222 THE PATH OF SCIENCE 



or could direct and, to a great extent, meets the most serious 

 objection to the operation of the project system. 



For the direction of the service and the development prob- 

 lems, which in most laboratories represent the greater portion 

 of the work, the project system would be preferable were it 

 not for the fact that it costs a great deal more. The project 

 system requires a complete accounting system, a great deal of 

 reporting involving stenographic assistance and filing, and, 

 in addition, it consumes an immense amount of time, both 

 of the scientific staff and of the management of the company, 

 spent on the careful consideration of the various projects. In 

 many large laboratories, much of the time of the senior 

 scientific staff is devoted to conferences and committee meet- 

 ings at which the problems of the laboratory are discussed in 

 detail. This is so serious that some laboratories openly state 

 that it is undesirable for the best scientific men to be group 

 leaders since they are left little time for scientific work and 

 that the scientific experts should have their work directed by 

 a group leader who is essentially a business man with scien- 

 tific training. It is very difficult to calculate accurately the 

 relative costs of the two systems, but with certain simplifying 

 assumptions, it is not impossible to make guesses. 



1. Let us assume that in both systems the scientific men 

 are paid the same amount. 



2. Let us assume that in both systems the scientific men 

 are of the same average ability. 



Then the cost by the two systems per scientific man employed 

 for the same total amount of ^vork done can be measured by 

 the total cost of the laboratories. Accordino^ to available figr- 

 ures, the cost of some laboratories run by the project system is 

 approximately $10,000 a year for each scientific man working 

 in the laboratories; in laboratories without the project sys- 

 tem, in which the work is directed only by the scientific staff, 

 the average cost is of the order of $7000 per man per year.* 



* These figures date from 1930; they have undoubtedly increased, but 

 the proportion will be unaffected. 



