374 THE AMERICAN JOURNAL OF SOCIOLOGY
increase from that year to 1890 was 12 per cent, by simple average and 17 per cent, averaged according to importance. The increase in average annual earnings shown by the census from 1880 to 1890 is over 39 per cent.
Colonel Wright's report on the employment of women and children, having been, as we have seen, discredited by himself, seems hardly worthy of extended criticism. As far as it may be relied upon as showing anything at all, it indicates a decided increase in the employment of both women and children.
One of the eccentricities of this investigation is that it takes eighteen years of age as the line of division between adults and children, rendering its figures incomparable with those of the census reports or those of the labor bureaus and the factory inspectors of the various states.
It fails to indicate the extent of the increased female and child labor, because, as has been already shown, a statement of the increase of such labor in establishments that existed ten years ago and at the present time takes no account, of what is claimed to be the fact, that establishments employing a large proportion of women and children have increased in number more rapidly than establishments of a different character.
As a complete canvass of the whole country could not have been accomplished with the funds at Colonel Wright's disposal, it would seem that the end sought would have been more nearly attained by a complete canvas in one or more localities that might be considered fairly representative. This has not been attempted, but, instead, a canvass has been made of establish- ments variously located that are claimed to be representative.
An investigation of the character and location of these establishments seems hardly to substantiate this claim. We find Illinois represented by but eighteen establishments, having 889 employes in the former and 1290 in the latter period, while Georgia is represented by fifty establishments, with 4297 employes in the former and 9189 in the latter period. South Carolina also outranks Illinois in importance, being represented by three times as many employes in the former and five times
