The work of Lombroso set out with two original faults: the mistake of having given undue importance, at any rate apparently, to the data of craniology and anthropometry, rather than to those of psychology; and, secondly, that of having mixed up, in the first two editions, all criminals in a single class. In later editions these defects were eliminated, Lombroso having adopted the observation which I made in the first instance, as to the various anthropological categories of criminals. This does not prevent certain critics of criminal anthropology from repeating, with a strange monotony, the venerable objections as to the ``impossibility of distinguishing a criminal from an honest man by the shape of his skull,'' or of <p 4>``measuring human responsibility in accordance with different craniological types.''[2] [2] Vol. ii. of the fourth edition of ``The Criminal'' (1889) is specially concerned with the epileptic and idiotic criminal (referred to alcoholism, hysteria, mattoidism) whether occasional or subject to violent impulse; whilst vol. i. is concerned only with congenital criminality and moral insanity. But these original faults in no way obscure the two following noteworthy facts--that within a few years after the publication of ``The Criminal'' there were published, in Italy and elsewhere, a whole library of studies in criminal anthropology, and that a new school has been established, having a distinct method and scientific developments, which are no longer to be looked for in the classical school of criminal law. I. What, then, is criminal anthropology? And of what nature are its fundamental data, which lead us up to the general conclusions of criminal sociology? If general anthropology is, according to the definition of M. de Quatrefages, the natural history of man, as zoology is the natural history of animals, criminal anthropology is but the study of a single variety of mankind. In other words, it is the natural history of the criminal man. Criminal anthropology studies the criminal man in his organic and psychical constitution, and in his life as related to his physical and social environment--just as anthropology has done for man in general, and for the various races of mankind. So that, as already said, whilst the classical observers of crime study <p 5>various offences in their abstract character, on the assumption that the criminal, apart from particular cases which are evident and appreciable, is a man of the ordinary type, under normal conditions of intelligence and feeling, the anthropological observers of crime, on the other hand, study the criminal first of all by means of direct observations, in anatomical and physiological laboratories, in prisons and madhouses, organically and physically, comparing him with the typical characteristics of the normal man, as well as with those of the mad and the degenerate.