360 MISSES K. FOOT AND E. C. STROBELL : RESULTS OF 



have the spot as strong as variolarius, 3 have a spot so small and faint that it 

 is scarcely perceptible (the lower two spechnens of photo 62, and the lowest 

 specimen of photo 66), and 3 are quite typical intermediates (the 3rd and 

 5th specimens of photo 64 , and the second specimen of photo 66). Thus 

 12 of these males have the variolarius spot, 3 are intermediates, and 3 are 

 almost without a spot. It therefore seems obvious that a Mendelian ratio is 

 not shown in any of the experiments, it does not appear in the Fi hybrid 

 generation, in the Fg hybrid generation, nor in this generation from an 

 Fj hybrid $ by pure variolarius cf . 



The above-mentioned cross (Fj hybrid ? X jiure variolarius ^) is im- 

 portant because it shows that the genital spot is not a sex-linked character. 

 The test of whether a character is linked to a factor that determines sex is 

 usually made by the " reciprocal cross, '^ which in this case would be pure 

 servus ? X pure variolarius ^ . It is assumed that the parent that is hetero- 

 zygous for sex can transmit a SL-x-liuked character only to the opposite sex, 

 and therefore none of the males from such a cross should inlierit the spot. 

 We were unable to attempt this cross, but the above-mentioned experiment 

 (Fi hybrid $ x pure variolarius S ) ^hows quite as conclusively as a " reci- 

 procal cross " whether the genital spot beliaves as a sex-linked character. 

 This cross demonstrates that the genital spot does not behave as a sex-linked 

 character, for it shows that the male variolarius can directly transmit the spot 

 to his male offspring, for these offspring show the greatly increased strengtli 

 of its inheritance from the pure male as compared with its inheritance from 

 the Fi males. 



We cannot explain this exclusively male character by simply assuming 

 that the constitution of the eggs inhibits the expression of the spot in 

 the female, for this leaves unexplained the fact that the spot is partly or 

 wholly inhibited in the Fx males. 



We have been unable to harmonize our results either with the Mendelian 

 or non-Mendelian (blend) type of inheritance — the great variability of the 

 Fi hybrids being the most obvious difficulty. The assumption of multiple 

 factors may be satisfactory as an explanation of the variability of the Fg 

 intermediates; but it does not explain the variability of the F, hybrids. 

 On the assumption of multiple factors the F^ hybrids should be alike — 

 subject only to minor variations (fluctuations). This_, however, is not the 

 case — 2 have no spot and 9 are variable intermediates. 



Castle's assumption of change in potency of a given factor or factors 

 seems more in harmony with the facts, for this offers not only an explana- 

 tion of the variability of the Fg intermediates, but admits any degree of 

 variability in the F^ hybrids — even to the extent of reversing dominance. 



