510 



THE AMEBIC AN NATURALIST [Vol. LII 



lower half of Table III there is an example of this kind. 

 In the first generation from alate parents there is a 

 minority of males; in the second generation no males. It 

 is not improbable that, if a long line had been bred from 

 alate parents, there would be a progressive decrease in 

 the proportion of male offspring in the sexual phase of 

 the cycle. 



3. Tables IV, V, and VI also contain confirmation of 

 the conclusion drawn from the earlier tables, namely, that 

 at any given time winged viviparous parents produce 

 more wingless viviparous offspring than do wingless 

 parents, and that in the sexual phase males are produced 

 chiefly by the wingless parents, sexual females by winged 

 parents. 



Discussion 



Although the most striking results of the foregoing ex- 

 periments may appear to be the fact that winged vivip 

 arous females produce mostly wingless females in the 

 parthenogenetic part of the cycle and sexual females in 

 the sexual part, whereas the wingless viviparous females 

 produce chiefly winged females in the parthenogenetic 

 phase and males in the sexual, nevertheless the clue to 

 the explanation of this phenomenon is more nearly dis- 

 coverable in the progressive change in the frequency with 

 which all forms occur in successive generations. Thus, 

 there is a transition from a preponderance of apterous 

 females early in the cycle to a predominance of winged 

 females later. There is likewise, in the sexual portion of 

 the cycle, a transition from males to sexual females. 

 This latter transition has been demonstrated in the off- 

 spring of wingless mothers, and is indicated as probable 

 in the offspring of winged females. 



These transitions imply a gradual change of some sort, 

 presumably in the metabolism of the animals. While the 

 difference between a male and a sexual female, or be- 

 tween an apterous and an alate viviparous female, may be 

 a definite morphological difference such as a difference 

 in chromosomes, so that an individual is either the one or 



