RETROSPECT AND PROSPECT 331 



tends to put everything in the highest classification as a matter of caution, 

 and the scientist who wants everything in the lowest classification as a 

 matter of speed and ease of operation should be under continuous re-exami- 

 nation. The fact that part of a program should be carefully protected is not 

 always a valid reason for throwing safeguards (and delay) around other 

 parts. Time can be saved by adequate and continuous attention to this 

 aspect of a research program. 



As for clearance, the military authorities should not wait until an emer- 

 gency arises to decide what scientists they can trust. They should be pre- 

 pared to evaluate the evidence produced by an investigation and to stand 

 back of their findings. Clearly, this would not be an easy matter in view 

 of the amount of difficulty a disgruntled person can make, but national 

 security is too serious a matter for the issue to be dodged. 



Compartmentalization of information can be carried too far, and prob- 

 ably was by OSRD. In theory, it is sound. Its disadvantages can be largely 

 overcome by a continuous review by an adequate staff or by periodic con- 

 ferences of Program Chiefs in which each would outline his progress and 

 his problems in sufficient detail to permit his colleagues to determine where 

 an exchange of more detailed information would be profitable. Even pro- 

 grams which should be carried on independently of each other may have 

 components which are common and on which an exchange of information 

 would save valuable time and manpower. 



Contracting. The OSRD contracting procedure was effective. The contract 

 was deliberately designed to insure the maximum freedom for the exercise 

 of scientific imagination within the limits governing the proper expenditure 

 of Government funds. By contrast, the Service research contracts in exist- 

 ence at the time NDRC was organized seemed to enmesh the scientists in 

 such red tape as to make effective work difficult. 



With the transfer of OSRD projects to the Services incident to the de- 

 mobilization of OSRD, a more reasonable contract has obtained a foothold 

 in both the Army and the Navy. Whether it will become firmly established 

 is still uncertain. Judiciously used, a flexible contract can tie academic and 

 industrial research facilities into the miUtary research program in a very 

 effective manner. 



In time of war it will be possible to divert up to lOO per cent of the 

 country's research facilities to military problems and to expand them if 

 necessary. In time of peace, however, these research facilities quite properly 

 are used primarily for the advancement of science in the case of academic 

 institutions or the improvement of the competitive position of a company 

 in the case of industrial laboratories. Even in time of peace, however, some 

 portion of both academic and industrial facilities can be obtained for mili- 

 tary research. This percentage will vary from institution to institution and 

 industry to industry, but it might be reasonable to expect a maximum of, 



