A census of irrigation was taken in 1890 and decennially thereafter, also a special census in 1902. A census of drainage was made in 1920. While these are not strictly production censuses, they present data for crops grown in irrigated or drained areas. The irrigation census is confined to 19 States in the western part of the country, where irrigation is a recognized feature of agriculture practice. It covers the area of land under irrigation, capital invested in irrigation enterprises, character of irrigation, crops grown, prices of land and water rights, water used per acre and physical condition of irrigation works. The drainage census presents information concerning the amount of land reclaimed by drainage, crops produced thereon, land needing artificial drainage, character of drainage, capital invested in drainage enterprises, and the amount necessary to complete the work.
Information regarding fisheries was collected at the decennial censuses from 184o to 1890 and at a special census in 1908. Since that date, the bureau of fisheries of the department of commerce has made statistical studies of certain districts from year to year to ascertain the fish catch. Thus figures for the entire United States can only be made up from data covering a different year for each district. The bureau of fisheries also collects annual data on the shad fisheries of the Hudson river and the Potomac river and monthly data on the fish catch at Seattle, Wash., and at certain New England ports.
Statistical data pertaining to the production of lumber, lath and shingles are collected and published annually, as are also data for pulpwood consumption and wood-pulp pro duction. The quantities of forest products used in various other manufacturing industries—such as wood distillation and the manufacture of dyestuffs and extracts, excelsior and veneers— and the amounts of cooperage stock and turpentine and rosin pro duced are shown in the reports of the biennial censuses of manu factures. The quantities and values of bark, wood, leaves and extracts used by tanners and the number of cross-ties and poles purchased are reported biennially, and the consumption of mine timbers is reported decennially. Certain data pertaining to forest products of farms and the expenditures for forest products on farms have been obtained in connection with the censuses of agriculture.
The census figures described above are, with the exception of the data on fisheries, collected at the intervals named by the Bureau of Census of the Department of Commerce, in accordance with the census law, which specifies the years for which the various censuses are to be taken and their general nature and makes answers compulsory and confidential.
The Bureau of Census was established on a permanent basis in 1902, previous censuses having been taken by a temporary force recruited for the decennial censuses. Besides the production censuses mentioned here, the bureau also collects data on popula tion, hospitals and benevolent and penal institutions, births and deaths; wealth, public debt and taxation; finances of cities and States; water transportation; distribution; and current business conditions. The census data on agriculture, irrigation and drain age are collected by means of a personal canvass by field agents of the bureau of the census, and at the decennial censuses the same agents collect agriculture figures and population figures. The information is secured directly from the operator of each farm or from an official of each irrigation or drainage enterprise, and filled in on the forms provided. For the other censuses, most of the returns are secured by mail, so far as is practicable, since these are made out chiefly by organizations. Lists of organizations are kept up-to-date by means of directories, information from trade associations and chambers of commerce, trade papers and other sources, using the previous census list as a basis. Where returns have not been obtained within a reasonable time, field agents visit the delinquent firms. At recent biennial censuses of manufactures. officials of local chambers of commerce have been appointed local agents for the canvass.
As the returns reach the bureau of the census, they are edited for omissions, errors, discrepancies, etc., and often correspondence must ensue with the individual firms to straighten out the matter. For the biennial census of manufactures alone, this correspondence amounts to over roo,000 letters. After the report has been edited, the information con tained in it is taken off on to punch cards by operators trained in this kind of work. Each punch on the card by its particular place denotes certain information on the schedule. The punched cards are next run through an automatic machine which sorts them according to the various classifications to be used, and then they are run through an electric tabulating machine which mechanically counts the number of cards, adds the data recorded in each field of the cards and automatically prints the totals. In this manner a quick and accurate count is obtained. For manu factures, the tabulation is usually made with adding machines.