This chapter describes a continuation of a previous study (12) that appeared in the Twenty-Seventh Yearbook of this Society. The earlier report was based on intelligence measurements of husband, wife, and two or more children in each of 105 families. The present study is based on 997 cases in 269 family groups, including the previous 105 families, and 164 additional families of two or mare members.Studies of familial resemblance impose no direct control on either heredity or environment, and for this reason cannot yield any direct answer to the problem of the relative contribution of heredity and environment to individual differences. 2 In spite of this, such studies 1 Assistance in the preparation of these materials was furnished by the personnel of Works Progress Administration Official Projects Nos. 65-3-5403 and 65-3-5406. In connection with the field work, acknowledgment is due to the Social Science Research Council of Columbia University for a grant in aid and to Professor A. T. Poffenberger as representative of the Council. 3 We do not wish to overemphasize the difference between the biometric and experimental approaches in this field, particularly where human traits are concerned. The biometric studies are never entirely without the equivalent of some experimental control, and the experimental studies never completely achieve the desired control of all relevant factors. Thus, the present biometric study, having as its locale a comparatively homogeneous rural environment, achieves some control of interfamilial environmental variations. The studies by Pearson (18), Elderton (6), and Wilcocks ( 22) have exploited this type of experimental feature of purely biometric studies; unfortunately, these investigators overlooked the importance of the ratio of intrafamilial to interfamilial environmental variation (10), and the possibility of a significant positive correlation between the two.