624 THE AMERICAN JOURNAL OF SOCIOLOGY
That there were establishments omitted from the canvass at each census cannot be denied. The omissions in 1 890 are referred to in the reports. The extent of these omissions at prior censuses no one can determine. I, however, have an idea of the extent of the omissions in 1890, and recall the fact that some of the largest ship-building establishments of the country, some of the largest sugar refineries, paper mills, cigar factories, and establishments engaged in other industries, absolutely refused, or willfully neg- lected, to furnish the information required for the census. A number of cases were referred to the Attorney General to institute legal proceedings in order to secure the data, but the suits were abandoned, for various reasons, among others because the infor- mation could not be secured in time to be included in the reports. While the omission of these establishments may not materially affect the totals, they undoubtedly offset the inclusion of the minor industries referred to, and tend to counterbalance the results of the more thorough canvass.
The establishments of productive industry being largely con- centrated in cities, a special effort was made at both the Tenth and Eleventh Censuses to make a complete canvass of the cities. It is in the cities, if anywhere, that the canvass of 1890 is more complete than that of 1880, for I am confident that in the rural districts, as a whole, the canvass of 1880 resulted in securing a larger number of reports than did that of 1890. A greater num- ber of cities were especially canvassed in 1890, and this with their possibly better canvass accounts for any increase that may be due to the character of the enumeration.
The extent that the possibly better canvass of the cities has affected the grand totals cannot be determined. The only industries that could be perceptibly affected are those generally conducted in small establishments, such as carpentering, black- smithing, tailoring, etc. In these industries a very large number of establishments control but a very small percentage of the total product for all industries. The canvass for the large establish- ments was in all probability as thorough throughout the entire country in 1880 as in 1890. In view of these facts it is absurd to
