PARTHENOGENETIC PURE LINE OF APHIS AVENGE FAB. 89 



mean of the fraternity for the sixtieth generation is found to 

 represent a length slightly greater than that of the parent of 

 this fraternity. 



Is there any evidence of inheritable effects of selection in this 

 long pedigree of 44 generations? Our only conclusion must be 

 in the negative. There is not the least evidence of a general 

 trend upward in the fraternal means as we advance generation 



TABLE IX. 



CONTINUATION OF DATA FOR ISOLATION n. LAST FOUR GENERATIONS. 



Individual Length in Millimeters Fraternal Mean. 



II84I 1.47 



11845 I.I7 



1 1 846 1.51 



IIS49 1.24 



118410 1.27 



118411 I.4O 



1-343 



11852 1. 54 



11 853 I-36 

 II8B4 1-47 

 IIS55 I-IO 

 I I 856 1.36 

 II857 1. 01 



1.307 



iiseib 1-54 



Il862b 1-39 



Hs63b 1.40 



1-443 



1 1 875 -99 



iisv6 0.97 



1187? I-36 



1.107 



after generation in this isolated subline. The fluctuations are 

 many and at times great, but the series continues on in the same 

 general way. These fluctuations have been partially analyzed, 

 and in those cases where extremely low or high variatons 

 occurred they were found to depend upon environmental con- 

 ditions (temperature and food supply chiefly). If there is a 

 cumulative effect in selecting extreme variants of each fraternity 

 for a large series of generations, the results obtained in this long 

 series of my pure line of Aphis avena Fab. fail to show it. 



