reached in November-January.  (Legludic, Attentats aux Moeurs, 1896, p. 16.) In Germany, Aschaffenburg finds that sexual offences begin to increase in March and April, reach a maximum in June or July, and fall to a minimum in winter (Monatsschrift fuer Psychiatrie, 1903, Heft 2).  In Italy, Penta shows that sexual offences reach a minor climax in May (corresponding, in his experience, with the maximum for crimes generally, as well as with the maximum for conceptions), and a more marked climax in August-September (Penta, I Pervertimenti Sessuali, 1893, p. 115; id. Rivista Mensile di Psichiatria, 1899).

Corre, in his Crime en Pays Creole, presents charts of the seasonal distribution of crime in Guadeloupe, with relation to temperature, which show that while, in a mild temperature like that of France and England, crime attains its maximum in the hot season, it is not so in a more tropical climate; in July, when in Guadeloupe the heat attains its maximum degree, crime of all kinds falls suddenly to a very low minimum.  Even in the United States, where the summer heat is often excessive, it tends to produce a diminution of crime.

Dexter, in an elaborate study of the relationship of conduct to the weather, shows that in the United States assaults present the maximum of frequency in April and October, with a decrease during the summer and the winter.  “The unusual and interesting fact demonstrated here with a certainty that cannot be doubted is,” he concludes, “that the unseasonably hot days of spring and autumn are the pugnacious ones, even though the actual heat be much less than for summer.  We might infer from this that conditions of heat, up to a certain extent, are vitalizing, while, at the same time, irritating, but above that limit, heat is so devitalizing in its effects as to leave hardly energy enough to carry on a fight.” (E.G.  Dexter, Conduct and the Weather, 1899, pp. 63 et seq.)

It is not impossible that the phenomena of seasonal periodicity in crimes may possess a real significance in relation to sexual periodicity.  If, as is possible, the occurrence of spring and autumn climaxes of criminal activity is due less to any special exciting causes at these seasons than to the depressing influences of heat and cold in summer and winter, it may appear reasonable to ask whether the spring and autumn climaxes of sexual activity are not really also largely due to a like depressing influence of extreme temperatures at the other two seasons.

Not only is there periodicity in criminal conduct, but even within the normal range of good and bad conduct seasonal periodicity may still be traced.  In his Physical and Industrial Training of Criminals, H.D.  Wey gives charts of the conduct of seven prisoners during several years, as shown by the marks received.  These charts show that there is a very decided tendency to good behavior during summer and