204. Memorandum From Secretary of Defense McNamara to President Kennedy 0

This is in reply to your memorandum of 21 July 19621 on the way in which the new conclusions about nuclear test detection were evaluated and disseminated in the Government.

The information disclosed by Dr. Carl Romney of the Air Force Technical Application Center (AFTAC) at the 3 July meeting called by ACDA 2 had been developed in final form by Dr. Romney only over the previous weekend, and had been approved for dissemination outside AFTAC by [Page 515] Dr. Romney’s superiors. Dr. Romney had anticipated that the ACDA meeting would be much smaller than it actually was, and he had not had an opportunity to present his report to the chairman of the meeting, the Deputy Director of ACDA, before the meeting actually took place. The information could not, in my judgment, have been developed at an earlier date.

Apart from ensuring that chairmen of similar meetings inform themselves in advance of the content of scientific reports to be rendered, I believe the best way to avoid a recurrence of this situation is to provide for review of information to be disseminated outside ACDA by the Director of Defense Research and Engineering, as I have already directed.

The detailed facts and chronology underlying the 3 July disclosure are as follows:

The new information disclosed at the 3 July meeting related not to our ability to detect small underground shots from a distance, but rather to the fact that it now appears there are significantly fewer earthquakes in the Soviet Union, producing signals equivalent to an underground nuclear explosion of a given yield in a given medium, than had hitherto been assumed. Measurements of the Soviet underground explosion of 2 February, the French underground explosion of 1 May, and the U.S. Aardvark explosion of 12 May, all taken together, gave the first indication that a correction factor previously used for determining Soviet earthquake statistics was in error. These explosions occurred during a period when our measurement abilities were rapidly increasing, primarily as a result of the Vela program of nuclear testing detection research. The information obtained from the 1 May shot was not available for interpretation until June. When taken together with the information from the February shot, it led Dr. Romney to direct a further analysis of the 12 May data. This analysis was received by Dr. Romney at the end of June. At the same time he was working on a study requested by ACDA to evaluate the technical capabilities of a system of existing external national seismic stations to detect and identify Soviet nuclear explosions. On 29 June, he decided to work over the weekend in order to test the hypothesis that the earlier accepted conclusions about the level of Soviet earthquake activities were in error, so that he could include the results in his report to ACDA on 3 July. It should be noted that the previously accepted estimates of Soviet earthquake statistics had several times been presented to both government and consulting scientists, and had been unchallenged.

Dr. Romney consulted his superiors about the form and content of his report, including the new evidence that the number of earthquakes in the Soviet Union might have been previously overestimated. They [Page 516] instructed him to include this information, since they thought it was relevant and timely.

The 3 July meeting was attended by some 25 officials from the White House, AEC, ACDA, State, CIA, and DOD, including Dr. Wiesner, Ambassador Stelle, Mr. Fisher, and Dr. Long of ACDA, and Commissioner Haworth of AEC.

As you know, after the 3 July meeting there was a great concern that if these new estimates should leak to the press, the government would be in the embarrassing position of being accused of withholding information that would tend to ease the inspection problem in a nuclear test ban, and it was, therefore, decided that a news release be made. The Department of Defense concurred reluctantly with this position, on the grounds that (1) a release of this estimate alone would tend to exaggerate its significance, and it would give the impression of a major technical breakthrough; and (2) the new estimates had not been independently evaluated outside AFTAC. Therefore, the Department issued a release covering all Vela results and including the new Soviet earthquake estimates as only one of these results; and, in addition, prior to the release, a group of technical experts reviewed the AFTAC conclusions. Although some skepticism was expressed, the group generally agreed that the AFTAC conclusions were reasonable. Since then, the analysis of much more data tends to confirm the AFTAC conclusions.

AFTAC is recognized to have special scientific competence associated with detection of nuclear tests, not found elsewhere in government. Because of this competence, AFTAC has continuously received high priority assignments not only from other Defense agencies but also from such organizations as State, CIA, AEC, and even the Joint Committee on Atomic Energy. It has been very difficult for AFTAC to establish relative priorities amongst this continuous input of requirements. Therefore, I have decided to assign to the Director of Defense Research and Engineering the responsibility for establishing these relative priorities, and for the staff supervision of AFTAC in all technical activities not primarily a matter of intelligence collection and processing. The assignment will include responsibility for the evaluation, control, and dissemination of new results of AFTAC technical analyses.

Robert S. McNamara 3
  1. Source: Kennedy Library, National Security Files, Kaysen Series, Nuclear Testing, Vol. II, 7/62-2/63. Secret. In a July 28 note to McNamara, attached to a carbon copy of this memorandum, Harold Brown wrote that Adam Yarmolinsky, Special Assistant to the Secretary of Defense, and Jack Ruina, Director of the Advanced Research Projects Agency, Defense Research and Engineering, had drafted the memorandum on the previous day and, regarding the final paragraph, Solis Horwitz, DOD’s Director of Office of Organizational and Management Planning, “concurs in the proposed assignment of responsibility. So do I, though somewhat reluctantly.” (Washington National Records Center, RG 330, OSD Files: FRC 66 A 3542, Atomic 000.01-400.112, 1962)
  2. Reference is to National Security Action Memorandum No. 174, July 21, addressed to the Secretaries of State and Defense and the Director of Central Intelligence, which expressed the President’s puzzlement at the release of new data on the U.S. ability to detect small underground nuclear tests and requested these agency heads to “consider how matters can be reorganized to avoid a repetition of this unfortunate event.” (Department of State, S/S-NSC Files: Lot 72 D 316, NSAM 174) Secretary Rusk responded to NSAM No. 174 in an August 15 memorandum to the President. (Ibid.) Marshall S. Carter, Acting Director of Central Intelligence, replied in an August 31 memorandum to the President. (Ibid.)
  3. See footnote 3, Document 196.
  4. Printed from a copy that indicates McNamara signed the original.