1. In order to test our previous conclusions concerning the tumor rate in hybrid strains, we carried out additional hybridizations. For this we selected strains which differed markedly in their tumor rate and which had been followed through a number of generations and found constant in their behavior. Hybridizations between strains or families of a similar (either high, medium, or low) tumor rate served as control experiments. In these cases the offspring showed a tumor rate similar to that of the parents. 2. In selecting for hybridization various groups of Cream mice representing a very low tumor strain and English Sable mice representing a high tumor strain, we obtained in the majority of cases hybrid strains with a tumor rate intermediate between that of both parent strains. In a few instances it approached somewhat the high tumor strain of the English, and in a few others the low tumor rate of the Cream. 3. In several hybrids between the high tumor strain English and the low tumor strain Silver, the latter of which was split off from the English, the high tumor rate of the English prevailed. A number of mice which served for hybridization were followed throughout their life and found to behave typically as to tumor rate. 4. If we omit the strains in which both parents had a similar tumor rate, we found the higher tumor rate to be dominant in twenty-four hybrids. In seventeen of these the mother strain dominated, and in seven the father strain. In nineteen hybrid strains the tumor rate was intermediate. In ten strains the lower tumor rate dominated. In eight of these the mother strain prevailed and in two the father strain. The low tumor rate was, therefore, dominant in approximately 18 per cent of the strains. 5. There does not seem to be a fixed rule as to dominance in the tumor rate. In a considerable number of cases, and especially in well analyzed cases, the result was intermediate. 6. Altogether in twenty-five of our hybrid strains the mother strain, and in nine strains the father strain prevailed. In nineteen strains the result was intermediate. We conclude that both father and mother strain may dominate and that the tumor rate is not in the strict sense of the term a sex-linked character. However, the fact that the mother strain prevailed in a much larger number of our cases than the father strain, and that several times (but not in all cases), in reciprocal crosses, the hybrids followed the tumor rate of the mother strain, suggests the possibility that as far as the hereditary transmission of mammary cancer in mice is concerned, the mother may be more potent than the father, and that perhaps under certain quantitatively varying conditions the mother strain may dominate over the father strain. This statement is merely a tentative conclusion at the present time and needs further investigation. 7. The results of these investigations confirm our previous conclusion that in the majority of the crosses which we observed, the cancer rate is either intermediate between those of father and mother strain, or that it follows the tumor rate of the parent with the higher rate and only in a relatively small number of instances the cancer rate follows that of the parent strain with the lower tumor rate. On the whole, the heredity of cancer rate and cancer age follows the blending type of hereditary transmission. 8. While there is a distinct relation between high tumor rate and early cancer age, our observations make it probable that cancer rate and cancer age are to some extent independent of each other. 9. On the whole the different generations, including F(1) and F(2) of the various hybrid strains, showed a concordant tumor rate and tumor age. 10. If we consider, then, our results as a whole, we may conclude that in crossing strains which differ in their tumor rate no rule of dominance which applies equally to all cases seems to exist. In a certain number of crosses the results are undoubtedly intermediate. In these instances the tumor rate and to some extent also the tumor age behaves in a manner similar to characters which differ in father and mother merely in quantity as in the length of organs. From such intermediate results all kinds of gradations exist, leading on one side to dominance of the strain with the higher tumor rate and on the other side to dominance of the strain with the lower tumor rate. However, in our experiments dominance of the strains with the higher tumor rate greatly predominated over the opposite extreme. 11. Our results on the whole are, therefore, in certain respects comparable with the inheritance of sex which Goldschmidt studied in hybrid strains of the gypsy mother. Here also all gradations from the male to the female were observed in the offspring. Goldschmidt assumes that in different hybrids there are created different quantities of certain substances which like enzymes determine according to their quantity the velocity of chemical reactions and the amount of certain substances produced. The latter determine in the hybrids the quantitative variations in the character which is analyzed. According to Goldschmidt, multiple allelomorphs, which in our experiments seem to determine the heredity of spontaneous cancer, depend upon differences in the quantity of a substance present in the different individuals or varieties. In whatever way we may conceive of the character of multiple allelomorphs, our results make it probable that multiple factors are involved in the heredity of cancer in mice.