STUDIES OF VARIATION IN INSECTS 



279 



and only one half as abundant as the typical soror in 1895, is 

 now (1901, 1902) two thirds as abundant.* 



Fig. 53. Frequency pohgon of the variation in eljtral pattern of 905 speci- 

 mens of the flower beetle, Diabrotica soror, collected at Stanford University, 

 1895. 



A pertinent question in this connection is that touching the 

 cause of this increasing frequency of a certain variation in this 

 species. Is the form with both middle spots fusing to form a 

 broad transverse bar being naturally selected? Or, in the light 

 of our observations of Hi^^odamia with its great variety of 

 color pattern among individuals exposed for a season to the 

 struggle for existence, are we not to consider seriously the pos- 

 sibility of the condition in Diabrotica being brought about by 



* Since this paper went to press, 905 individuals of D. soror, collected Oct., 

 1904 from same locality as series of 1895, 1901 and 2, show the following per- 

 centages of the various patterns : 20.9 with all twelve spots separate; 615.4 with 

 spots of middle pairs fused; 13.5 with miscellaneous patterns. 



