PARTHENOGENETIC PURE LINE OF APHIS AVENVE FAB. 63 



TABLE II. 



SHOWING MEASUREMENTS OF CORNCILES OF INDIVIDUALS OBTAINED IN ISOLATION 4, 



AND COMPUTATIONS OF THE FRATERNAL MEANS (THE AVERAGE RATIO OF THE 



CORNICLE LENGTHS TO WIDTHS FOR ALL MEMBERS OF A FRATERNITY). 



Selection Made for Increasing the Body Length, Isolation No. 5; 



and also Selections Made in a Check Isolation (No. 6} for 



Decreasing the Body Length. 



Because stature, or body length, is one of the most commonly 

 studied of the fluctuating variations, and also because it is a 

 character well known to be inherited in some of our higher ani- 

 mals, I decided to see if by continued selection in a pure line it 

 could be affected. Experiences with isolation No. 3, having 

 taught me that size was influenced easily by temperature changes 

 and probably by other changes also, I decided to eliminate errors 

 from these sources by running a check isolation, or subline. And 

 in order to test more rigorously the effects of selection, in this 

 check strain the shortest individuals were picked out in each 

 instance. Selections were begun in the twenty-sixth generation 

 in each isolated subline, and continued until the thirty-second 



