Significant differences have been found in the spermatocyte chiasma frequencies of ten strains of mice differing in genetic susceptibility or resistance to the spontaneous occurrence of malignant tumors. In three resistant and five susceptible strains there is a positive correlation between susceptibility and low chiasma frequency, and between resistance and high frequency. Two interrelated strains may be exceptional, but their cancer and leukemia records are not clearly established, and one of them, though classed as resistant, is somewhat anomalous genetically.The study comprises a total of 16 tests, each made on 25 complete cells. It was made as an indirect test of a working hypothesis on the mechanism of genetic susceptibility. This hypothesis is based on: (a) the assumed validity of the somatic cell mutation theory of cancer in its most general formulation, (b) the conclusion that gross chromosome irregularities commonly found in tumor cells cannot in themselves be the cause of cancer, though they may be a direct result of some less obvious lack of correlation in the mitotic process, (c) the fact that in different organisms the existence of genes controlling chromosome structure and behavior and other aspects of cell division has been established, (d) the observation that in tumors themselves the chromosomes are commonly found to be "split" precociously, and (e) the observation that "splitting" of meiotic chromosomes is correlated with the absence of chiasma formation. The hypothesis postulated that in strains of mice genetically susceptible to the spontaneous occurrence of tumors, though not yet having cancer, the chromosomes might be found to be "split" slightly in advance of the normal period, and that this could be measured indirectly, but most precisely, by its anticipated effect of causing lowered chiasma frequency at meiosis. The precocious "splitting" would itself be a primary result of some genic difference in the constitution of susceptible animals and might either—through causing grosser chromosome irregularities—lead directly to neoplastic growth, or be a primary morphological concomitant of an underlying cause common to both phenomena.