from the normal is confined to some single offense. Nevertheless, single offenders constitute, as we have already seen, a very considerable proportion of our prison population. If this classification of criminals is correct, it is evident that it is very important both in studying the causes of crime and in devising practical measures for dealing with the criminal class; for the instinctive criminal, the habitual criminal, and the single offender manifestly need very different methods of treatment. One of the gravest faults of the criminal law and of penal institutions hitherto is that they have not provided for the different treatment of different classes of criminals. The Extent of Crime in the United States.--According to the United States census there were in prisons on June 30, 1904, a total of 81,772 prisoners above the age of five years serving sentences. Of this number 77,269 were males and 4503 were females; again, 55,111 were whites, and 26,661 were colored. Classified according to the prisons in which they were found, 53,292 were in state penitentiaries, 7261 were in state reformatories, 18,544 were in county jails, and 2675 were in city prisons. These were only the persons serving prison sentences. An unknown number were in county and city jails awaiting trial and serving out fines. Again, it must be remembered that this was simply the prison population on a single day, June 30, 1904. During 1904 there were, according to the census, 149,691 persons committed to prisons to serve sentences. To all of the above we must add also the 23,034 juvenile delinquents who were found, on June 30, 1904, in the juvenile reformatories of the United States. Unfortunately we have no figures from previous censuses with which we can compare the above, as the census of 1890 and previous censuses included prisoners awaiting trial. In 1890, however, there were, deducting the 15,526 awaiting trial and serving out fines, 66,803 persons above the age of five years serving sentences. These prison statistics, however, give us little idea of the actual amount of crime in the United States, because they include only the persons committed to prison to serve sentences, and do not include the vast number who escape the meshes of the law or who simply receive fines, or whose sentences are suspended. It is estimated by competent authorities, basing their estimate upon the number of known convictions of crime in certain large cities, that there are not less than 1,000,000 convictions for crime, annually, in the United States--including, of course, convictions for both felonies and misdemeanors. That this is not an excessive estimate may be indicated by the fact that in the state of New York alone in 1900, a year before the custom of suspending sentence on probation came so largely into vogue, there were nearly 100,000 commitments to prison.