Nevertheless it is a fact that the relation of cause and effect between drink and crime has recently been denied, with the aid of arguments based upon statistics. M. Tammeo opened the discussion by observing that the countries of Europe and the provinces of Italy distinguished by the largest consumption of alcohol, show lower ratios under the worst crimes of violence. He gave to his remark a relative and limited value, for he only denied that the abuse of liquor was the most active cause of crime. After him M. Fournier de Flaix, maintaining the same proposition with the same statistical arguments, and admitting that ``alcohol is a special scourge for the individual who indulges in it,'' yet concluded that ``alcoholism is not a scourge which menaces the European race.'' And he repeated that the nations which consumed the greatest quantity of alcohol show a slighter frequency of crime, especially against the person. Lastly M. Colajanni enlarged upon the same proposition, using the statistical data so fully set out by M. Kummer, and drew a still more positive conclusion, that ``there is a lack of constancy, regularity, and universality in the relations, coincidence, and sequence, as between alcoholism and crime and suicide; so that it is impossible to establish any statistical relation of cause and effect between these phenomena.'' Passing over the grave errors of fact in M. Colajanni's brochure, I will only observe that this pro<p 119>position is a pure misapprehension of statistical logic. If we once admit (and unfortunately it cannot be denied) the bad influence of alcohol on bodily and mental health, in the form of spirits as well as of wine--as to which it is not correct to say that the southern departments are not consumers of alcohol--it cannot be maintained that alcohol, which is physically and morally injurious to individuals, is not hurtful to nations, which are but aggregates of individuals.