252 ORGANIZING SCIENTIFIC RESEARCH FOR WAR 



invited to carry on their own research so far as they felt it must be done on 

 a top secret basis. There was no refusal to accept top secret research; rather 

 OSRD called for a conference before top secret projects were submitted so 

 that it might be convinced both that the classification was justified and that 

 OSRD was the only place where the desired work could be done. No top 

 secret project was in fact submitted to OSRD nor was any conference in- 

 volving OSRD called to consider submitting one. It is understood, however, 

 that several projects which were tentatively proposed as top secret were 

 marked down to secret before their submission to OSRD. The normal tend- 

 ency to play safe makes for overclassification and the interposition of a 

 "caution" sign in the case of top secret projects was a desirable step. 



When the Army or Navy submitted a project to OSRD, it indicated the 

 recommended classification, which was invariably adopted by OSRD. Occa- 

 sionally, the two Services would submit almost identical projects, but with 

 different classifications, and then an attempt would be made to get agree- 

 ment as to which OSRD should use. For projects originating in OSRD, the 

 tentative classification was established by the originating divisions; as the 

 divisions were working on related classified Service projects, there was little 

 difficulty in determining the appropriate classification. Circulars were issued 

 by the Executive Secretary from time to time on such topics as how to 

 determine whether an item should be classified, and how to mark, store, 

 transmit or reproduce classified items. 



As a general rule, OSRD contracts bore the same classification as the 

 projects to which they were devoted. The correspondence between contracts 

 and projects was not complete, however, as in some cases several projects 

 were handled under the same contract. Normally, the contract carried the 

 highest classification of any project worked on under it, with the contractor 

 being separately authorized to handle projects of lower classification accord- 

 ing to their respective rules. On occasion this was reversed, so that a contract 

 would carry a confidential classification when most of the projects under 

 it were so classified, and an occasional secret item would be called to the 

 attention of the contractor with instructions to apply to it the rules govern- 

 ing secret matter. 



One criticism of the OSRD practice which probably would apply to 

 security precautions generally was the persistence of a classification after the 

 reason for its establishment had ceased to exist. A periodic review of all 

 classified items would doubtless have shown many for which the classifica- 

 tion could have been lowered or even removed. The volume of current 

 work always seemed to be so great that the time could not be spared for 

 such a review until after the end of hostilities when there was a general 

 declassification. In retrospect it seems possible that the saving in time re- 

 sulting from handling documents of lower classification would have justi- 

 fied strenuous efforts to find the time for reclassification at an earlier date. 



