Y. TANAKA is the first author who discovered the linkage groups in the silkworm, Boanbyx mori, and a remarkable fact that no crossing over takes place in the female of this insect. According to his experiments, the S (striped. larval marking) and Y (yellow blood color) are linked with each other, the coupling ratio being very near to a 3 : 1 ratio, which corresponds to 25 crossing over. He noted however, that the ratio is susceptible to considerable variation, it came so low as 2 : 1 at least in a single family. In the review of TANAKA'S paper, STURTEVANT suggested the presence of a modifying factor as a cause of variation in the crossing over value. In order to study the influence of X rays upon the crossing over value, I started my experiments with a. single F1 egg-batch produced by mating between a striped yellow female (race Cltinayang, a Chinese bivoltine) to a normal white male (race Yamatoniskiki, a Japanese bivoltine). Both races had been obtained from different sources from those of the strains used by TANAKA, and no blood relation could be suspected between them. When the F1 striped yellow males were back cro ssed to the double recessive females, I found that c. o. v. (crossing over value) is markedly different in different families, and that it might be desirable to study, at first, the nature of the variation in question. I selected "high" class, i. e. the families that showed a high value of crossing over, and "low" class, i. e. the families that showed a lower crossing over percentage, in F1 generation, and the mating was accomplished among the same class respectively. Thus I continued the selective breeding so far as to sixth generation. It goes without saying that the c. o. v. of the SY/sy individuals in a certain generation is calculated from the result of the next generat on obtained by mating them to the double recessive sy/sy females. The "high" series maintained a high c. o. v. in succes s ive generations, segregating no "low" class from it. The average c. o. v. was alw a ys low in the "low" series, but it often segregated families of intermediate or still higher class. The intermediate class produced the offspring of high, intermediate and low values, roughly in a 1 : 2 : 1 ratio. From the results mentioned above, we may assume a dominant modifying factor CII which reduces the S-Y crossing over value, in a homozygous condition, by 12 %. Consequently c. o. v. in males with the genetic constitution SYCII/syCII is about 12 %, while it is about 24 % in normal case or in SYcII/sycII. Owing to the imperfect dominancy of CII the individuals heterozygous for this factor give rise to an intermediate c. o. v., i. e. 17-18 %. Besides CII gene, which I may call the major modifier, I have found one or more minor modifiers that influence, but in a slight degree or by 1-4 %, the c. o. v. of the "high" class as well as that of the "low" class. Seasonal fluctuation of c. o. v. is also noted. There is some indication that the essential factor of the seasonal fluctuation is the influence of temperature, which is especially striking in the "high" class. The average c. o. v. of the "high" class was 21.48 % in 30℃, while it was 25.86 % in 19°C, and it was intermediate when reared in an intermediate temperature. Thus it is evident that high temperature reduces c. o. v. between the striped and yellow. The influence of temperature is hardly perceivable in the "low" class.