INHERITANCE IN HYDATINA 81 



equal in viability, and the two reciprocal crosses were approxi- 

 mately equal. 



Variability of the duration of the egg stage. The effect of cross- 

 ing on the variability of the duration of the egg stage was tested 

 in two successive generations, beginning with a cross between 

 lines C and D. The standard deviation of the hatching time 

 for inbred eggs of these two lines were widely different (1.09 

 and 7.14 respectively). The standard deviations of the two 

 reciprocal crosses were intermediate between those of the inbred 

 eggs, but were very unequal (1.^7 and 5.64); each cross was 

 nearer its female parent, in point of variability of hatching time. 



This inequality existed only in the egg stage, however, for 

 when the two reciprocal crosses {C X D and D X C) were in 

 turn inbred and reciprocally crossed, the variability of hatching 

 time did not materially differ in the four lots of eggs (standard 

 deviation being 2.30, 2.42, 2.19 and l.bo, respectively). 



2. The effect of selection on viability of eggs. In hybrid lines. 

 In the lot of eggs designated (C X D) X (C X D), which were 

 only once inbred from an Fi (and heace hybrid) parthenogenetic 

 line, selection was made for families of high and low viability, 

 respectively. From the selected eggs, parthenogenetic lines were 

 reared, and among their inbred eggs, selection for families of 

 high and low viability was again made. Three successive selec- 

 tions resulted in an average difference of about 21 per cent in 

 the viability of the eggs; but practically all this difference was 

 produced by the first selection, and was not thereafter appreciably 

 increased. 



Selection in relatively 'pure' lines. Three successive inbreed- 

 ings, as described above, must have changed many heterozygous 

 characters to the homozygous condition, if random segregation 

 and recombination occur; hence the parthenogenetic lines L3 and 

 H3 must have been relatively 'pure,' that is, homozygous. Was 

 the difference in viability between L3 and H3 fertilized eggs a 

 permanent product of selection, or had it been maintained only 

 by continued selection? 



Random selection from L3, without reference to the viability 

 of the parents, resulted in a lowered viability in L4. This may 



THE JOCHNAL OF EXPERIMENTAL ZOOLOGY, VOL. 15, NO. 1 



