IO2 



H. E. EWING. 



by the side of each antennal drawing is a heavy line representing 

 the length of the individual possessing the antenna, also magni- 

 fied to the same degree as the antenna itself. In the case of 

 individual 7.31! the antenna is practically of the same length as the 

 body. In the case of 7si7 the antenna is considerably less than 

 two thirds as long as the body. All intergrades were found 

 between these two extremes, showing that these variations are 

 only fluctuations. 



FIG. 13. Camera lucida drawings, all of the same magnification, showing ex- 

 tremes of fluctuating variation in the lengths of cornicles trom three individuals 

 (93512, 9356, 9362) of the 936 fraternity. The heavy black line in each case represents 

 the body length, same magnification, of the individual whose cornicle is drawn 

 by the side of it. 



Fig. 13 shows us a similar variation in the length of the cor- 

 nicles in comparison with the length of the body in a single 

 fraternity. The cornicles of individual 9 36 I2 (the left of which 

 is shown at top of Fig. 13) are 25.5 per cent, of the total body 

 length of the same individual. The cornicles of individual 9 3 e2 

 (one of which is represented at the bottom of Fig. 13) are only 

 17.3 per cent, of the body length of this individual. 



The variation in the relative size of contiguous parts is shown 

 in Fig. 14, where we have represented camera lucida drawings, 

 all of the same magnification, of the third and fourth segments of 

 the antenna? of three different individuals of the F 9 fraternity. 

 Here the large segments are the third and the small ones the 

 fourth. Observe that there is almost no variation in the length 



1 The letter F is here used to denote the subline or isolation, and refers to isola- 

 tion No. 3 (only one isolation was running during the ninth generation). 



