of which parental example would not play an important part, such as arson, damage, or sexual offences, the parental correlation is found to be greater than in stealing or burglary, when the influence of parental example would be likely to have most effect. The result arrived at is that the criminal diathesis, revealed by the tendency to be convicted and imprisoned for crime, is inherited at much the same rate as are other physical and mental qualities and pathological conditions in man, and that the influence of parental contagion is, on the whole, inconsiderable, relatively to the influence of inheritance, and of mental defectiveness, which are by far the most significant factors discovered in the etiology of crime. Other environmental factors which are commonly alleged as the 'causes' of crime, _e.g._, illiteracy, alcoholism, poverty, etc., are examined statistically, so far as the data at the disposal of the author furnish ground for valid scientific conclusion. These alleged causes are, in reality, nothing more than the co-existence of associated phenomena, and until such association is analysed by statistical methods, causation, in the strict scientific sense, cannot be demonstrated. Thus, to take a general instance: poverty and illiteracy are often described as the 'causes' of crime, but as more than a third of the population of Great Britain belongs to the class of general labourers, who are presumably both poor and illiterate, such a statement can mean no more than that there is a more frequent association of criminal acts with persons living on a low rather than on a high economic scale. The exact numerical measure of the association can only be obtained by elaborate statistical comparison, the data for which are not in existence.