(1) Conditions connected with the family life have a great influence on crime; indeed, inasmuch as the family is the chief agency in society for socializing the young, perhaps domestic conditions are more important in the production of crime than any other set of causes. We cannot enter into the discussion of the matter fully, but we have already seen in former chapters that demoralized homes contribute an undue proportion of criminals. It is estimated by those in charge of reform schools for delinquent children that from 85 to 90 per cent of the children in those institutions come from more or less demoralized or disrupted families. Illegitimate children notoriously drift into the criminal classes, while dependent children who grow up in charitable institutions are prone also to take the same course. Domestic conditions have of course an influence on the criminality or non-criminality of adults. This is best shown perhaps by the fact that the great proportion of criminals in our prisons are unmarried persons. Thus the United States prison census of 1904 showed that 64 per cent of all prisoners were single persons. Statistics from other countries are practically the same. This means that, on the one hand, the family life is a preventive of crime, and on the other that the socially abnormal classes who drift into crime are not apt to marry. (2) Industrial conditions also have a profound influence upon criminal statistics. Economic crises, hard times, strikes, lockouts, are all productive of crime. Quetelet, the Belgian statistician, thought that the general rule could be laid down that, as the price of food increases, crimes against property increase, while crimes against persons decrease. At any rate, increase in the cost of the necessities of life is very apt to increase crimes of certain sorts. The various industrial classes show a different ratio of criminality. In general among industrial classes the least crime is committed by the agricultural classes, while the most crime is committed by the unemployed or those with no occupation. The census of 1904 showed that 50 per cent of all prisoners that year were non-agricultural laborers or servants. (3) The demographical conditions, conditions concerning the distribution and density of the population, have an influence upon crime. In general there is more crime in the cities than in the country districts. The statistics of all civilized countries seem to show about twice as great a percentage of crime in their large cities as in the rural districts.