CROSSING EUSCHISTUS VARIOLA RIUS AND EUSCHISTUS SERVUS. 357 



cancelled, by the £act> that the two generations are not equally represented, 

 and cannot therefore be justly compared. We have 190 males of the F 

 generation in which the exact condition of the spot can be determined, and 

 only 11 males of tlie Fi generation ; and therefore it would not be reasonable 

 to conclude from such inadequate data, that the Fj generation of these 

 hybrids is less vnriable than the Fg generation. We are inclined to believe 

 that if the number of the Fi hybrid males were equal to that of the Fg 

 generation, they would show both extremes of the inheritance of the spot, 

 and a variability quite as pronounced as we find in the Fg males. 



Some of the males of the Fg generation might be assumed to be pure 

 dominants and recessives, for some have the spot as fully developed as that 

 of pure rariolarhis (one or more of the specimens of photos 15, 23, 26 28 

 32, 34, 36, 40, 41, 46, 48, and 55), while others have no spot whatever 

 (one or more of the specimens of photos 16 to 39, 43, and 51 to 56). 

 The intermediates show the spot in a very variable degree of intensity, but 

 if these latter are assumed to be " visible heterozygous forms '' of Mendelian 

 segregation, there should be an equal number of pure dominants and reces- 

 sives, whereas there are only 19 that have the perfect mriolarius spot, and 

 74 in which the spot is absent. 



If we would attempt to find a theoretical explanation of the results we 

 must first devise a formula that will work for pure variolarius, and the facts 

 compel two assumptions in the making of such a formula. First we must 

 assume that the female is homozygous for the spot factors, and second that 

 she has an inhibitor for the spot and that she does not transmit this inhibitor 

 to her male offspring. The necessity for this last assumption seems obvious, 

 because the spot is never inhibited in the male. If we would assume 

 however, that both sexes can have an inhibitor, an assumption that some 

 facts in the hybrids demand, then we must assume that it requires a double 

 dose to inhibit the spot ; and to insure the spot being a constant character 

 we must assume that in the male the inhibitor is not only heterozygous but 

 sex-linked. But this assumption, which would hold for variolarius, would 

 not work for the cross, because the female hybrids would in this case be 

 heterozygous instead of homozygous for the inhibitor, and would therefore 

 have the spot. If to obviate this difficulty we assume that servus also has 

 an inhibitor, this would inhibit the spot in the Fi hybrid females, but in 

 none of the Fjt hybrid males, whereas in fact, the spot is completely or 

 partially inhibited in all the Fi hybrid males. 



The necessity of the assumption that the female variolarius is homozygous 

 for the spot factors is quite obvious, for the genital spot could not remain 

 constant if we assume that the female is heterozygous for the spot factors 

 unless we add the unwarrantable assumptions that the female is heterozyoous 

 and the male homozygous for sex, and that the spot factors are linked with 

 the sex-factor. 



30* 



