How closely the degrees of immorality are associated with the degrees of deficiency remains one of the most important problems to be answered authoritatively by the correlation of these traits when properly measured. That the greater degrees of immorality and of deficiency are on the whole associated and not opposed we have good reason to believe, but there are undoubtedly examples in which the degree of immorality or delinquency is out of proportion to the degree of deficiency. The fact that certain instances are found of moral imbeciles without corresponding intellectual deficiency, which has been noted by Stern (_188_, p. 75) and by Anton (_67_), does not of course determine the direction of the tendencies. We must base our deductions as to the danger of delinquency among lower and higher grades of deficients on our knowledge of the general tendencies. Are morons, relative to their numbers, more dangerous to the community than lower grade deficients? We must not make the absurd deduction that because morons are most numerous they are most likely to be delinquent and should therefore be most carefully isolated or supervised. B. THE CORRELATION OF DEFICIENCY AND DELINQUENCY. Modern statistical methods afford the ultimate quantitative tool for determining the cause of delinquency, whether or not we also require that the data should be assembled under experimentally controlled conditions. The rapid strides which have been made in answering this fundamental question of criminology may be judged by noting the treatment of it in such a work as Goring's compared with the impressionistic literary style which has prevailed. Illustrations of particular cases, opinions subconsciously formulated by experts from wide experience in dealing with delinquents, even the votes of the majority of leaders in the field, give way before the acid test of measurement of tendencies in human traits just as poorer methods succumbed in the Middle Ages in the realm of the physical sciences. Quantitative determinations can no longer be brushed aside with a smile on the supposition that statisticians are the biggest liars. They must be answered by better data or more refined methods. The form of the discussion of social questions has changed. Correlation is a powerful new weapon for attacking these problems which promises to go far beyond the range of earlier blundering methods.