230 



The Mechanism of Evolution in Leptinotaksa 



has become a dumping-ground for differences of this general class. My first 

 proposition that " environment '^ was responsible for the range seems to have 

 been answered negatively in a certain and emphatic manner. 



Two possibilities remain : either the range is the product of " impurities " 

 in the substances or due to minute differences in the constitution of the material 

 which the methods used did not entirely isolate. I sought to test the question 

 of the possible existence of minute " germinal variations " which might be 

 small differences in the composition of the substance itself, through mating 

 individuals of identical index, to determine if it were possible to obtain within 

 the group minutely characterized races. 



Table 32. 



No. of mating. 



Parents, 

 index. 



Progeny. 



Fraternity No. 



Range of index. 



Totals. 



10. g VI(BA)A 



10. g VI(BA)B 



10. g VI(BA)0 



10. g VI(BA)D 



10. g VI(BA)E 



10. g VI(BA)F 



10. gVn(BAA)A .... 

 10. g VII (BAB) A .... 

 10. gVn(BAC)A .... 

 10. gVII(BAD)A .... 

 10. gVn(BAE)A .... 

 10. gVn(BAF)A .... 



gVn(BAA) ., 

 g Vn(BAB) . . 

 gVn(BAC) . 

 gVII(BAD) .. 

 g Vn(BAE) . . 

 g VII (BAF) . , 

 g VIII BAAA 

 g Vni(BABA) 

 g Vin(BACA) 

 g Vni(BADA) 

 g Vni(BAEA) 

 g Vra(BAFA) 



fM. 

 IF 

 f M. 

 IF. 

 JM. 

 IF. 

 (M. 



If. 



(M. 



If. 



f M. 



If. 



(M. 



If. 



fM. 



If. 



fM. 



If. 



fM. 



If. 



fM. 



If. 



f M. 

 IF. 



59 



30] 

 29 J 



37] 

 22 j 

 341 

 43) 

 32) 

 35 



59 



77 



67 



40) 

 45) 



5?! 

 17) 

 25 j 



86 



321 67 

 25 j ^^ 



Total 934 



In October 1907, I mated from stock 6, line 10 g VI (BA) (table 33), 6 

 pairs, when both parents had the same index, but reared under the conditions 

 of the general culture quarters. The progeny obtained from these pairs matured 

 at the end of November and showed some indication of the influence of the 

 method that had been practiced. Early in December I again mated from the 

 same lines pairs having the same indices as in the first instance, and obtained 

 a second generation early in January 1908, in which there were unmistakable 

 indications of the action of the operations in some of the lines, but by no means 

 in all, a condition that might mean one of several possibilities. All hibernated 

 from the middle of January to the first of March 1908, when they were forced 

 from hibernation and mated. In table 33 I have shown the conditions of the 

 series in the two generations of 1907. 



