566 THE AMERICAN JOURNAL OF- SOCIOLOGY
service." This occupation the author of this work regards as one of those "connected with woman's traditional sphere as housewife," although it is evident that almost the entire number were employed in what are practically factories, as servants who do the laundry work in the ordinary household would be reported as domestic servants and not as laundresses. The increase in the number reported as laundresses was from 108,198 in 1880 to 216,631 in 1890. If, while excluding milliners, dressmakers, and seamstresses, we include this class as engaged in manufacturing and mechanical industry, we have an increase of 64 per cent. An unbiased writer, so well informed regarding the census as the author of this work, would hardly intimate that any increase in female employment indicated by a comparison of occupation statistics of the tenth and eleventh censuses is probably due to a more complete enumeration, but would instead state that the figures doubtless come far short of showing the actual increase, because the returns of enumer- ators of population from which the tables of occupation are compiled are admittedly less complete and reliable at the eleventh than at the preceding census.
Further illustration of this author's bias in the use of statistics we find in this connection on p. 78, where he says :
Figures for Massachusetts comparing the proportionate number of females to males in manufacturing industries in 1885 and 1895 show that at the former period the number of females was 33 per cent, and at the latter period 34.6 per cent, of the number of employe's. The change, therefore, has been insignificant.
In a footnote it is said : "For details in regard to the different industries see Massachusetts Statistics of Manufactures, 1895, p. 203." Referring to the volume and page designated, we find a table showing percentages for 1885, 1894, and 1895, in which the percentage for 1885 is as stated, but that for 1895 is given as 35.09 instead of 34.6, as this author quotes it. We also find it explained, in the accompanying text, that the figures for 1885 have no direct bearing on those for 1894 and 1895; those for the earlier period being from the census and cover- ing all establishments, while those for 1894 and 1895 are from the annual reports, in which only the principal establishments are included.
One is led to wonder what the author would consider a significant increase in a state whose industries, for many years, have largely been those employing nearly as many females as males, if that indicated by his erroneous comparison is in his opinion insignificant.
The statistics of Massachusetts, as presented in the census of 1895,
