From the foregoing data we have drawn the following conclusions, realizing that the number of cases dealt with is too small, and the data in many of the cases too meagre, to more than suggest them as tentative or probable.1. Normal traits predominate in manic-depressive psychoses (164 to 120).2. Abnormal traits are found in about an equal proportion in dementia præcox and manic-depressive psychosis (98 to 104).3. Certain abnormal traits—reticence, peculiarity, precosity—are found in much larger proportions in dementia præcox than in manic-depressive psychosis (27 to 4).4. Normal personalities are found fairly frequently in both diseases, but more frequently in manic-depressive psychosis (33 to 26).5. Abnormal personalities are much more frequent in dementia præcox than in manic-depressive psychosis (16 to 7).6. The "shut-in personality" or the tendency to it is found almost exclusively in dementia præcox (10 to 2).7. If all doubtful and abnormal personalities (i. e., the normal—mixed + all abnormal as in Table II) are conceded to predispose to dementia præcox, on the one hand not more than half (24 cases, 48 per cent) of the dementia præcox cases are so predisposed; and on the other, a third (17 cases, 34 per cent) of the manic-depressive cases have the same predisposition.8. The special "shut-in personality" is found in too small proportion (20 per cent) to substaniate fully the claims made by some writers for its prevalence as an etiological factor in dementia præcox.The study suggests three main lines of future work:1. The gathering of fuller data as to make-up and characteristics, from several sources in each case whenever possible.2. The making of similar comparisons of make-up in all the psychoses, but especially in the high-grade defectives and epileptics.3. The study of characteristics and personalities as possible determinants of reaction-types or of symptoms, or as modifying the clinical picture, rather than as causes of the onset of the psychosis.