While partial correlation affords an ideal approach to answering the question of causation, it has been used only to a very limited extent. The necessary data for comparing the closeness of relationship of various suggested causes of delinquency are not available and too few who are interested in social problems have appreciated the significance of the method. We should, therefore, lay especial emphasis on the measurement of the correlation of deficiency and criminality by Goring. He laboriously assembled the only data which are sufficiently extensive to allow much reliance to be placed upon their statistical reduction. In his use of correlation, moreover, he acted under advice from the main center for this work at the Galton Laboratory in London. If those who were “mentally defective” under Goring's designation were always convicted of crime and none of those who were not defective were ever convicted of crime, the measure of the relationship between criminality and deficiency would be expressed by a correlation coefficient of +1.00. If there were no relationship whatever between deficiency and criminality the coefficient would be 0.00. If the deficients were never convicted of crime and the non-deficients were always criminal the coefficient would be -1.00. Intermediate degrees in the relationship of these tendencies would then be represented by decimals which would be either positive or negative, depending upon whether the traits were associated together or were opposed. The coefficient which he found for the male population was +.6553, which was much higher than that for any other constitutional or environmental factor which he measured. In calculating this correlation Goring regarded 10% of the criminal male population as defective. He found that this was in agreement with the common tendency in English convict prisons to class officially about this portion of the criminals as defectives and needing care. He also assumed that 0.46% of the non-criminal male population in England and Wales was defective, the proportion suggested by the report of the Royal Commission on Feeble-mindedness. By a careful computation he calculated that 7.2% of the males either have been or will be convicted of crime before they die. He then constructed the four-fold table on the basis of these estimates as applied to the 948 convicts whom he examined as to their mental condition. The coefficient was then calculated by Pearson's method for a four-fold table. This method assumes that the mental ability and the tendency to criminality are distributed normally in the population and that the difference in numbers between the criminal and the non-criminal, deficient and non-deficient are not too great. In case the percentage of defectives among the criminals were taken as 20% instead of 10% the correlation would be increased to .79.