Masao So and Yoshitaka Imai 327 



occurs before their birth, though at that time they did not prove it by 

 direct evidence. 



Recently Kirkham (1917) and, independently, Ibsen and Steigleder 

 (1917), however, have confirmed this view from their embryological 

 studies, where the homozygous yellow mice were expected to perish. 

 They brought forward evidence of marked mortality among the embryos, 

 especially in the early stage or soon after the eggs attached themselves 

 to the mother's uterine membrane. 



As was indicated above, the modified ratio obtained in the " Daruraa " 

 X "Daruma" cross is entirely similar to that of the yellow mice. Accord- 

 ing to normal Mendelian expectation one-third of the " Daruma " young 

 born in the cross " Daruma " x " Daruma " may be expected to be 

 homozygous in regard to the D factor. So far as the writers have 

 tested such animals, however, none of them proved to be homozygous. 

 Thirty of them were used for this purpose, and all of them were found 

 to carr}' the factor in the heterozygous condition. Formerly Little, 

 also testing twenty-one of them coming from a similar origin, obtained 

 the same results. If this number is added to the writers' figure we 

 may safely conclude that the " Daruma " mouse carrying the D factor 

 in the homozygous condition is either never bom or matured. It may 

 be possibly thought that as the mortality occurs soon after birth, yet 

 before the time at which the pattern character can be recorded, one- 

 third of the " Daruma " mice meeting with such a fate may not be 

 actually recognized. To answer this question the original record was 

 carefully examined, but the data led to the negative conclusion. So 

 there seems to be no mistake in saying that the homozygous " Daruma " 

 race do perish at the embryological stage like the yellow mice, though 

 direct evidence supporting this view is not in hand. 



Mating : " Daruma " x Self. 



When homozygous selfs are mated with the "Daruma" race we 

 should expect equal numbers of " Kasuri " and self in the subsequent 

 generation, as the latter parent is invariably of heterozygous genetic 

 composition. The actual numbers obtained from these matings are 

 represented in the following table : 



The number is rather small, but the results evidently agree with 

 expectation. 



