THE CENSUS OF 1900 35 



The census act prescribes what inquiries shall be undertaken and, 

 in large part, what questions sliall be asked. These questions are 

 asked of every individual, of every owner of a farm, and of every 

 manufacturer in the United States, all of whom are required to answer 

 under penalty of law, and are liable to prosecution if false answers 

 are given. For this purpose a small army of investigators is essential, 

 numbering in the aggregate fifty thousand people. The country is 

 divided into three hundred districts, each of which is put under the 

 control of a supervisor, and for each subdivision an enumerator is 

 a})pointed, who is expected to make a return for from 2.000 to 4,000 

 of population. The statistics of manufactures are severally collected 

 by special agents. The enumerators are all required to complete their 

 work in thirt}^ daj's from June 1, 1900, while more time is given to 

 collectors of statistics of manufactures. All these facts are reported, 

 on schedules, which constitute what may be called the raw material 

 with which the Census Office has to deal. 



Second, the Census Office itself may be regarded as a great manu- 

 facturing establishment in which this raw material is collected into 

 printed books. Referring only to the population, it may be said that 

 this conversion involves four distinct processes. In the first of these 

 the facts recorded on the schedules are transferred to cards, one card 

 for ever^nndividual enumerated, in which holes are punched accord- 

 ing to various possible answers to questions contained in the schedule- 

 There are on each card two hundred and forty distinct positions which 

 any particular hole may occupy. The position of the hole shows its 

 significance. The second process is that in which these cards are 

 counted by electricity. The electrical counting-machine used in the 

 last census is the invention of Herman Hollerith. It is so contrived 

 that needles passing through the })unched holes on each card form 

 electrical connections which operate clock-faced dials, showing num- 

 l)ers corresponding to each individual fact or com})inati()n of facts. 

 The third process consists in entering the number on result slips and 

 combining them in tabular form as copy for the i)rinter. The final 

 process is the setting-up of the type and the preparation of the 

 stereotyped plates for the press. 



All this is very simple in theory and in ])ractice, ])ut it involves 

 an enormous amount of work. The work done in tlic last census 

 was equivalent to between 6,000 and 7, <)()() years for one nian. The 

 weight of the cards used was 200 tons, and of the schedules returned 

 by tlie enumerators 150 tons. There is not a day during the continua- 



