Foreign Relations of the United States, 1958–1960, Volume III, National Security Policy; Arms Control and Disarmament, Microfiche Supplement

515. Memorandum of Conversation Among Principals of Geneva Test Group¹

Washington, January 8, 1960

SUBJECT

• Course of Action to be Pursued Upon Resumption of Nuclear Tests Conference, January 12

PARTICIPANTS

• State

  • Acting Secretary Dillon
  • Mr. Farley—S/AE
  • Mr. Popper—S/AE
  • Mr. Baker—S/AE
  • Mr. Mark—S/AE
  • Mr. Dubs—SOV

• White House

  • Dr. Kistiakowsky
  • Mr. Gordon Gray
  • Mr. Keeny

• D.O.D.

  • Secretary Gates
  • Gen. Loper
  • Mr. Knight
  • Gen. Fox

• C.I.A.

  • Mr. Allen Dulles
  • Mr. Brent
  • Dr. Scoville

• A.E.C.

  • Chairman McCone
  • Gen. Starbird
  • Dr. English
  • Dr. Walske

Mr. Dillon referred to his letter of January 7 to Foreign Secretary Lloyd outlining in general terms the course of action to be pursued upon resumption of the Geneva negotiations January 12th and noted that the main element to be filled in was the question of the threshold. He called upon Dr. Kistiakowsky to present recommendations based upon the work of a group of advisers which had met on Wednesday to consider this question.

Dr. Kistiakowsky said the group of scientists which had worked both in California and here, had concluded it is technically possible to define a threshold in terms of the magnitude of seismic events detected by the control system. He distributed copies of a definition of magnitude based upon straightforward measurements which, apart from minor quibbles, would probably be acceptable from a technical standpoint even to Soviet scientists. He said that the technical group had concluded that above a magnitude of 4.25 the definition he had distributed could be readily applied. Above magnitude 4.5 it would be possible to test safely up to about half the corresponding kiloton yield since the magnitude would be uncertain by some 0.1 to 0.2 magnitude units and the yield of a device can not be predicted in advance with complete accuracy. He said the one minor point on which they had not reached complete agreement was merely a matter of choice, namely how far out you should go with the instruments used to measure the critical magnitude. He distributed a table indicating the number of events that would be located and unidentified at various magnitudes if the U.S. definition and criteria were applied. The table is attached. He said that the group had decided to recommend, although it did not represent a purely technical judgment, that a magnitude of 4.75 would be a useful point at which to define the threshold. This magnitude corresponds to the Blanca shot in the Hardtack series, so we have a definite fix on instrument readings for this magnitude. If the Soviets stick to their own interpretation of the Hardtack data they would, under that threshold, not be able to test more than two kiloton devices. The U.S. on the basis of its interpretation would feel free to test devices up to and possibly above 10 kilotons, without the use of decoupling, in a medium comparable to that of the Rainier shot. The level of the threshold to be chosen, however, is flexible. If it is higher there is less justification for a substantial monitoring system and the U.S. position becomes correspondingly less defensible. If it is lower the number of inspections becomes quite large if a substantial deterrent is to be maintained. The incentive to undertake concealment by use of decoupling also becomes greater. He noted that it was considered technically feasible at a cost of some £25 million to construct a hole large enough to contain a 70 kiloton explosion which was not substantially larger for test purposes than the legal tests at 4.75 magnitude. Thus the 4.75 level suggested represents a compromise between these conflicting considerations. In response to questions as to what magnitude scale the 4.75 magnitude represented, Dr. Kistiakowsky said that it was based on the Gutenberg and Richter scale as developed in 1956. The numbers of unidentified events in the table were calculated by the rigorous standards of U.S. criteria set forth at the recent technical conference. Opinion differs as to how much further a good seismologist could narrow down the choice. Estimates vary from 30 per cent to a factor of 10. A fair guess is that the unidentified shown on the table would be reduced by a factor of 2. If seismic research continues, this factor can be improved. One point on which delegation scientists were sensitive in recent technical discussions was the Soviet reminder that little had been done for improvement of the system by U.S. scientists since 1958.

He said that the group had also made non-scientific estimates of the number of inspections that might be suggested to the principals for consideration. This could be either a percentage quota or a numerical quota. The pluses and minuses of the two alternative methods of calculation include the following factors:

A percentage basis (1) affords protection against yearly variations in the number of earthquakes and existing uncertainties in our estimates, and

(2) if based only on the number of unidentified events, presents a much greater incentive to the Soviet Union to participate in improvements of the technical capabilities of the system because such improvements would reduce the number of unidentified events; and gives the Soviets an incentive to install stations in Communist China in order to reduce the number of unidentified events eligible for inspection.

A numerical basis (1) Would presumably be easier to negotiate since it accepts the quota proposal and makes the exact level of inspection clear in advance.

(2) Avoids the need for criteria on which there was no meeting of minds among Soviet-U.S. experts.

(3) Is easier to apply in practice.

(4) Is simple and understandable from the standpoint of public relations.

There are several alternative means of expressing the level of inspection.

(1) We could say 10 per cent of all located seismic events above the magnitude of 4.75 were to be inspected. This method would by-pass the criteria problem but would provide little incentive to improve the system unless the treaty also called for reductions in the number of inspections in connection with improvements.

(2) We could call for inspection of 20 per cent of all located seismic events deemed eligible for inspection by the criteria the U.S. has proposed. This proposal would be based on the proposal we have already tabled that inspection of 20 per cent of events below 5 kilotons be inspected. If this basis for calculation were proposed, the percentage would apply to the number of events unidentified without stations in Red China until such stations would be installed.

(3) The third alternative would be to simply say that there would be 10 inspections per year.

Mr. Dillon said it makes little difference what exact level is selected within the general range Dr. Kistiakowsky recommends. The question depends not only on technical factors but may involve specific military needs of the Department of Defense and AEC. If, for example, from a technical standpoint we could accept a 10 kiloton threshold, and DOD had strong needs for tests in the 15 or 20 kiloton range, this factor could be taken into account. Dr. Kistiakowsky said the group of scientists had not explicitly considered the question of U.S. testing needs in its recommendation, but that Dr. Harold Brown of Livermore had joined in the recommendations. He is thoroughly familiar with our testing needs and presumably took this factor into account. Mr. Gates said these technical questions relating to testing needs were mostly for AEC to decide. He noted that after all these complex calculations we had arrived back at the Lloyd proposal for 10 inspections. Mr. Dillon commented that Lloyd’s 10 inspections had applied to a comprehensive ban whereas the present suggestion was for 10 inspections applied to events above approximately 20 kilotons. General Loper stressed that all unidentified events should be eligible for inspection. Mr. Gates said that just as military needs are usually over-estimated in the first instance as a margin of safety, we should play it safe in our initial proposal regarding the level of inspection. Mr. Dillon said that we might, in fact, need a higher deterrent factor than 20 per cent and recalled our previous proposal that all events above 5 kilotons be inspected.

Dr. Kistiakowsky said the figure of 10 would include all events we thought really suspicious after our seismologist had intelligently applied diagnostic criteria to the unidentified events.

Mr. Dillon reverted to the relationship of military needs to a threshold and commented that, if you needed a 30 kiloton shot for weapons development, this could be legitimately conducted by the use of decoupling. That would not be cheating since the size of the signal rather than the size of the explosion would be the criterion of legality. Mr. Gray asked whether there was any point in thinking in terms of a control system including Communist China. Dr. Kistiakowsky said that in this estimate he was looking at the column of figures calculated without stations in Communist China. He said that there might be substantial advantage stipulating that testing allowable under the threshold be accompanied by advance notification of the time and place of the tests. We could argue that this builds into the proposal a mechanism for improvement of the system. Mr. Dillon asked what were the prospects for progressing to lower thresholds. He noted that Dr. Kistiakowsky said it was difficult to apply the definition below 4.25. He asked whether, for example, a threshold of 4.6 would be feasible. Dr. Kistiakowsky said we could go even further down than 4.6 if the auxiliary criteria could be codified. He mentioned that Dr. Oliver already speaks of a factor of 10 reduction in unidentified events by unilateral application of the auxiliary criteria.

Mr. Dillon said it was important from a public relations standpoint to emphasize that one threshold would not apply forever. We should emphasize our willingness, as better methods of development are to be put into effect, to lower the threshold progressively. This could also be done by increasing the number of on-site inspections. Mr. Gates questioned the idea of stressing “we could do better later”, since it implies the U.S. threshold quota we now propose is too high. Mr. McCone said it could also be argued by the Soviets that we should call off all tests below a threshold unilaterally from the beginning.

Dr. Kistiakowsky said we still have on the books a proposal to ban all tests in outer space. There is a well established fact of gaps in detection of capabilities at great distances, even though testing at such distances would cost a great deal and could not be conducted on a large scale for some years. Our policy with respect to such a ban is, however, primarily a political rather than a technical problem. Mr. Dillon suggested that we might go up to 100,000 kilometers on the basis of ground based instruments instead of proposing the total ban based upon very expensive satellite systems. Mr. McCone said he felt the same way.

Mr. Gates said he felt he must reflect the very strong JCS concern about including an underwater ban. Political difficulties with underwater tests are much the same as with those in the atmosphere. We should not, however, propose to include such a ban at an early stage. A very real military problem is involved. Without further experimentation we simply don’t yet know how to kill a submarine with nuclear weapons without endangering the crew of the destroyer delivering the weapon. We would have to give in on the ban of underwater tests in the last analysis, but it would be unwise to begin by doing so. Dr. Kistiakowsky asked whether bubbles from an underwater shot could contaminate the atmosphere. Mr. Gates said this was not known. In addition there was the difficulty of possible decoupling of underwater shots too if you are willing to spend enough. He believed the political factors were equal to the military factors in the case of underwater explosions, but believed it was good not to start with the full offer.

Mr. Dillon noted that the JCS also wants to test in the atmosphere. Mr. Gates agreed but said the underwater question was somewhat different and would involve not weapons development but learning to use weapons already in our stockpile. It is a safety problem to be resolved. We don’t want to use weapons that would kill our own people.

Dr. Kistiakowsky said that a 10 kt underwater shot could be conducted provided one million tons of concrete were used, were made in a structure the size of the Empire State Building, and were placed two kilometers deep in the ocean and when, of course, the Department of Defense is rich enough to do this.

Mr. Dillon said we should all go home and study the figures that had been presented and should get additional figures from Dr. Kistiakowsky for the intermediate ranges between 4.75 and 5. There would be no great difference to us between 4.75 and 4.85, for example, and we would need the advice of the DOD and AEC in making a choice.

Mr. McCone asked whether this was a question of adequate safeguards or if the elbow room we needed for military requirements is a legitimate consideration. Mr. Dillon said we could move the threshold somewhat on the latter basis. The problem with the threshold of 5 is that it may not provide sufficient justification for control systems in the Soviet Union. We have little feeling about the precise level so long as it provides adequate justification for the system we propose. The Soviet criteria would have identified events only above the Blanca range. In this sense the Soviets too were admitting a threshold of about this level, so that it would not be in a position to object too effectively to our proposal. Dr. Kistiakowsky noted by way of clarification that a statement that the technical situation was difficult below 4.25 applies only if we assume the Geneva system. We could go lower if the system were modified as regards either spacing or instrumentation.

Mr. McCone said that while further study is required he was in a position to make some comment even now on the threshold problem. He said that this would seem to be an extremely complicated type of negotiation; that it has many pitfalls from the negotiating point of view and that it gives Tsarapkin considerable room to maneuver us into an uncomfortable position. He said we should think carefully as to whether we want to start down this road at all. We should decide in advance with the UK just what kind of a treaty we are negotiating, that is, whether our aim is a comprehensive ban or whether it is our intention to stop testing only in the areas where effective safeguards can be accomplished. If the latter is agreed, both with the UK and Russians, we may safely go this way.

Mr. Dillon said he agreed that we should not get into specific numbers in the Geneva negotiations unless the Russians agreed in principle to the threshold approach. As to the level of threshold, Mr. McCone said that recognizing the political factors the AEC would prefer the threshold of 5 but could live with a threshold of 4.75.

Mr. Gates asked whether it is realistic to expect that, if you can test legally below a threshold, it would be politically possible to conduct an extensive test program below it—for example, 10 tests a year. Mr. Dillon said he didn’t think the question of whether you have a threshold would make any difference in the political reaction to a test program. [illegible in the original] would be equally concerned about underground testing whether it was permitted by a treaty with the Soviets or not. Testing would be easier with the threshold to the extent the threshold would permit us to say that we would in the future ban all tests that were controllable. The initial agreement would show our willingness and good intentions. Mr. Farley added that in addition, the kind of a treaty we were discussing would dispose of the pollution problem.

Mr. McCone said we now know that the threshold is technically feasible; the next question is whether it is good. Mr. Dillon said that he thinks it is a good approach. He believed we should say we want a phased treaty covering the atmosphere (leaving the underwater question aside for now) and underground tests above a given threshold. We should point out this approach by-passes the criteria disagreement with Soviet scientists. We should say we do not want to go into detail on the proposal until it is agreed in principle. We should not get into the outer space question unless the Soviets agree on the underground part of our proposal. Then we could say once the principle of threshold was accepted, that a comparable principle should apply in outer space. We should tell the U.K. as soon as possible what our ideas are on a specific threshold. Mr. Gray asked what the U.K.’s attitude would be. Mr. Dillon said that they may very well accept the threshold proposal. They could be expected to proclaim a moratorium on tests below the threshold and would want the U.S. to do the same. Mr. Gray commented that we might ask the U.K. in this case whether they would want the information from our own “immoral” tests.

Mr. Allen Dulles noted that the higher you put the threshold the fewer on-site inspection you get. Mr. Dillon said you wouldn’t get many any way.

Mr. McCone said that a high threshold reduces the incentive for decoupling. The Soviets, he believed, would discuss a number more readily than a percentage. The Soviets might also propose that we keep the same number of inspections and move the threshold down. He said the AEC believed the simplest method would be to base inspections upon a percentage of all events located. We can couple this with a statement of willingness to renegotiate the level as detection and identification capabilities improve. We would want a higher percentage depending on whether Red China was included in the control system. In response to a question, Mr. Keeny said Red China was relatively more important than other areas of Asia because it was not only adjacent to the Soviet Union but adjacent to highly seismic areas within the Soviet Union.

Dr. Kistiakowsky commented on the project Vela. So far little money has been made available and little work had been done. Defense has quite recently made available $8 million for the project. In addition to this project, however, there are two other research needs: (1) research on instrumentation for ground-based detection in outer space, and (2) research on satellite instrumentation for detection. General Starbird said the AEC was spending $2 million in development of outer space instruments and another million in construction of a hole for underground tests. General Loper said the original memorandum of agreement last April did not appear to call for such an expensive program as called for by the more far-reaching objective stated for Project Vela which ARPA had designed.

Dr. Kistiakowsky said we needed research programs in all three fields, seismic, ground stations for high altitude detection, and satellite instrumentation. Mr. McCone said the AEC was making preparations for the decoupling shots but the announcement of these preparations had been delayed since November, first because of the U.N. debate and then because of the technical talks in Geneva. We could not make the announcement as easily after January 12 as now because it might result in some embarrassment to the resumed Geneva discussions. Since now seems to be the best time he proposed the text already cleared interdepartmentally should be released. Mr. Farley confirmed that we had no objection to the text and added this would seem to be a suitable time. Mr. Dillon agreed but believed the President should see it, because of his personal interest in these matters. Mr. Gray agreed the President should see it. General Starbird said that failure to make the announcement had not prevented preparations so far. They had proceeded at the Nevada test grounds under the general implication that we were merely keeping the test grounds in order. There have been no leaks so far but it would be difficult to seek sites for additional tests in Louisiana if we were unable to state our purpose.

Mr. McCone asked for discussion of the question of whether we should make the announcement at all. Mr. Dillon commented that the President’s attitude is often “why say something if we don’t have to”. Mr. Gates said it is not a problem of DOD concern. Mr. Dillon said we need to go ahead and get the sites we need for additional tests, and the desirability of the announcement would seem to depend on whether the announcement was necessary in order to proceed with site selection. Mr. McCone said he is willing to take the risk of proceeding without announcement, but it should be taken in the recognition that at some point we will have to make the announcement after the fact. It was agreed to defer the announcement and to proceed with the selection of sites.

Mr. Dillon said the group should meet again next week after it had considered the threshold proposal advanced by Dr. Kistiakowsky. He said that Mr. Merchant would represent the Department at the meeting. In the meantime instructions to the Delegation now would be along lines of the January 6 letter to Selwyn Lloyd, pertinent paragraphs of which he read to the meeting.

Attachment

M	N	Nu	Nu*	YR
5.25	32	3	8	110
5.00	59	8	27	43
4.75	105	25	70	19
4.50	185	95	150	8
4.25	335	245	300	3.3
4.00	580	490	545	1.45

M	Magnitude
N	Total number of earthquakes in the USSR of magnitude greater than M
Nu	Number of unidentified events in the USSR of magnitude greater than M
Nu*	Number of unidentified events if stations are limited to USSR and US
YR	Yield equivalent in KT for Rainer coupling

1. Source: Discussion of appropriate threshold for underground testing. Secret. 8 pp. NARA, RG 59, Secretary’s Memoranda of Conversation: Lot 64 D 199.