alleged causes of crime, such as alcoholism and epilepsy, are not more than accidental associations with crime, themselves depending upon the high degree of relationship which is admitted to exist between defective intelligence and crime. So far then, the conclusion is that English criminals are selected by a physical condition and by a mental constitution which are independent of each other: that the one significant physical association with criminality is a generally defective physique, and that the one vital, mental constitutional factor in the etiology of crime is defective intelligence. The question of the respective influence of heredity and environment is next considered by Dr Goring. The family histories of 1,500 convicts are examined, and two important relations are demonstrated (1) that the percentage of criminal offspring increases progressively according to whether neither parents, the mother only, the father only, or both parents are criminal: (2) that the percentage of criminal offspring becomes steadily greater as the age of the children increases from 14 to 23. With regard to age, the interesting fact results that the mean age of criminal enlistment is 22, with a deviation of nine years; and 14 to 32 may be regarded as the age when the chance of inherited criminal disposition is most likely to reveal itself--the modal age at first conviction is about 19. It appears also that the probabilities of conviction are greatly increased when a brother has been convicted, and the greatest intensity of the fraternal, as well as of the paternal association, occurs in families tainted by the crimes of stealing and burglary, _i.e._, the taint of habitual and professional criminality. But though the tendency for crime to recur in families already criminally tainted is an indisputable statistical fact, it is not in itself a fact of heredity. It may be due to contagion within the corrupted home into which a criminal is born. The solution of the question as to which of the two influences, heredity or contagion, is predominant, cannot be determined by observation alone--there are numberless instances pointing one way or the other--it can only be determined by a statistical examination of family statistics, where the possible influence of each factor has been eliminated. The high degree of association between criminality in husband and wife would, at first sight, seem to furnish proof of the influence of contagion, it being a relation where heredity can be eliminated, but when it can be shown that every other married female criminal is the wife of a criminal husband, and that four out of every five alcoholic wives have alcoholic husbands, the theory of contagion gives way to a theory of 'associative or selective' mating among criminals, due to the universal tendency prevailing in every department of life, of like to mate with like. So again, if we eliminate contagion, _i.e._, if we examine crimes in the perpetration