crimes against property relatively to the number of crimes against the person increases considerably as we advance towards the north." Another eminent student of French criminal statistics, M. Tarde, comes to very much the same conclusions as Quetelet; he admits that a high temperature does exercise an indirect influence on the criminal passions. But the most exhaustive investigations in this problem have been undertaken in Italy, by Signor Enrico Ferri. After a thorough examination of French judicial statistics for a series of years, Ferri arrives at the conclusion that a maximum of crimes against the person is reached in the hot months, while, on the other hand, crimes against property come to a climax in the winter.^[15]

In testing these opinions respecting the influence of climate upon crime, we are obliged, to some extent, to have recourse to international statistics. But these statistics, as has already been pointed out, owing to the diversity of customs, laws, criminal procedure, and so on, do not easily admit of comparison. So much is this the case that we shall not make the attempt as far as these statistics have reference to crimes against property. In this field no satisfactory result can, at present be obtained. The same remark holds good in relation to all offences against the person, with the exception of homicide. This, undoubtedly, in an important exception; and it arises from the fact that there is a greater consensus of opinion among civilised communities respecting the gravity of homicide than exists with regard to any other form of crime. Murder in all its degrees is a crime which immediately causes a profound commotion; it is easy to recognise; it is more likely than any other offence to come to the ears of the authorities. For these reasons this crime lends itself most readily to international comparison; nevertheless, differences of judicial procedure, legal nomenclature, and different methods of classification stand in the way of making the comparison absolutely accurate. These differences, however, are not so great as to render comparison impossible or worthless; on the contrary, the results of such a comparison are of exceptional value, and go a long way to determine the question of the effect of climate upon crimes of blood.

Assuming, then, with these reservations, that such a comparison can be instituted, let us see to what extent murder, in the widest sense of the word, including wilful murder, manslaughter, and infanticide, prevails in the various countries of Europe. In ordinary circumstances this task would be a laborious one, entailing a minute and careful examination of the criminal statistics and procedure of many nations. Fortunately, it has recently been accomplished by Dr. Bosco in an admirable monograph communicated in the first instance to the Journal of the International Statistical Institute, but now published in a separate form. Bosco's figures have all been taken from official sources, and may, therefore, be accepted as accurate; but, before tabulating-them, it may be useful to make an extract from the explanatory note by which they are accompanied. "As the composition of the population, with respect to age, varies in different countries, and as it has to be remembered that all the population under ten years of age has no share, at least under normal conditions, in the crime of murder, it has seemed to me a more exact method to calculate the proportion of murders to the inhabitants who are over ten years of age, than to include the total population. For those States where a census has been recently taken, such, for instance, as France and Germany, the results of that census have been used; that is to say, the French census of May, 1886, and the German census of December, 1885. For the other States the population has been calculated (adding the excess of births over deaths to the results of the last census) to the end of the intermediate year for each period of years to which the information relates; that is to say, to the end of 1883 for Belgium, and to the end of 1884 for Austria, Hungary, Spain, England, Scotland and Ireland. As the information respecting Italy refers to 1887 only, the population has been estimated up to the end of that year. The division of the population according to age (above and below ten) has been obtained by means of proportional calculations based on the results of the census for each State. In the case of France and Germany, however, it has been taken directly from the census returns."^[16]

Homicides of all kinds in the following European States:—

What is the import of these statistics? We perceive at once that Italy, Spain and Hungary head the list in the proportion of murders to the population. In Italy, out of every 100,000 persons over ten years of age, eleven in round numbers are annually convicted of murder in one or other of its forms; in Spain eight are convicted of the same offence, and in Hungary five are convicted. These three countries are conspicuously ahead of all the others to which our