110 H. E. EWING. 



But the most noteworthy case of erratic paedogenesis was 

 found, however, in the offspring of 11435, a normal wingless 

 adult, that left 13 young. Of these 13 young, 9 reached the 

 stage of maternity. Of these 9, 3 were paedogenetic,u 4 gi, 11492, 

 and 11495. These paedogenetic nymphs were all of the winged 

 form, and reached their stage of reproduction along with the 

 adults of the wingless form, which is much sooner than the time 

 allowed for the development of the normal winged adult forms. 

 The psedogenetic nymphs were all large, vigorous individuals, 

 and reproduced normally. The individual n 49 i gave n young 

 before she was killed. Of these young, i was reared to maturity, 

 producing a normal wingless adult. Another, 11492, gave 14 

 young before being killed,jand of these, 13 were reared to maturity, 

 producing in every case normal wingless adults. The third 

 individual, 11495, gave birth to 16 young before being killed, of 

 which 4 were reared to maturity, all being normal wingless forms. 



Other instances of erratic psedogenesis were met with, but 

 these suffice to show the nature of their appearance, and the 

 fact that this tendency toward paedogenesis is not inherited. 

 Paedogenesis in this pure line appeared to be due to the arrest- 

 ment of the growth of the immature individual, the development 

 and functioning of the reproductive system being unhindered. 

 It was probably induced by low temperature changes during a 

 critical period in the development of the individual. At any 

 rate adverse conditions must have been responsible for this 

 arrested development. 



SUMMARY OF RESULTS. 

 In Regard to Heredity. 



I. Six different fluctuating variations in a parthenogenetic 

 pure line of Aphis avence Fab. were tested to see if there was any 

 summation effect of selection on these variations. Selections 

 were made for ten or more successive generations in the case of 

 three of these characters ; for forty-four successive generations 

 in the case of one character; and were carried out in both plus 

 and minus directions in the case of two characters. In all of 

 these cases no summation effect was produced by selection. 



