But although no physical type peculiar to criminals can be demonstrated, certain physical differences in criminals have emerged, and it is in the examination of these differences that Dr. Goring attempts to establish a theory of criminality more simple and reasonable than that which refers them to the presence of a definite criminal type. From a comparison of the stature and weight of the general population, published in 1882 by the British Association for the Advancement of Science, he shows that, (apart from differences due to class differentiation,) in physique, as measured by stature and weight, criminals, with the exception of those convicted of fraud, are markedly differentiated from the non-criminal sections of the community. This physical inferiority, however, must not be associated with any condition of degeneracy, atavism, or other defect, mental or physical, originating spontaneously, but all the evidence points to the truth of the theory that these bodily conditions are "selective factors" determining, to some extent, conviction for crime. It may be imagined that as good physique determines occupation, so a bad physique predisposes to a criminal career. It also facilitates arrest by the Police, and apprehensions are considerably fewer than offences committed. It is, too, generally observed that persons of good physique are less irascible and prone to violence, and the case of the incendiary would show that a weakly man has recourse to a mean act from motives of revenge, not being capable of an act requiring physical force. "Fraudulents," it is true, are not selected for crime, for they resemble, in weight and stature, the law-abiding public; but they are an exceptional case, which, while destructive of a theory of degeneracy, is not necessarily inimical to the theory that physique selects crime. Dr. Goring does not deny that there is a possibility that this physical inferiority may tend to become an inbred characteristic of the criminal classes, the convicted fathers having sons who inherit their diminutive stature, and thus, in course of time, an inbred differentiation of the criminal classes might result. That this may be so is illustrated by statistics, which show that industrial and reformatory school children are consistently on the average one inch shorter in stature, and several pounds less in weight, than any other class of school-children of the same age in the United Kingdom. Nothing more than this can be conceded to the Lombrosian School. The only fact at the basis of criminal anthropology is that thieves, and burglars, and incendiaries (_i.e._, about 90 per cent. of all criminals) are markedly differentiated from the general population in stature and body-weight. There is no other scientific foundation than this for the extravagant doctrines of the "Positive" School.