Approved For Release 2000/08/10 : CIA-RDP96-00787ROO0100060001-6 .tvtm!m:~. \International/ TO Mr. Manfred Gale (HQDA, DAMA-ZD) FROM H. E. Puthoff SUBJECT sRi/Gale Committee Meeting of 24-25 July IdEND DATE 3 August 1979 LOCAT10N Bld. 44 CIC We appreciated having an opportunity to discuss with you and members of your committee several aspects of our work during your recent visit to SRI. We welcome this kind of dialog and exchange of views with individuals of diverse backgrounds, who are willing to focus on issues of importance to us and our program. There were certain aspects of the meeting, however, which we felt were unsatisfactory, but for which we have specific remedial actions to propose. As you know from our agenda, we had intended to cover four major areas of interest: (1) an historical overview, including past applications to problems of interest to clients; (2) scientific questions with regard to RV protocols, judging, statistics, and so forth; (3) the goals and planned activities on present client programs; and (4) committee discussion with program remote viewers. Unfortunately, with the extended ad hoc discussions that took place during the opening presentation on historical material (Part 1) and the need to press on to a discussion of planned efforts on present client programs (Part 3), we all but skipped Part 2 on scientific questions (for example, none of the slides were shown). This seemed expedient at the time, given the apparent inappropriateness, in a group of such diverse interests, of getting involved in detailed discussions on specific technical issues (e.g., nuances as to protocol procedures, statistical evaluation techniques, etc.). This may have been "penny-wise and pound foolish," however, for all of us at SRI realized at one time or another in casual discussions with individual committee members during dinner breaks, etc., that there was a profound lack of understanding of some very basic issues. To cite some specific examples: (1) It was expressed that the RV target pool should be chosen and maintained by someone further removed from the experimentation than the principal investigators, and that access to the target pool should be limited. The above is in fact the procedure already in force. An independent target selection team generated the target pool and turned them over to the Project SSO who keeps them stored in a secure container. Targets are then withdrawn only under supervision of the SSO (who records their withdrawal) at the beginning of RV sessions. In Approved For Release 2000/08/10 : CIA-RDP96-00787ROO0100060001-6 SRI 2903 2/78 Approved For Release 2000/08/10 : CIA-RDP96-00787ROO0100060001-6 Mr. Manfred Gale Page 2 3 August 1979 particular, the RV session interviewers (Targ and Puthoff) had never seen the target list until your committee asked that it be brought to the conference room. (2) Concern was expressed that apparently successful RV results in our protocols might be due instead to a combination of remote viewer guesses of known Bay Area landmarks and/or artifactual sharpening of a subject's narrative by cues from the interviewer who is at least educated, if not directly knowledgeable as to the target possibilities. (We think that is was understood that the inter- viewer never knows the particular target.) The corollary to this concern is that although the RV function may exist, our protocols can't demonstrate it because of the above-stated circumstances. In fact, the statistical procedure we use takes into account at the outset the possibility that the target pool might be completely known to both remote viewer and interviewer--that it could be that The remote viewer and interviewer were poring over the target li.st during the session; in short, that the RV series is to be treated as belonging to that class of studies in which the elements of the target pool are known a priori to both remote viewer and interviewer, as in studies involving numbers or cards as targets. (In fact, we would go this apparent criticism one step further and assert that it would be naive to assume that by any change in protocol one could in principle avoid the assumption that the remote viewer knows the target pool.) Thus, when there is a statistically significant result in our protocols, the cause must lie elsewhere than remote viewer guess or interviewer cueing, as these possibilities are handled at a fundamental level by a statistical procedure that assumes the worst. This fact must be understood by the committee members if they are to assess the SRI program results properly. (3) It was repeatedly suggested that it would be better to dispense with the interviewer in an RV session so as to have a "cleaner" protocol, and that the use of an interviewer is somehow a methodological flaw. Such a suggestion indicates a complete failure to comprehend that success as obtained by use of the SRI RV protocols (as opposed to other procedures) is in large part due to an innovative design which incorporates a division of labor between remote viewer and interviewer designed to mirror the two primary modes of cerebral functioning; namely, the nonanalytic cognitive style (brain function) that pre- dominates in spatial pattern recognition and other holistic processing (and is hypothesized to predominate in psi functioning), and the Approved For Release 2000/08/10 : CIA-RDP96-00787ROO0100060001-6 Approved For Release 2000/08/10 : CIA-RDP96-00787ROO0100060001-6 Mr. Manfred Gale Page 3 3 August 1979 analytical.cognitive style that predominates in verbal and other analytical functioning. (Only very experienced remote viewers appear to have the ability to handle both cognitive styles simultaneously.) The removal of the burden of analytical functioning (by the interviewer) during exercise of the RV faculty appears to be an important ingredient for success. A change in this aspect of the methodology may give the appearance of a cleaner protocol (it is not), but it may also yield the lower level results characteristic of classical academic (as opposed to operational) studies. Our client programs, on the other hand, force us to develop techniques that yield results, and it cannot be stressed enough that we have developed the appropriate statistical procedures to handle such possibilities as remote viewer guessing and interviewer cueing or sharpening. To the degree that the committee's overall evaluation of the SRI program will touch on technical issues such as these, we would suggest that it is incumbent on the committee members to obtain a more thorough grounding in the technical details than was possible in the few hours available during the July meeting at SRI. We would doubt, for example, that anyone would be in a position to critique our present RV statistical procedures fairly and justly (not that there wasn't the potential because of the expertise represented) simply because we did not have an opportunity to present them, and, being a new approach in our program (Scott's direct-count-of-permutations method), we have not discussed them at any length in publications available to the committee. Since the approach (a) is specifically designed to handle narrative material of the remote viewing type, (b) takes into account the possibility of potential remote viewer/interviewer guessing and/or knowledge of the target pool, and (c) has been thoroughly investigated, used, and documented in the academic parapsychology community as the most conservative reference statistic to fall back on, one cannot assess remarks such as (1) - (3) above without at least an intuitive understanding of the assumptions and implications of this approach. Beyond this, there were other items that 'to our mind need clearing up, such as confusion of our work with other labs' work and statements, and little awareness that we have done considerable experimentation with simpler paranormal functions than remote viewing (e.g., computer-automated number-perception tasks, binary card tasks) to establish certain parameters of the more complicated functioning. (Such scientific legwork we did not present to the committee, as it seemed beyond the scope of interest expressed at the time.) Approved For Release 2000/08/10 : CIA-RDP96-00787ROO0100060001-6 Approved For Release 2000/08/10 : CIA-RDP96-00787ROO0100060001-6 Mr. Manfred Gale Page 4 3 August 1979 To remedy the shortcomings inherent in the brief orientation meeting of 24-25 July, to ensure that we do justice to the multitudinous concerns of the committee, and to ensure that the SRI program obtains a fair hearing from the committee, we propose that: (1) all committee members receive from your office a copy of this memo and the attached protocols to gain a better understanding of our procedures; (2) further discussions be held in smaller groups between SRI personnel and those members of the committee who are especially concerned with specific issues (e.g., protocol procedures and statistical approaches with Drs. Synder and Tang). We are available for such interactions at Menlo Park, Washington, or elsewhere, and would appreciate the opportunity to resolve these issues at the earliest convenience. Out of such interactions we could also expect as a bonus to define with greater clarity those scientific issues that need to be pursued in more detail should fundamentally-oriented research programs be set up in the future, as they must. Since we are mutually tasked by our sponsor to provide the best technical assessment possible of a field fraught with complex and subtle issues, we believe that additional actions such as we have proposed are necessary if we are to fulfill our mutual goals and responsibilities. Enclosed are copies of the memo and protocols for all the committee members, which we would appreciate your distributing. Also enclosed for use at your discretion are additional copies for Maj. Gen. E. R. Thompson, Dr. Ruth Davis, and the Hon. Walter Laberge. We remain at your disposal as to arrangements should you decide to follow up on the further interactions we have proposed. Approved For Release 2000/08/10 : CIA-RDP96-00787ROO0100060001-6 I ApprUqd For Release 2000/08/10 : CIA-RDP96-00787ROO0100060001-6 August 1979 0 STANDARD REMOTE VIEWING (RV) PROCEDURES: LOCAL SITES FC~11 Prepared by: H. E. Puthoff, Ph.D. R. Targ E. C. May, Ph.D. 333 Ravenswood Ave. * Menlo Park, California 94025 AInte Waked For R04eass-2CiDO/O(NIM'5104A-ROPga4]D-M7-ROODIO0060001-6 NQIL4~;07 Approved For Release 2000/08/10 : CIA-RDP96-00787ROO0100060001-6 STANDARD REMOTE VIEWING (RV) PROCEDURES (LOCAL SITES) Our present standard remote-viewing (RV) procedures are similar to those in our Proc. IEEE paper, "A Perceptual Channel for Information Transfer over Kilometer Distances: Historical Perspective and Recent Research."I The elements of the protocol, each of which is addressed below, consist of (1) basic procedural design; (2) remote viewer/interviewer roles; (3) target pool selection; (4) target storage and access; (5) remote viewer orientation; (6) interviewer behavior; (7) target person behavior; (8) post-experiment feedback; (9) evaluation procedure. 1 Approved For Release 2000/08/10 : CIA-RDP96-00787ROO0100060001-6 Approved For Release 2000/08/10 : CIA-RDP96-00787ROO0100060001-6 I BASIC PROCEDURAL DESIGN At the beginning of a trial, a remote viewer is closeted with an interviewer in an isolated windowless room of the Radio Physics Laboratory in the SRI complex to await an agreed-upon start time. At the same time a target person is sent, without communication with the remote viewer or interviewers remaining at SRI, to a target location somewhere in the San Francisco Bay Area (-500 square km). The target is determined by random-number access to a target pool of sealed travelling orders previously prepared by an independent experimental team and kept locked in a secure safe. The target pool consists of more than 50 target locations chosen from a target-rich environment. During a predetermined viewing period of 15 minutes duration, the remote viewer is asked to render drawings and describe into a tape recorder his impressions of the target site being visited by the outbound target person. The interviewer with the remote viewer is kept ignorant of the target and is therefore free to question him to clarify his descriptions without fear of cueing (overt or subliminal) as to the particular target. Since general knowledge of the San Francisco Bay Area target region on the part of the remote viewer and interviewer must be taken as a given, and since particular knowledge of the contents of the target pool is revealed as a series progresses, one must take into account the possi- bility that any particular description may be artifactually sharpened. (Such sharpening can in principle increase the apparent quality of the result only if there is functional remote viewing to begin with. It cannot in the absence of ESP produce an inflated result.) This sharpening possibility in the presence of an already functioning RV capability is handled in the statistical evaluation of the results by conservatively assuming at the outset that the series is to be treated as belonging to that class of studies in which the elements of the target pool are known a priori to both remote viewer and interviewer, as in studies involving numbers or cards as targets. 2 Approved For Release 2000/08/10 : CIA-RDP96-00787ROO0100060001-6 Approved For Release 2000/08/10 : CIA-RDP96-00787ROO0100060001-6 When the target person returns to SRI following the remote viewing period, the subject is then taken to the target site so that he may obtain direct feedback. Following a series of such trials over a several-day period, a formal blind judging procedure (described below) is used to evaluate the data and quantify the results. 3 Approved For Release 2000/08/10 : CIA-RDP96-00787ROO0100060001-6 Approved For Release 2000108110 : CIA-RDP96-00787ROO0100060001-6 II REMOTE VIEWER/INTERVIEWER ROLES An important methodological aspect of the SRI RV protocols revolves around the fact that the remote viewer/interviewer team constitutes a single information gathering unit in which the remote viewer's role is designed to be that of perceiver/information source, and the interviewer's role is designed to be that of analytical control. This division of labor is designed to mirror the two primary modes of cerebral functioning; namely, the nonanalytic cognitive style (related to brain function) that predominates in spatial pattern recognition and other holistic processing (and is hypothesized to predominate in psi functioning), and the analytical cognitive style that predominates in verbal and other analytical functioning. 2-4 (Only very experienced remote viewers appear to have the ability to handle both cognitive styles simultaneously.) The interviewer role, removing as it does the burden of analytical functioning during exercise of the RV faculty, appears to be a key element in generating the level of success required in operational programs, and we attribute the success of the SRI RV protocols in large part to this innovative design which appears to provide an appropriate match to the required functioning. 4 Approved For Release 2000108110 : CIA-RDP96-00787ROO0100060001-6 Approved For Release 2000/08/10 : CIA-RDP96-00787ROO0100060001-6 III TARGET POOL SELECTION Target locations in the San Francisco Bay Area are selected by a team of two Radio Physics Laboratory personnel who are not involved as inter- viewers in the experiments (to prevent direct knowledge of the target pool by the interviewers). The locations are chosen to satisfy the following criteria: (1) Target sites must be within a half-hour drive of the SRI Menlo Park complex so that a uniform target access time exists for all experiments. (2) The target pool is constructed to contain several targets of various types--that is, several fountains, several churches, several boathouses, and so forth--specifically to circumvent analysis strategies of the type "there was a fountain yesterday, so it is unlikely that there is a fountain today." Furthermore, targets of different types are not chosen to be particularly distinct from each other, so that overlapping features exist. In this manner the content of a given target, determined by random entry into the target pool, is essentially independent of the contents of other targets ("open-deck" design). (3) The definition of what constitutes each target is established in advance of the entire RV series by written descriptions on a set of 3" X 5" target cards. (Ex: Four Seasons Restaurant, on El Camino Real, just north of San Antonio Road. Stand under the entry arch and feel the bricks.) These cards constitute the outbound team's instructions at the beginning of the trial, and the judge's target list during the evaluation phase. 5 Approved For Release 2000/08/10 : CIA-RDP96-00787ROO0100060001-6 Approved For Release 2000/08/10 : CIA-RDP96-00787ROO0100060001-6 IV TARGET STORAGE AND ACCESS The target cards are numbered and placed in individual envelopes, similarly numbered, by the target selection team; they are then turned over to the project Special Security Officer (SSO) who maintains them in a GSA-approved secure container. At the start of an RV session the interviewer, remote viewer, and target person rendevous in the laboratory and establish the trial start time (30 minutes hence). The target person then leaves the laboratory for the SSO station, generates a random number in the presence of the SSO by the use of the random-number function on a Texas Instruments Model SR-51 hand calculator, obtains the associated envelope (which is recorded by the SSO), and departs for the target site. 6 Approved For Release 2000/08/10 : CIA-RDP96-00787ROO0100060001-6 Approved For Release 2000/08/10 : CIA-RDP96-00787ROO0100060001-6 V REMOTE VIEWER ORIENTATION During the period that the target person is enroute to the target, the interviewer and remote viewer have a period to relax and discuss the protocols. The goal of the interviewer during this period is to make it It safe" for the remote viewer to experience remote viewing. For the initial orientation of a new remote viewer, this typically includes a discussion as to how remote viewing appears to be a natural rather than abnormal function, and that many people appear to have done it successfully. The remote viewer is told that memory and imagination constitute noise in the channel, and therefore the closer he can get to raw uninter- preted imagery, the better. He is encouraged to report raw perception rather than analysis, since the former tends to be correct while the latter is often wrong. Since remote viewing is a difficult task, apparently similar to the perception of subliminal stimuli,6 it takes the full attentive powers of the remote viewer. Therefore, the environment, procedures, etc., are designed to be as natural and comfortable as possible so as to minimize the diversion of attention to anything other than the task at hand. No hypnosis, strobe lights, or sensory-derivation procedures are used, since in our view such (novel) environmental factors would divert some of the subject's much-needed attention. 7 Approved For Release 2000/08/10 : CIA-RDP96-00787ROO0100060001-6 Approved For Release 2000/08/10 : CIA-RDP96-00787ROO0100060001-6 VI INTERVIEWER BEHAVIOR The interviewer arranges ahead of time to have pen and paper available for drawing, and a tape recorder. The room lighting is somewhat subdued to prevent after-image highlights, shadows on eyelids, etc. When the agreed-upon RV trial time arrives, the interviewer simply asks the remote viewer to "describe what impressions come to mind with regard to where the target person is." The interviewer does not pressure the remote viewer to verbalize continuously; if he were to, the remote viewer might tend to embroider descriptions to please the interviewer, a well-known syndrome in behavioral studies of this type. If the remote viewer tends toward being analytical ("I see Macy's") the interviewer gently leads him into description, not analysis. ("You don't have to tell me where it is, just describe what you see.") This is the most important and difficult task of the interviewer, but is apparently necessary for good results, especially with inexperienced remote viewers. It is also useful for the interviewer to "surprise" the remote viewer with new viewpoints. ("Go above the scene and look down--what do you see? If you look to the left, what do you see?,') The remote viewer's viewpoint appears to shift rapidly with a question like this, and the data come through before the viewer's defenses activate to block it out. The shifting of viewpoint also obviates the problem of the remote viewer spending the entire session time giving meticulous detail on a relatively trivial item, such as a flower, which, even if correct, generally will be of little use in assessing the session. (Once a remote viewer feels he sees something, he tends to hang on to this perception rather than commit himself to a new viewpoint.) It is important to recognize again that with the division of 8 Approved For Release 2000/08/10 : CIA-RDP96-00787ROO0100060001-6 Approved For Release 2000/08/10 : CIA-RDP96-00787ROO0100060001-6 labor outlined in Section II it is the interviewer's (not the remote viewer's) responsibility to see that the necessary information to permit discrimination among the range of target possibilities is generated, the remote viewer's responsibility being confined to exercise of the RV faculty. The remote viewer is encouraged to sketch what he sees, even over his objections that he is not an artist,,can't sketch, etc. He may do so throughout, or wait until the end of the session if intermittent drawing would distract his concentration. Since drawings tend to be more accurate than verbalizations, this is an extremely important factor for good results. 9 Approved For Release 2000/08/10 : CIA-RDP96-00787ROO0100060001-6 Approved For Release 2000108110 : CIA-RDP96-00787ROO0100060001-6 VII TARGET PERSON BEHAVIOR After obtaining a target card in the manner described in Section VP the target person proceeds to the target site indicated. He is asked to come upon the target location at the starting time so that his view of it is fresh at the beginning of the remote viewing period. He is to then simply pay attention to the environment as dictated by instructions on the target card. At the end of the agreed-upon target viewing time of 15 minutes the target person returns to the lab. 10 Approved For Release 2000108110 : CIA-RDP96-00787ROO0100060001-6 Approved For Release 2000/08/10 : CIA-RDP96-00787ROO0100060001-6 VIII POST-EXPERIMENT FEEDBACK When the target person returns, and after all the raw data has been turned over to the SSO, the interviewer, remote viewer, and target person proceed directly to the target site for feedback. This helps to develop the remote viewer's sense of which aspects of his mental imaging process are correct, which are incorrect. This appears to bring the RV trial to closure for the remote viewer, so that when he has a following session, his mind is no longer involved with wondering how he did on the previous one. Only a very experienced subject can function well time after time without feedback, so this is done for each trial to optimize the potential for success. Approved For Release 2000/08/10 : CIA-RDP96-00787ROO0100060001-6 Approved For Release 2000/08/10 : CIA-RDP96-00787ROO0100060001-6 IX EVALUATION PROCEDURE In a sense, the most critical part of the remote-viewing procedure is the evaluation procedure. Any single experiment in remote viewing, even if perfect, could in principle be dismissed as possibly a coincidence. Further, any result less than pefect might be called into question as a generalized "grass is green, sky is blue" transcript that fits every target. Only blind differential discrimination of transcripts across a series of targets can provide a basis to discriminate between these dismissals and the RV interpretation. To obtain a numerical evaluation of the accuracy of a standard six- trial :remote viewing series with a given remote viewer, the results are subjected to independent judging on a blind basis by an SRI research analyst not otherwise associated with the series. In preparation for judging, the remote viewer's tapes are transcribed. The resulting transcripts are then edited only to the extent of deleting information which might act as artifactual cues to a judge, such as references to other targets, or phrases which might indicate the temporal order of the transcripts. The transcripts (including associated drawings) and target cards, each arranged in their own random order different from the order of target usage, are then turned over to the judge. The judge is instructed to visit the target locations on the basis of the target card instructions, and to blind rank order, on a scale of 1-6 (best to worst match), each of the six transcripts against each of the six target sites, generating a 6 X 6 matrix as in the example shown in Table 1. 12 Approved For Release 2000/08/10 : CIA-RDP96-00787ROO0100060001-6 Approved For Release 2000/08/10 : CIA-RDP96-00787ROO0100060001-6 TYPICAL 6 X 6 JUDGING MATRIX Distribution of Judge's Rankings TRANSCRIPT LETTER TARGET A B C D E F 1 (D 5 3 1 4 6 2 5 G) 2 6 4 3 3 5 6 0 2 3 4 4 1 5 3 0 4 6 5 5 4 2 6 0 3 6 6 4 2 5 3 A precise measure of the statistical significance of the matrix of target/transcript relations is given by a direct-count-of-permutations method of great generality.a It is an exact calculation method requiring no approximations such as normality assumptions. Furthermore, the judging process that went into generating the matrix is not required to be inde- pendent transcript-to-transcript nor target-to-target. Finally, the statistical evaluation procedure is general enough that, in addition to being applicable to the blind rank order procedure in use at the present time, it can be applied to analyses in which numerical estimates of target/transcript correspondences are made on the basis of other rank-order or rating scales. This includes rating 1-7, zero to complete correspondence; arbitrary scale rating arrived at by some complex procedure involving many factors such as occurs in multiple-judge voting; cases in which, for a 13 Approved For Release 2000/08/10 : CIA-RDP96-00787ROO0100060001-6 Approved For Release 2000/08/10 : CIA-RDP96-00787ROO0100060001-6 given target, several transcripts are given the same rating, all transcripts are rated zero, a few transcripts are assigned rank order numbers and the rest are assigned the mean of the remaining rank order numbers, and so forth. The only requirement is that no artifactual information is provided as to the order of targets and transcripts. In particular, it can be shown that if targets are used with replacement or are non-orthogonal, then the method applies even in the case in which there is trial-by-trial feedback and the target pool is known a priori to both remote viewer and interviewer. Thus the possibility of interviewer cueing or subject guessing based on a priori knowledge of the target pool is handled at a fundamental level by a statistical procedure that assumes the worst. The argument is as follows. In the absence of knowledge as to which transcript was generated in response to which target, one observes that in setting up the matrix there are n: possible ways to label the columns (transcripts), given any particular order of the rows (targets), and vice versa. Thus, there are n! possible matrices which could be constructed from the raw judging data, all of them equally likely a priori in the absence of knowledge as to the order targets and transcripts. Each has its associated sum of on the diagonal corresponding to a possible alignment of targets and transcripts. The significance level for the experiment is then determined by counting the number of possible matrices that would yield a result (diagonalsum) equal to or better (i.e., lower sum of ranks in the rank- order case, higher sum of scores in the correspondence-rating case, etc.) than that obtained for the matrix corresponding to the key, and dividing by n: This ratio gives the probability of obtaining by chance a result equal to or better than that obtained in the actual judging process. For the 6 X 6 matrix used as an example (Table 1) we have, by direct computer count of the 6: matrices obtained by interchanging columns, -3 p = 2/6: X 10 2.8 14 Approved For Release 2000/08/10 : CIA-RDP96-00787ROO0100060001-6 Approved For Release 2000/08/10 : CIA-RDP96-00787ROO0100060001-6 This statistical procedure, in use for more than two decades by many researchers, was specifically designed to handle narrative material of the remote viewing type, and it cannot be stressed enough that it is constructed sufficiently conservatively so as to apply even in the limiting case in which the target pool is completely known a priori to all involved, thus handling any possible contamination due to remote viewer guessing or interviewer cueing in protocols of the type used in the SRI RV procedure. 15 Approved For Release 2000/08/10 : CIA-RDP96-00787ROO0100060001-6 Approved For Release 2000/08/10 : CIA-RDP96-00787ROO0100060001-6 REFERENCES 1. H. E. Puthoff and R. Targ, "A Perceptual Channel for Information Transfer over Kilometer Distances: Historical Perspective and Recent Research ft Proc. IEEE, Vol. 64, pp. 329-354 (March 1976). 11 2. J. Ehrenwald , Cerebral Localization and the Psi Syndrome," J. of Nervous and Mental Disease, Vol. 161, No. 6, pp. 393-398. 3. R. Ornstein, The Nature of Human Consciousness, San Francisco, CA: Freeman, 1973, Ch. 7 and 8. 4. R. W. Sperry, "Cerebral Organization and Behavior," Science, Vol. 133, pp. 1749-1757 (1961). It 5. H. F. Dixon , Subliminal Perception and Parapsychology: Points of It Contact , Proc. of the XXVII Annual International Conference of the Parapsychology Foundation, Inc., New York (in press). It 6. C. Scott , On the Evaluation of Verbal Material in Parapsychology: A Discussion of Dr. Pratt's monograph," Jour. Soc. Psych. Res., Vol. 46, No. 752, pp. 79-90 (June 1972). 16 Approved For Release 2000/08/10 : CIA-RDP96-00787ROO0100060001-6