218 



The Mechanism of Evolution in Leptinotaksa 



Corresponding lines from the average and the large weights were also carried, 

 but are not shown in the figure because they would only complicate it with the 

 larger number of curves and add nothing to the results in that they were precisely 

 similar to those shown in the A series. 



•"fS 



30 



20 



10 



(30 



FW 



20 



rao 



20 



F, ( 



Fi < 



30 



20 



10 



Fig. 31. — Schematic representation of resuits obtained in thie attempt to breed 

 weight as a single character, showing distinctly that it was not possible to purify 

 this material with respect to weight and reduce it to uniform pure lines. 



In the years 1907, 1908, and 1909 I made a similar set of determinations upon 

 races of L. signaticollis from the Eancho Basosco location, these being bred out 

 in line cultures and selected on the basis of Aveight. The general result was the 

 same as in the two preceding instances, and there were no results but a lot of 

 values subtended by the animals in question and having no certain relation to the 

 organism. 



