was present as at the start. The type 

 appears highly resistant to differing external conditions and is in reality well 

 organized to persist in many different conditions of nature. 



In the laboratory this type has been subjected to most varied tests in the 

 effort to prove it to be in composition a masked heterozygote. When crossed 

 with signaticollis the reaction is, when it is obtained at all, a duplicate of the 

 reaction from which the type arose, with about equal numbers of the two types 

 present in F^, and the undecimlineata type that comes out in this test 

 breeds true. 



