Abstract
1. As mentioned in the preceding paper (3), hybrid sterility among rice varieties seems to be due to that genes necessary for the development of gamete are duplicated and a hybrid between varieties having different recessive genes produces double recessive or sterile game tes in a deflnite ratio. When a gene (A or a ) controlling a character was linked with such a gamete-developnlentgene, the segregation ratio of A : a would be modified aild it should be expected that sterile plants be higher in frequency among heterozygotes Aa than among homozygotes AA or aa. 2. The relation between the gene above mentioned and the gamete-development-genes is inferrable from a certain method of calculation based on the phenotypic ratio of A : a and the differences in the mean as well as in the variance of fertility between phenotypically dominant plants and recessive ones in the F2 or back cross of a sterile hybrid. This was verifledin dealing with the change. of segregation ratios of grain tip pigmentatioh gene and glutinous gene in some varietal crosses.
