the arrangement of the various occupation designations with respect to classes differed somewhat from that of 1880, principally in the grouping of fisheries and mining with agriculture and in providing a class for persons engaged in professional pursuits separate from those engaged in purely domestic and personal service, thus making five main classes instead of four, as formerly, and with a still further reduction in the number of occupation designations to 218.

The passage of the act, approved March 3, 1899, which governed the taking of the Twelfth Census, marked an important departure in census-taking in this country. By this act the work at the decennial period was limited to four subjects,population, mortality, agriculture, and manufactures; but provision was made after the completion of the decennial work for the collection of the statistics relating to many special subjects. This step became necessary because of the encyclopedic character of the work of the two preceding censuses, at which the attempt was made to carry on a host of special work substantially coincident with that required for the general enumeration of population and the collection of the statistics of industry; and, therefore, in these respects the census of 1900 was a marked improvement over its immediate predecessors. It did not appreciably reduce, however, the work attending the general enumeration of population, and the requirements with respect to the return of occupations by the census enumerators were substantially the same as in 1890. In the presentation of the statistics there were 303 designations of occupations, arranged under 140 occupation groups, and conforming, as nearly as possible, to the general classification used at the preceding census.

These are, in brief, the facts of the enumeration and presentation of occupations in the Federal census; but what of the conditions under which the work has been conducted? There need be no consideration in this regard concerning the first two groups, comprising the earlier efforts to secure occupation data, because in each case the basis differed and the returns were not inclusive of all the gainful workers. For the third group, com

prising the last four censuses, the basis for the return and presentation of occupations established by General Walker in 1870 has been, in the main, observed, although the effectiveness of many of the occupation designations, as describing the groups of workers therein included, has been very much lessened on account of the increasing degree to which labor has been subdivided. The results have been presented in varying detail as to the number of occupation designations used, but covering for each census substantially all the gainful workers of both sexes and of all ages.

There has been at each of these censuses forced conditions of work, arising out of a wholly temporary organization and with entirely too short a time to prepare for so great an undertaking. And thus it would seem to have been a well-nigh impossible task, particularly since 1880, to bring together suddenly a large clerical force; to provide for their housing in widely scattered quarters; to map out a complete plan of work to govern the operations of both the office and field force; to prepare and print the necessary schedules and instructions, amounting in 1900 to very nearly 26,000,000 copies; to provide for the appointment of several hundred supervisors, and through them for the selection and equipment of a small army of enumerators,―more than 53,000 in 1900; and in time to begin work, under the law, on June 1, barely one year after the passage of the main census act, to say nothing of supplemental legislation at a much later date by which the plan of work was considerably modified. It is evident, therefore, that under this stress of work there could be, even at the Twelfth Census, but little, if any, provision made, except through the printed instructions, for training the field force or for the close supervision of their work during the course of the enumeration; nor was there, indeed, much opportunity for the detailed examination of the completed schedules before they were required to be sent to the central office at Washington. The supervisors, it is true, were furnished with instructions concerning the manner in which the schedules in general were to be scrutinized for omissions and possible error, but, aside from establishing the completeness of the canvass in

each census district, the necessities of the work, under the time limit prescribed for its completion, demanded a degree of expedition in the return of the schedules which precluded their critical examination, even though the supervisors' offices were fully equipped for the work, which was not always the case. As a consequence, the census officials were practically helpless in the matter, and in general were compelled to accept the entries on the schedules as they came to them, and to tabulate them accordingly. As the census has been carried on to date, therefore, the return of occupations in the detail essential to their proper classification has rested largely upon the printed instructions to the census enumerators and the integrity and interest with which they have performed their duties under them.

Fortunately for the Thirteenth Census, this era of census adversity is or ought to be largely now a matter of history, and, as we approach the active period of this great work, there is much reason to hope for improvement at many points, not the least of which is with respect to the return and classification of occupations.

The first and foremost basis of this hope rests in the fact of a permanent bureau, with a strong nucleus of trained experts and experienced clerks. The former temporary census organization was fatal to concerted effort, and the preparations covering so wide a range of operations were too hurried, and therefore necessarily incomplete and insufficient in many particulars. The permanent organization should have been effected at least twenty years ago, and every census superintendent from General Walker down has given freely of his testimony as to what the outcome was likely to be in the near future unless there was a material change in the plan of work. Happily, the present Director, having a permanent bureau to begin with, can look forward with hope to the coming of the Thirteenth Census, and emphasize, as he has already done before the committees of Congress and in his annual reports, some of the agencies through which must come material improvement in the conditions surrounding the work at the period of the general enumeration, for example: (1) the change of the date of enumera

tion from June 1 to April 15, thus providing for a more nearly complete enumeration of the population-particularly that in cities-than has ever before been possible; (2) the omission of the mortality schedule and the entire withdrawal of the manufactures schedule from the general canvass, thus limiting the work of the census enumerators in cities to but one schedule,population, and in country districts to but two schedules,population and agriculture; (3) the provisions for a smaller enumeration district, for increased and graded compensation for both supervisors and enumerators, better organization for and much closer supervision of the field work, opportunity for longer preliminary training of both supervisors and their assistants, and, as a consequence, the possibility of better provision for the examination in and correction from the supervisors' offices of all the returns, including occupations; and last, but not least, (4) the amplification of the returns with respect to occupation, so as to include the class of worker (employer, employee, or working on own account) and, if possible, by whom employed, thus affording, to the extent that the inquiries are successful, the means for the better classification of the occupation data, particularly in industry. Although the census officials have not been unmindful of the desirability of securing information of this character, as is done in some of the European countries, it has not been deemed advisable heretofore to attempt it, because of the temporary character of the census organization and the "forced" conditions of the work generally, as already explained.

Thus it will be seen that it is the purpose to strengthen the work of the enumeration in every way possible, but with all our hopes we must not lose sight of the fact that there are limitations which are inherent and to an extent ineradicable. We may hope to improve materially the quality of the returns, but we shall not be able to get just the facts we are seeking. From the nature of the case it is difficult, first, to secure everywhere enumerators that in all respects fill the bill; and, second, even though they be of the right stamp, the answers to the inquiries cannot always be secured at first hand, and so the

returns will not state the facts as explicitly as is desired and as the instructions require. In theory, it is possible, under a permanent organization, to provide for the early selection, appointment, and training of the enumerators,-far in advance of the need for their services,-but, as a matter of fact, many changes are bound to occur in any list of original selections, however or whenever made up, before the census day is at hand; and experience teaches us that there are likely to be, too, many vacancies to be filled on the morning of the day when the work begins and throughout the course of the enumeration as well. The causes for changes in enumerators are many and varied,-sickness, death, new employment offered, change of mind, indifference to work, because found to be too uninviting, too exacting, or not sufficiently profitable, so that the original list of selections is subject to considerable revision at a later day, a "waiting list" becomes a necessity, and it is often necessary and desirable, when the work is well advanced towards completion, to utilize the services of an enumerator who has proven his worth in one district in carrying on or completing the work in a second district. So large a body of enumerators is, after all, as in times past and in most countries, likely to be made up of good, bad, and indifferent workers, widely scattered as they are over a large area of country, and largely dependent for their selection, early equipment, and instruction upon a body of supervisors who have had, on the whole, no previous census experience, and who, as business men of their respective localities, cannot devote much time to purely preliminary study in advance of the actual preparatory work, so that the matter of successful enumeration finds its foundation largely in the ability of the supervisor to direct the work of the enumerators, and in the use of good business sense in meeting the emergencies which will arise inevitably during its progress, and which, in all respects, cannot be anticipated through previous instruction, either written or oral.

In the preparations for the Thirteenth Census one other step is to be taken which, so far as it may help to clear up the situation with respect to occupations, is undoubtedly the most