Through statistics, then, we are placed in possession of the facts relating to crime and criminals, and facts alone can give the color of reason and good sense to all measures of reform, to all projects looking to the suppression of crime and the elevation of the criminal classes. Statisticians, therefore, whatever may be their theories, whatever their pet views about crime and criminals, deserve well of the community; for without their close and painstaking work the most ingenious theorist and the best-inclined philanthropist would be utterly at sea; for as Phidias could not have chiselled his unrivalled Zeus without the marble, neither can the most zealous reformer advance a foot without clear and well-tabulated statistics. For this reason we bid especial welcome to the interesting monogram of Mr. Dugdale, which is a monument of patient and laborious exploration in a field of limited extent. It is evident that he did not set about his work in a _dilettante_ spirit, but spared no effort and avoided no inconvenience—and his inconveniences must have been many—to ascertain the utmost minutiæ bearing on his topic. He has not contented himself with adhering to the methods of inquiry usually in vogue, but has added to the law of averages, which ordinary statistics supply, individual environments and histories which may be considered causative of general results, and as such are the key to common statistics. “Statistics,” he says, “cumulate facts which have some prominent feature in common into categories that only display their static conditions or their relative proportions to other facts. Its reasoning on these is largely inferential. To be made complete it must be complemented by a parallel study of individual careers, tracing, link by link, the essential and the accidental elements of social movement which result in the sequence of social phenomena, the distribution of social growth and decay, and the tendency and direction of social differentiation. To socio-statics must be allied socio-dynamics. Among the notable objections to pure statistics in the present connection is the danger of mistaking coincidences for correlations and the grouping of causes which are not distributive.”