The purpose of the present study is to scrutinize the character of twins seen from the physiological aspect by the analytical measurement of their G. S. R. which appeared under the situation of mental stimulus, and to confirm the intra-pair mutual resemblance based on heredity, and the degree of the influence of heredity-environment factors. The results have been compared with the findings about the heredity-environment factors involved in mental and physical qualities which have already been reported. The subjects studied are 38 pairs of twins (21 paires of identical twins, 15 pairs of fraternal twins, 2 pairs of unlike-sextwins) and 20 pairs of the comparative groups of unrelated individuals. The latter group was made up of children of the like-sex and of the same birth date, and were picked up from the class in which the selected twins were enrolled in school. The stimulation for the G. S. R. is constituted by 41 questions, orally put to the subjects, relating to the need and frustration, and those affairs and situation which were believed to arouse ellotion of fear and unrest.As for the intra-pair resemblance of the recorded curves of the G. S. R., calculations of the ration of the concurrence between the pairs were made on the basis of the existence of the response curves, the number of the curves corresponding to each stimulation, and the degree of the general intrapair resemblance in the recorded response curves. Moreover, about the ratio of the response curves to the number of stimulation, and the number of responses, the time of reaction, the degree of provisional responses corresponding to each stimulation which were all treated by the definite method, the intra-pair mean percentile deviations (E) and the ratio of heredity and influence of the environment of the twins based on F. Lenz's formula, and the intraclass correlation (r') by the analysis of variance were calculated in order to examine the heredity factors by comparing the resemblance between the oviparous distinctions.1) The ratios of concurrence between the pairs on the basis of the existence of the response are as follows : 77.6% in identical twins, 70.1% in fraternal twins, 63.0% in the comparative group. The ratios of concurrence about the number of responses corresponding to each stimulation are as follows : 47.6% in identical twins, 41.2% in fraternal twins, 32.9% in the comparative group.2) Through all the diagnostic items, the identical twin group showed the smallest, the fraternal twin group comparatively large, and the comparative group the largest, mean percentile deviations (E). (Table 3).3) The hereditary numerical values calculate by the Lenz's formula from E are as follows : Ratio of response 1.55, Number of response .60, Time of response 1.41, Degree of provisional response .31. But, when the results of the calculated mean probable errors of these numerical values are taken into account, the hereditary factors are not recognized to be greater than the environmental factors.4) The intraclass correlation coefficients obtained through the analysis of variance are : .442 - .713 in identical twins, .227 - .513 in fraternal twins, -.115 - .321 in the comparative group. The r's in the identical twins has the significance of at 1% level, and the numerical value of 70% of r's in fraternal twins has the significance at 5% level, but r's in the comparative group have no significance even at 5% level. (Table 4).These results show nearly the same thing as the results of the Kraepelin Numerical Addition test, The Adjustment inventory, The Moral judgement test and Rorschach test, but when compared with the results of the finger print, physical qualities, intelligence and scholarship, faculty of bodily movement and special faculties (the results of which have already been reported by the present author), the intra-pair resemblance is in a low state.