]l4 the bureau of science 



of the variety of work carried on in a single day in the entire 

 Bureau with its dozen divisions and sections. The analyses or 

 examinations of samples of a given kind which are performed 

 by the laboratory cannot often be segregated because with few 

 exceptions they must be done promptly without waiting for 

 others to accumulate. 



Routine work does not prevent the Bureau from fulfilling its 

 lawful function of conducting independent scientific research 

 and working out the problems of economic importance, provided 

 it be of a legitimate character. By a small nominal charge such 

 as we make in most cases, we are able to exclude unimportant 

 samples submitted out of curiosity and only of momentary in- 

 terest ; without such a charge we would be obliged to devote our 

 entire time to routine work at the sacrifice of all our research 

 work. The schedule of prices on the average does not cover more 

 than the actual cost of routine analyses or examinations, on 

 account of the extreme variety of the work. Certain types of 

 analyses are made gratis exclusively for other branches of the 

 Government, and in these cases it has been necessary only to 

 approximate the cost for statistical purposes so the charges in 

 the schedule may be more or less than the actual cost of the 

 routine analyses or examinations. The charges for unusual and 

 miscellaneous work are too low. The large variety of work 

 makes economy of time difficult; the lack of room makes it im- 

 perative to disassemble apparatus as soon as work is completed 

 instead of keeping it set up for similar work, as the space is 

 needed for new work, and the irregularity with which requests 

 are received interferes with the intertwining of suitable problems 

 for research and investigations. The irregularity is partially 

 overcome by enlisting the routine employees as secondary workers 

 on investigations. The cost of performing routine work of a 

 given class varies from time to time, depending upon conditions. 

 In the United States a laboratory often secures a large amount 

 of one class of work and can invariably employ cheap assistance 

 and carry it on empirically. For instance, there are laboratories 

 in the States where they do nothing but analyses of boiler water. 

 In spite of their cheap labor, in almost every case their charges 

 are higher than ours. Where the volume of any given class of 

 work performed by the chemists, bacteriologists, assayers, or 

 other employees is large, we are able greatly to reduce the unit 

 cost and especially are we able to do so when we train and use 

 the services of intelligent apprentices. At the present time we 

 have an apprentice mechanical soil analyst, an apprentice milk 

 analyst, and apprentice cement testers. Each of these is able 



