The net result, for our purposes, of studies so far made of mental decay is not very helpful, because (1) the original status of the subjects is never known, (2) the psychoses have been mixed in experiment, without preliminary test-knowledge of the characteristics of each, if any, and (3) the method of correlation, which has been used, is not suited to show selective enfeeblement of mental functions. Every study made has suffered from one or more of these hindrances to interpretation. The information gleaned from them is much the same as that already gleaned from studies of the undeteriorated, namely, that among people (whether sane or insane) those who hold a certain rank within a group in one function tend also to hold a similar rank within that group in other functions. The question of _selective_ enfeeblement of a function within a group of the insane remains unanswered. The investigators of the demented have, however, made a particular contribution in pointing the way to a new source of light. For the study of mental decay, when carried on by adequate methods, extremely difficult of attainment, is sure to throw light on the relationships among mental functions. From it we shall learn whether some functions remain intact, with impairment of other functions. V. IS INTELLECT INHERITED AS A UNIT? There are still other approaches to the study of the constitution of intellect. One is through the investigation of heredity. The question is whether intellect is inherited as a unit, or whether some different formula is indicated. If intellect is a unit character, subject to but one determiner in the germ-plasm, then it should act as an “all or none” capacity in its appearance among offspring of given matings. Children should be separable into distinct groups, each having a different median with respect to intellect, _i.e._ those who have intellect and those who lack it. The methods of mental measurement teach us plainly that intellect is not inherited in this way. Instead of a broken curve, indicating a division of children into those who inherit and those who fail to inherit a unit character, we obtain the curve already demonstrated, which is continuous and symmetrical. There is but one diversified group of children, with respect to intellect—not distinct groups. The inheritance of intellect does not, therefore, follow the simple formula of unit characters, as does the shape of peas, the color of rabbits’ coats, or eye-color in man. The trait we measure and name as general intelligence is a complex, resulting from the incidence of a great number of functions, acting together in a great number of ways, yet cohering in respect to amounts found in given individuals. Possibly each of the indefinitely numerous functions, which thus appear to act together as man’s intellect, may be a unit character, inherited according to Mendel’s formula. Such a possibility is at present purely speculative.