Kehrer[289] has made a careful analysis of the facts brought out by Kraepelin's statistical diagram showing the various age groups represented by his manic-depressive cases. "From the fifteenth year of life, at which age manic and melancholic attacks are most frequent (about twenty-five per cent), the curve of the manic attacks falls steadily (with only two important rises at the thirty-fifth and the forty-fifth years) until it becomes less than five per cent at the seventieth year, while the curve of the melancholic conditions with equal constancy increases (with the exception of the fifty-fifth year only), especially between the forty-fifth and fiftieth years, from fifty-two to seventy-four per cent and finally to eighty per cent. On the other hand, the curve of the manic first attacks falls steadily from 28.5 per cent at the twentieth year to 3.5 per cent at the sixtieth, with a slight increase at fifty from 12.7 per cent to 13.4 per cent, while in the male sex the same curve shows no further increase after the thirtieth year, when it reaches its maximum (33.8 per cent) and even shows a particularly sharp fall, from 22.2 per cent to 5.9 per cent, between the fiftieth and sixtieth year.... Based on this diagram Kraepelin concluded that the depressions of the involutional period, which did not show special symptoms of some other disease entity, could not be differentiated from those of the earlier periods of life." Specht,[290] Hübner and Stransky have subscribed to these views. Stransky expressed the opinion that "there is nothing in the form of these depressions, either with or without anxiety, by which they can be distinguished from those recognized as manic-depressive insanity and that neither the course nor the age of onset offer any convincing argument for their clinical independence." Rehm, on the other hand, held that there were depressions of the involutional period of life corresponding to Kraepelin's melancholia and not belonging to manic-depressive insanity. He described these as lacking the constitutional taint and characterized by a slow onset, without previous attacks, fatigability, outspoken egocentric conduct, hypochondriacal delusions of the deteriorative type and the appearance of hallucinations. Bleuler,[291] Bumke, Seelert, Albrecht and others still hold to the integrity of involution melancholia as a distinct entity. "These forms," as Bleuler expresses it, "have as a rule a much more protracted course. They progress slowly for one or two years, continue to be mild, reaching their height in several years, and decline slowly to their final conclusion. The inhibition is obscured by great restlessness, genuine agitated forms are common, they tend to recidivism much less than the others and show also much less heredity." Albrecht, in 138 cases of functional psychoses of the involutional period, only thirty-two of which were in men, diagnosed eighty-two as genuine involution melancholia. In none of his cases did he find an