◆ SpookStack

36. (Continued) ‘ 43, For the better part of two years, CIA had been check- ing information obtained from refugee, defector, and agent sources with NPIC whenever it was apparent that the informa- tion was of a kind that could be verified or negated by aerial reconnaissance, In May 1962, NPIC began publishing a series of formal listings (Photographic Evaluation of Information on Cuba) ain which these reports Were evaluated in the light of photogra— phic.coverage. In the 7 issues of this publication between 31 May and 5 October NPIC examined 138 raw reports referred to it for comment. of this total, only three cited missile activity which could not be’ linked directly to the SA-2 and cruise mis- sSile deployments. NPIC'’s evidence negated those three. 44, When the first indications of build-up began to come in in August, *these procedures were further tightened. CIA current intelligence was ordered orally by the DD/I's office on about 14 August not to publish any information on the con- struction of missile bases in Cuba until they had been checked out with NPIC. (This instruction was in the field of intelli- gence technique rather than of policy; it had no relation to later restrictions; (see para 50). Between 14 August and mid- October this office sent NPIC 13 memoranda asking for a check on 25 separate reports containing information which was thought to raise the possibility of Soviet offensive weapons in Cuba, A great many more such reports were checked with NPIC_informally by telephone. In all cases, NPIC either lacked the” necessary coverage or made a negative finding. 45. On 20 August, the COMOR Targeting Working Group (chaired and staffed largely by CIA) set up the first compre- hensive card file system for Cuban targets. An example of its procedures is the handling of targets in the Sagua La Grande area. Based on refugee reporting, the COMOR Targeting Working Group on 27 August pinpointed four farms in this area as sus— pect missile sites, Readout of the 29 August coverage showed an SA-2 site near Sagua La Grande which apparently was the~basis for the reported activity there, and the target card was changed to show a confirmed SA-2 site. It should be noted that know- ledge that this site was in the area could have led analysts to misinterpret any subsequent reports of MRBM activity as part of the SAM development ,but in fact no such reports were received. 46. By September, the volume of agent and refugee report- ing had become very large indeed. During the month 882 re- ports on internal activities in Cuba were disseminated, exclu- Sive of telegraphic dissemination. (The CIA clandestine col- lectors report that their output represented only the small pub~ lishable fraction of the raw material collected.) A substantial -24-