PARTHENOGENETIC PURE LINE OF APHIS AVENJE FAB. 1 03 



of the fourth segment, but a great variation in the length of the 

 third. In F 9 i6 the third segment is not more than a fourth 

 longer than the fourth segment; in F 9 I5 the third segment is 

 almost twice as long as the fourth one. 



Variation in size is probably as noticeable as any other of the 

 fluctuating variations. In Fig. 15 we have drawings of the 

 same magnification of two individuals obtained in isolation II of 



,1 





FIG. 14.% Camera lucida drawings, all of the same magnification, showing the 

 extremes of fluctuating variation in the length of the third as compared with the 

 fourth segment of the antenna of three individuals of a single fraternity (Fa). The 

 third (long) and fourth (short) segments of an antenna of a single individual are 

 drawn^together to same scale. 



the pure line of Aphis aventz. The larger figure is of individual 

 I I 50 i, length 1.99 mm.; the smaller one of 11538, length 0.99 mm. 

 In other words one individual is slightly more than twice as long 

 as the other, and from this we should judge it to be about 8 times 

 as heavy (the two not being weighed). These are the extremes 

 obtained in 459 individuals, representing all those obtained during 

 the first 40 generations of isolation 11. 



Not only do we find fluctuations in regard to size in the indi- 

 vidual and its parts, but in shape also. In Fig. 16 we have 

 drawings of the right cornicles of three individuals of the 427 fra- 

 ternity. What a variation in shape! In the case of the right 

 cornicle of 427! we find the base almost twice as broad as the distal 

 end. In the right cornicle of 4 87 4 the base is no wider than the 

 distal end. This latter cornicle is strongly incrassate at its tip, 



