The statistical method is one of the most effective for discovering the etiology of crime, and my readers will see that I use it much of the time. Nevertheless, it must not be forgotten that this method, however important, is only one among several. It is chiefly valuable in finding the direct causes of crime; as to indirect causes it gives us much less information. I would call attention to the following example: several authors have proved that there is an inverse connection between economic conditions and crimes against persons, i.e. that these crimes increase in times of prosperity (this does not apply to recent times, however). The conclusion has often been drawn from these facts that an improvement in the lot of the working-class would lead, as a law of nature, to an increase in crimes of violence. This is a typical example of the insufficiency of the statistical method alone. When we seek for the cause of this phenomenon in another way, in the structure of society, we discover that it is to be found in the low moral and intellectual condition of the working-classes. There can be no law of nature at the root of the matter, else the well-to-do classes would be the violent criminals “par excellence.” It goes without saying that just the contrary is the case. To sum up, I conclude that statistics furnish a powerful means of discovering the causes of crime, provided they are used critically and carefully. The statistical method is not the only one; to be a good criminologist, it is necessary to be a statistician, but it is necessary to be a sociologist also. [Note to the American Edition: Upon the relation between sociology and statistics see also Žižek, “Soziologie und Statistik.” In recent years there has been a violent controversy over the statistical method. It was opened by Hoegel in his “Die Grenzen der Kriminalstatistik” and “Kriminalstatistik und Kriminalaetiologie”; then came Wassermann with his “Begriff und Grenzen der Kriminalstatistik”, and “Georg v. Mayr als Kriminalstatistiker und Kriminalsoziologe und die moderne Methodenlehre”, which goes farther and denies almost all value to the statistical method. On the other side, we find v. Mayr in his “Statistik und Gesellschaftslehre”, II, p. 441, “Forschungsgebiet und Forschungsziel der Kriminalstatistik”, “Kriminalstatistik und Kriminalaetiologie”, and Wadler in his “Erkenntnistheorie und Kriminalstatistik.”] CHAPTER III. THE ITALIAN SCHOOL. [88] I. C. LOMBROSO. In “Crime, Its Causes and Remedies”, [89] one of his last works, Professor Lombroso treats, among other things, of the influence of economic conditions upon criminality.