892 THE AMERICAN JOURNAL OF SOCIOLOGY
those in the southern states, we have, per one million, 913 prisoners in 1870, 962 in 1880, and 1072 in 1890. The statement of the populations of penal institutions on a single day is sufficiently near to the average populations at the period in question. But it is clear that the number of prisoners at any time depends not only upon the number that has been sent to prison, but also on the length of the sentences imposed ; for if in a given prison 100 convicts are received annually, with a uniform sentence of two years each, there will be in two years after opening the prison and in subsequent years 200 individuals in prison ; while, if in another prison 100 are received annually, "with a uniform sentence of three years, there will be at the end of the third year and subsequently 300 individuals in that prison. Therefore the reported population does not represent the crimes of a year. In 1890 the average sentence of those sentenced for definite terms was 3.88 years; in 1880 it was 4.14 years. As the census gives the sentences in detail a calculation may be made showing that whereas the prisoners reported in 1890 as committed for terms of one year and over were 45,115, tlie actual number committed was probably 15,295 ; and whereas 18,538 were reported as com- mitted for less than one year, the actual number so committed was probably 181,134. The former represent the serious offenders. The number of long-term prisoners, according to the census reports, increased 48.84 per cent, from 1880 to 1890, while, according to our calculation, the increase was 64.20 per cent. In 1890 there were rel- atively more short sentences given. If in one section long sentences prevail, prisoners will be relatively more numerous than in another where short sentences prevail. The census reports 6405 female prisoners, or 7.78 per cent, of the total number ; but since the average sentence for males was 4.07 years, and for females 1.59 years, the estima- ted percentage of female commitments would be 16.07. This accords with the expe- rience of foreign countries. By elements of population, according to the census of 1890, 50-4 1 P er cent, of the prison population was native white, 19.53 P er cent, foreign- born white, 28.78 per cent, negroes, 1.28 per cent, other elements. The estimate of commitments would be 49.96 per cent, native white, 29.42 per cent, foreign-born white, 19.12 per cent, negroes, 1.50 per cent, other elements. If ratios to the population are calculated, they should be with the adult population, not with the general population. The census gives an altogether incorrect idea of the relative frequency of different classes of crime. Homicide, which brings a long, usually life sentence, is represented by a large number of persons whose crimes were committed oftentimes many years before the date of the census. As a picture of crime at a given time the census report paints the situation too darkly. Crimes against the person and against property are not in any given year nearly so numerous as the census would lead us to infer, while crimes against society, receiving sentences averaging one-fifth as long as the former, are much more numerous than the census indicates. R. P. P'ALKNER, Annals of the Ameri- can Academy of Political and Social Science, January 1897. F.
An Unemployed Census. The German government has taken advantage of the opportunity afforded by its " Industrial Census " of June and its " Population Census" of December 1895,10 make a census of the unemployed. The results of this investigation are to be found in the Vierteljahrshefte zur Statistik des deutschen Reichs, 1896. The term "unemployed" was intended to include those dependent upon wages who were out of work for other cause than sickness. The number of such work-people found engaged in agriculture, industry, commerce, transport and domestic and other service was 15,497,632. Of these on June 14, 1895, 118,605, or 77 per cent., were not in work because of sickness, while 174,073, or 1.12 per cent., were not in work for causes other than sickness, that is were " unemployed. " This shows that in June the unemployed were 77 per cent, of the wage-earners, while in December it was 3.5 per cent. An examination of the problem in different industries shows that the most of this increase came from the building trades and so was clearly due to seasonal unemployment. The report says that if the statistics given are in error, they are in being an overstatement of the number of the unemployed. In the Labour Gazette, March 1897. M.
