other to make the entire enumeration or to complete the work which had been begun by another enumerator. In sparsely settled parts of the country it frequently occurs that no applications at all are received for the appointment of enumerators.

It may be mentioned also in this connection that in section 13 (page 15, lines 2 and 3) authority is given for the assignment of two or more enumeration districts to one enumerator, no distinction being made between urban and rural districts.

It is requested. therefore, that the proviso under consideration be amended by inserting, in line 21, page 17, after the word "shall,” the words so far as practicable," and also by inserting, in line 24, after the word "city," the words “township, or other civil division."

Since the phrase" within which his duties are to be performed," in line 22, page 17, obviously was intended to relate both to the word "district." in line 20, and to the word " subdivision." ending in line 22. the meaning will be made clearer by the insertion of a comma after subdivision." In fact, under a strict construction of the proviso as at present punctuated. the phrase "within which his duties are to be performed would relate only to "subdivision." The comma after the word "act." in line 21, page 17, was obviously inserted through error and should be deleted.

The proviso under discussion, revised in accordance with the foregoing suggestions, would read as follows:

Provided, That within the limits of continental United States each supervisor to be appointed or selected under this act shall be an actual resident of the district, and each enumerator to be appointed or selected under this act shall, so far as practicable, be an actual resident of the subdivision, within which his duties are to be performed; but an enumerator may be appointed if he be an actual resident of the city, township, or other civil division of which the subdivision in which his duties are to be performed is a part.

Senator SUTHERLAND. I would like to ask you a question or two, if you please. in relation to section 20, which provides that the enumeration of the population required by that section of this act shall be taken as of the 1st day of January.

Mr. ROGERS. Yes, sir.

Senator SUTHERLAND. What was the time of the taking of the Thirteenth Census? When was that taken?

Mr. ROGERS. April 15.

Senator SUTHERLAND. Taken as of April 15?

Mr. ROGERS. Yes, sir.

Senator SUTHERLAND. What are your reasons for changing that date, to the 1st of January?

Mr. ROGERS. I will take that up right now. In speaking of the change in the date of taking the decennial census to January 1, 1920. at yesterday's hearing before the Census Committee of the Senate. Hon. William D. Foulke, chairman of the census committee of the National Civil Service Reform League, stated that the representatives of the Department of Agriculture had testified "that January 1 would be the worst possible date for taking an agricultural census.' The office committee appointed by the Director of the Census, all of whom with the exception of one member had had experience during at least two decennial censuses, recommended the change of date from April 15, as used at the Thirteenth Census, to January 1, in order "to meet the requirements of the agricultural census, which, because

of the house-to-house canvass, must be taken in connection with the enumeration of population. The 15th of April, the date of the previous census, is not a good time of year in which to take a census of crops and live stock. At that date several months have elapsed since the crops of the preceding year were harvested, and in the interval the farm may have changed hands, so that the information is supplied by some one who has not direct personal knowledge of the facts. It is, moreover, the season of year when farm animals are born, and therefore it is a poor time at which to count live stock. Probably the best time for a census of agriculture is the fall, say October or November. But the census of population must be taken at the same time, and a serious objection to taking the decennial census in the fall is that every other census would conincide with a presidential election. Obviously, there would be difficulties in the way of conducting a successful, impartial, and scientific census in the midst of the excitement of a political campaign. So, all things considered, the committee came to the conclusion that January 1 was the best date for the census."

This change in date was unanimously recommended to the director by the leading officials of the Department of Agriculture, by college professors interested in agricultural work, by students of agricultural subjects, and by editors of agricultural newspapers. The Secretary of Agriculture appointed a committee representing the principal bureaus of his department to cooperate with the committee of the Bureau of the Census to aid in determining a date for the Fourteenth Decennial Census and in the preparation of a tentative schedule of agriculture for the enumeration.

The committee recommending this date was composed of D1. Le Grand Powers, chief statistician of the Division of Agriculture of the Bureau of the Census at the censuses of 1900 and 1910; Mr. William A. Hathaway, chief of division of the Division of Agriculture of the Bureau of the Census during the census of 1910, and now acting statistician for the American Telephone & Telegraph Co., New York City; Dr. George F. Warren, of Cornell University; Dr. Henry C. Taylor, of the University of Wisconsin; Dr. John Lee Coulter, of the University of the West Virginia; Dr. Clarence Poe, editor of the Progressive Farmer; Mr. Clarence B. Hurrey, United States Chamber of Commerce; Mr. W. J. Barrows, formerly chief of division of the Division of Agriculture of the Bureau of the Census, and now chief clerk in the office of the Third Assistant Postmaster General.

The committee of the Department of Agriculture was composed of Mr. L. M. Estabrook, Chief of the Bureau of Crop Estimates; Mr. Charles J. Brand, Chief of the Office of Markets and Rural Organization; Prof. W. J. Spillman, Chief of the Office of Farm Management; Mr. George M. Rommel, Chief of the Division of Animal Husbandry; and Dr. L. C. Corbett, Chief of the Division of Horticultural and Pomological Investigations of the Bureau of Plant Industry.

This change of date was also. indorsed by representatives of the Office of Indian Affairs, the Reclamation Service, Bureau of Plant Industry, Office of Public Roads and Rural Engineering, and by Dr. F. H. Newell, of the University of Illinois, attending the conference on irrigation, held December 19-22, 1917.

Later the change of date was indorsed also by the delegates attending the drainage conference, held February 18-20, 1918. This conference was composed of representatives of the Office of Public Roads and Rural Engineering of the Department of Agriculture, the Engineer Corps of the United States Army, and a number of leading engineers and college professors of the country.

Under date of April 22, 1918, former Director E. Dana Durand stated in a letter to the present Director of the Census "that, if adopted, this legislation would be a decided improvement, especially with respect to the contents of the schedules and the date of enumeration."

Senator SUTHERLAND. That has reference to the agricultural schedules rather than to population.

Mr. ROGERS. Well, that was submitted, and they, of course, discovered the dates of the census and the contents of the inquiry.

The statement made at the hearings yesterday by Mr. Foulke, of the National Civil Service Reform League, that an official of the Department of Agriculture had made a statement protesting the change of date of the census to January 1 was in error. The statement to which he referred was made by Dr. Pearl, a representative of the Food Administration, who stated that his opinion concerning the change of date to January 1 was based upon a talk with some of the men of the Bureau of Animal Industry of the Department of Agriculture. When Dr. Pearl's statement was called to the attention of Mr. George M. Rommel, Chief of the Division of Animal Husbandry of the Department of Agriculture, he made the following response, under date of May 2, 1918:

I can not quite agree with Dr. Pearl that January 1 is the worst possible date that could be taken. I think that the date of April 15, from the live-stock standpoint, is absolutely the worst because that is in the midst of the season when the young animals are being born. However, I do not believe it would be possible to get an ideal date. If you will recall, one reason why the representatives of this division agreed to January 1 was that the number of young animals was at a minimum and that this date coincided with the date used by the department.

I feel that the great necessity in the live-stock census at this time is to determine what is the actual breeding power of our live-stock industry. That we do not know at present. We simply have estimates which are confessedly inaccurate. The fact that January 1 is an objectionable date for the reason suggested by Dr. Pearl and Dr. Wright makes it all the more necessary to have an accurate count of the breeding animals on hand on that date so that we can distinguish between those animals which will enter more or less speedily into the food supply and those whose effect on the food supply is likely to be projected over one or two years. Of course it is extremely important for us to know what the meat supply is likely to be for the next six months, but the supply which is available to-day is practically gone six months from now. It is much more important for us to know whether our stocks of animals on hand at this time are sufficient to give us a necessary meat supply a year or two years from now.

That is his letter.

Senator SUTHERLAND. Well, the determination of the committee to take the census as of January 1, is based entirely upon the desirability of doing so from the standpoint of the agricultural schedule.

Mr. ROGERS. No, sir; but that is the best date that a decennial census can be taken, considering all the inquiries to be made. To go back to the question that I have just finished, if you care to hear what Secretary Houston said in regard to that, I will read that.

The CHAIRMAN. We will be glad to hear it.

Mr. ROGERS. I will read what Secretary Houston says:

I should think the time would depend upon the status of the crop. If the census is taken in January, 1920, the acreage harvested of each crop grown in 1919 can be readily ascertained by the census enumerators, because the questions will be asked after practically all crops for the year are harvested and before crops for the new year are planted, and while the facts relating to the previous crops are still fresh in farmers' minds. Also in most States tenants who grow crops in 1919 are still on the same farms in January, whereas if the census were postponed to April or a later month, many tenants would have moved. January is considered the best month for taking a census of live stock because the number is more stable than later in the year, and also because the number reported by the census in that month will be comparable with the annual live-stock estimates of the Bureau of Crop Estimates in January. Another advantage of taking the farm census in January is that many farmers will be available for service as enumerators and farmers can be interviewed at a time when they are not actively engaged in farm operations. January is also the month of inventories and an appropriate time for a census enumeration. The CHAIRMAN. Well, Mr. Rogers, have you anything else that you desire to put in the record?

Senator SUTHERLAND. I want to ask a few questions right along this line, if you please, Senator.

The CHAIRMAN. Pardon me, Senator; proceed.

Senator SUTHERLAND. How about the taking of the census at that time of the year, that is, up in the northern States? ·

Mr. ROGERS. Well, there is a discretion provided for in the bill, which I think would attend to that.

Senator SUTHERLAND. Do you not think that that would have to be exercised so generally as to in effect vitiate your census?

Mr. ROGERS. No; taking it as a whole, there is a very small section of this country which is extremely cold and inaccessible in the early part of January, and if a cold spell comes on, it can be taken under another date. Then, there are other features in regard to this date which should be given consideration. This date is right after the holidays, when the school-teachers and the boys are not busy, and the business men are not taking their vacations, and it is a time when the people are at home. They can be enumerated more easily because they are at home, and that would be the best time to secure the manpower to assist in the enumeration, and after the time the crops are just harvested, and the inventories of the business men and the farmers for the year made out, and every man knows exactly what wheat he has, what he has in corn and live stock and everything else. Then, too, the manufacturers, in the majority of cases, have finished the year's work and they have taken an inventory of the stock, and they know at that time of their accomplishments for the year. On the whole the reasons are so many that I would encumber the record if I were to try to detail them.

Senator SUTHERLAND. Well, I would say that it would seem to me that in States like West Virginia and the other northern States and the mountain States, that it would be a very cold time of year, and it would be very difficult to get the people to go out and take the census along the 1st of January, if there should be such cold weather as we have had

Mr. ROGERS. Well, there are some objections to any period of the year.

Senator SUTHERLAND. It seems to me that that same condition is likely to prevail all over the northern part of the United States

Mr. ROGERS. Well, climatic conditions might make January 1, as compared with April 15, not so preferable, but there are other conditions which must be considered, such as the men being busy on the farms and being hard to find, the school children being away at school, the business man hunting for orders and having no recollection of his business or any inventory of what he did during the last year. You will find in the cities and in the towns that the business men make the calendar year the record of their business.

Senator SUTHERLAND. It would seem to me that in that respect— that is, with reference to the climatic conditions which might prevail and continue throughout the entire month, especially in the northern States, where the cold weather starts in about Christmas and extends through January and February-that you might find it very difficult to get people to go out in weather of that kind when it is 10 degrees below zero or thereabouts

Senator NEW. Take last January, for instance.

Mr. ROGERS. Yes; last January was an exception, surely.

Senator SUTHERLAND. And take the snow and the cold weather; I think that it would be humanly impractical. Of course the eminent men that you have gathered about you in the discussion of this matter, I feel that their opinions are very valuable-no doubt they are, but at the same time we all know what weather conditions may prevail and frequently have prevailed during the entire month of January and continued along in February and a good part of March, so that we can almost count it as an exception when we do not have very severe climatic conditions through January and February, and to get people to go out in my section of the country and take the census in January, I think you would find it very difficult, if the weather there is like it usually is. More than that, they frequently have snow, and there may be 6 feet of snow

Senator NEW. There is one thing certain, and that is if you had a January like the January we had last year, we would have no difficulty in finding the people at home.

Senator SUTHERLAND. No: but the difficulty would be for them to get away from home, and you would find the people at home.

Mr. ROGERS. Well, when the enumeration is done by a representative who is stationed in that locality, he does not go very far.

Senator SUTHERLAND. Then, too, I should think there would be some danger of the enumerators guessing at it, instead of making the trips, and they would use the country telephones, and I think that they would not get the information accurately.

Mr. ROGERS. The enumerator would not have a large territory, and he would have the means of getting the information.

Senator SUTHERLAND. Yes

Mr. ROGERS. And he has a month in which to do it.

Senator SUTHERLAND. Well, it takes a good part of a month to cover the enumeration if it is scattered, even under normal conditions?

Mr. ROGERS. You are speaking of the country districts, I suppose! Senator SUTHERLAND. Yes, sir; I am speaking of the rural sections. Then you always allow a month