Nuclear Museum Logo
Nuclear Museum Logo

National Museum of Nuclear Science & History

J. Carson Mark’s Interview

Manhattan Project Locations:

Dr. J. Carson Mark joined the Manhattan Project in May 1945 with a delegation of British scientists. He worked in the Los Alamos Scientific Laboratory, and in 1947, went on to head its Theoretical Division. Mark stayed on after the end of the Second World War as part of the project aimed at developing the hydrogen bomb. In this interview, Mark addresses the challenges involved in making a hydrogen bomb, including the design process and the conflicts between other scientists in the laboratory. He also discusses the Soviet hydrogen bomb, and the problems their project faced, despite Soviet espionage in the United States.

Date of Interview:
September 12, 2022
Location of the Interview:


Carson Mark: We shouldn’t have been making this damn bomb without trying to keep it secret from [Joseph] Stalin. We should’ve been talking to him like [Niels] Bohr said. [Klaus] Fuchs believed and took it into his own hands to make sure that the conversation went on. Of course, he didn’t need to because Stalin knew anyway. Not the technical details, but the general facts.

Richard Rhodes: Well, I think people often forget that Fuchs not only carried information to the Soviets, he also carried it to the British. Yes? I mean, the proliferation went through England as well as to the Soviet Union after the war.

Mark: He carried it to the British?

Rhodes: Well, he shared, to put it another way, we shared the bomb with the British, and after the war when they went home, they then had to work on their bomb, right?

Mark: Yeah, well, their bomb wasn’t much of a surprise to anybody. Looked an awful lot like the one that the team had been involved in building here.

Rhodes: Yeah.

Mark: And, indeed, the bomb that [Igor] Kurchatov and his team build wasn’t very different either.

Rhodes: No, it wasn’t, it was very similar.

Mark: The first bomb the French built wasn’t very different. So, there’s a lot of misunderstanding. These things have a life and character of their own, which sort of dictates a lot of what happens, given the Fuchs or non-Fuchs.

Rhodes: No, in fact, the pattern with the Russians—I just finished drafting about 400 pages on this book I’m working on about the Soviet program and the espionage, mostly because of all the documents that the Russians have made publicly available.

Mark: Stuff in their archives.

Rhodes: Yes. A lot of which—

Mark: You say you have just done 400 pages of that.

Rhodes: Yeah.

Mark: How long does it take you to do 400 pages of anything? It takes me about a day and half to do one page.

Rhodes: Well, the research took about four years. The writing’s a lot faster, thank God.

Mark: Well, that’s because you’ve forced yourself to do a lot.

Rhodes: One of the things that came up that I found intriguing was the question of how much about the thermonuclear Fuchs was able to pass along. 

Mark: Well, nothing about ours.

Rhodes: No, not about the one later, but he clearly knew everything up through the Super Conference.

Mark: Up through—

Rhodes: April of ’46.

Mark: April of ’46, sure. Well, of course, we never used that either. We don’t think it could be used, and the Russians were working on that set of ideas under [Yakov] Zeldovich.

Rhodes: Yes. Well, [Andrei] Sakharov says that when he came on to Zeldovich’s program, he felt that there was some material that must have come originally from espionage.

Mark: That could be, otherwise, who would have had such a damn fool idea? I love the remark that Sakharov in his memoirs makes in that connection, that the great virtue of his first idea was that feasibility didn’t have to be established. It wasn’t open to arguments as to whether the process would work or not.

Rhodes: Well, because of Mike.

Mark: Huh?

Rhodes: You mean, because of the Mike shot here?

Mark: No. His first idea preceded Mike. Sakharov’s first idea was, well, in 19—soon after he went to the institution.

Rhodes: ’48, ’49?

Mark: ’48 or ‘9. And he had a thing he called the layer cake.

Rhodes: Oh, yes, right.

Mark: And, if you took one look at the layer cake and think of what needs to go on, you don’t have to doubt that something will happen.

Rhodes: Ah, right, good.

Mark: You don’t have to establish the feasibility of the process. But, if you took a look at the results of the Super Conference in April of ’46, or took a look at what Zeldovich was doing, the first thing you have to ask yourself is, “My God, will it work that way or not?”

Rhodes: Oh, I see.

Mark: Or, will it work at all? And, that’s what Sakharov said very quickly, that his first idea had the lovely feature that feasibility, as quite separate from the extent to which it works—

Rhodes: Why is it feasible? Because—

Mark: Oh, because you just take a look at the physical processes you’re asked to think of, or which are involved, and you say, “If I heat this up, something is bound to happen.” 

My question is: would it happen to an exciting extent or only to a poor extent? That—you have to work to get a feeling for. But, with the classical super, you had the basic question, and I believe we still have it, we don’t know if that crazy thing would work that way or not.

Rhodes: [Edward] Teller has talked over the years about the idea of radiation compression was discussed at that conference.

Mark: Oh, at that conference.

Rhodes: Is that true?

Mark: Not that I know of. The fact that radiation would be around doing something or other was obvious, was almost certainly referred to and was certainly in everybody’s mind. But, not in the sense that it later came to be thought of. Edward has, I know he’s been saying that, he’s been trying to make it completely obvious to everyone that [Stanislaw] Ulam didn’t contribute anything.

Rhodes: Yeah, right.

Mark: And, that has had a great effect on what he says.

Rhodes: Yeah, I think that’s very clear as one listens to it.

Mark: Those two people knew each other quite well, and I knew them both. Each was aware that the other was a pretty bright person, sharp intelligence. Ulam used to make witty, pointed, scornful, oh, really shamefully disreputable remarks about Teller when Teller wasn’t there. Once in a while, his feelings about Teller couldn’t have escaped Edward’s notice. Edward reciprocated those feelings generously. So, they each were talking down the other, and that went on for years.

Ulam felt that he invented the new approach to the hydrogen bomb, and Teller didn’t wish to recognize that, couldn’t bring himself to recognize it, has taken almost every occasion he could, not absolutely every one, to deny that Ulam contributed anything.

I think I know exactly what happened in the interaction of those two. Edward would violently disagree with what I would say. It would be much closer to Ulam’s view of how it happened. But, I’ve forgotten which of the recent books about Teller, there’s an interview quoted with Jay, what’s his name, the president’s science advisor.

Rhodes: Oh, yes, yes.

Mark: The guy who’s in the hospital.

Rhodes: Yes, I know who you mean.

Mark: A very angry remark on Edward’s part about of the intellectual quality as this claim that, “Well, he’s a good scientist, therefore, he wouldn’t do anything wrong.”

Rhodes: When the FBI was interviewing Fuchs back in ’49 and ’50—

Mark: When the FBI was doing—

Rhodes: Was interviewing Fuchs in prison in England, Klaus Fuchs, at one point they asked him about something to do with his knowledge about the Super, and he laughed and said, “Well, I invented the mechanism for exploding that thing.” And, he was referring, I think, to a patent that he had with [John] Von Neumann, and there’s no references to what the patent might have been.

Mark: I’ve seen the patent. I don’t know why it’s left to seem so mysterious. Of course, a lot of things are left to seem mysterious because they’re classified. And, they don’t need to be classified in many cases, but to explain what this is, which wouldn’t convey anything surprising to anybody, would not be allowed because it is classified, hasn’t been unclassified.

Rhodes: The only word that I saw connected with it, I think he described it as an implosion system.

Mark: Oh, well it was.

Rhodes: I guess I’m referring more to his claim that this was an important invention for at least the classical Super.

Mark: Well, it’s related to the classical Super. It was not in any important way different from the ideas that were built into the George shot that Edward Teller keeps referring to as an important approach to the Super, even though it was devised and thought of and pictured entirely as an adjunct to the classical Super.

Rhodes: That’s interesting, because I wanted to ask you exactly about the George shot, which the record is, at least the public record, is strangely silent about.

Mark: Oh, it’s a horrible mess.

Rhodes: Is the George shot something that you, here at Los Alamos, had programmed and planned along the way?

Mark: Not really. We, of course, checked through the details, produced the design, but it was not a thing that we were busily working at, except as a sequel to the Russian test.

Rhodes: That was its origin? As a response to a Russian test?

Mark: Well, it was a response to the Russian test of August, ’49. What shall we do? Well, Luis Strauss sort of automatically said, and I’m not sure to whom, I guess to anybody who’d listen, “We must make a quantum jump.” That Guff!

“We must go to work on the Super,” which he had heard glowing accounts back in 1944 or ‘5 from Teller.

So, before it was really decided that the lab should direct a lot of attention to the thermonuclear program, it had been thought in Washington that that’s what the country should do. Thought mainly by people who didn’t understand the state of affairs at all, didn’t know anything except the glowing pictures of what forty megatons might be like if realized.

Edward happened to be here at the time of the Russian shot. He had taken the year away from Chicago and was going to spend the whole year with the theoretical group at the lab. And, since it wasn’t a new idea to him that we should work on the thermonuclear weapon, it was not terribly surprising that the lab found itself, what shall we do that bears on improving our status in the thermonuclear weapon. That’s obviously what ought to receive attention now, says Edward. Not too many disagreed, except that the content of what should be done was not really fully indicated by just that statement. [George] Gamow happened to be here at the same time. He was not too unimaginative a person either.

It was agreed we must do something that relates to the thermonuclear business, what would be a good thing that we could turn attention to. There were at least two different proposals, one mainly talked of and introduced by Gamow and one which Teller favored. Both had to do with notions that fitted into the picture of the classical Super as one of the things we might be in a position to do, and that might be useful in that connection, would have to be done in that connection sometime anyway, if I went that route. And, it was from that rather vague scrambling of ideas that the detailed form of the George shot emerged. It was the pattern that Edward had favored. It turned out that it was very much the pattern of the Fuchs/Von Neumann patent.

Rhodes: Was it coincidental that is the pattern of the patent, or was it by deliberate choice?

Mark: Well, it wasn’t discussed in those terms, but it was indeed very similar to the idea that Fuchs and Von Neumann had sketched out in 1945, I guess. Look, it wasn’t, it wasn’t by any means a copy from that, but it did about the same thing, and related to the Super in about the same way as their patent, namely a possible way of getting the first step from the starting fission bomb towards heating deuterium to a point where it might burn. And, that’s what the George shot was about.

It was new experience. Well, I’d describe it a little bit as if it were old hat. It was doing things which we had never taken seriously to do before, which had been sort of sketched out in an arm-waving fashion.

Rhodes: Can you describe at all at this point what the classical Super was? Is it still totally classified, or is there something you can say about it?

Mark: I think it’s been written in a way, which with your research in the Russian archives won’t be mysterious. It was the idea that deuterium could be set burning if you got it hot enough, and perhaps in a fission bomb might provide the sort of temperature level that you would need. So, let’s think we have a long pipe full of liquid deuterium, and we have a fission bomb, which we set off at one end of it, with the idea that we will heat that end sufficiently that a burning wave will get started and proceed along the pipe, the burning wave being the reaction, thermonuclear reaction of deuterium. Now, there’s a classical Super.

There’s some questions that have to be filled in, answers have to be filled in. Amongst them, if you heat one end of a deuterium pipe like that, will a burning wave, in fact, go along, run along it. Will it detonate like a stick of high explosive? That’s kind of a central question. Of course, you’d have to ask how hot do I have to make it before that’ll happen. But, even if I make it very hot, will that happen? Open question. It was still open when [Harry] Truman said, “Let’s work on the hydrogen bomb.”

Rhodes: Yeah, yeah.

Mark: And, that was the picture that one had of how the hydrogen bomb would work. Poor Harry didn’t know that the damn thing had this characteristic that you still had to prove whether anything would proceed that way or not, it wasn’t clear. All the work that had been done was to attempt to establish that and it hadn’t managed. That still left the question: how hot does it have to be? And that still leaves the question, “Here’s my fission bomb. Will it provide the kind of temperature I need, and how do I pattern things so that it does that?” Another open set of questions.

Well, it was getting fairly clear that the fission bomb would not provide the kind of temperature that one would need. We didn’t know what temperature you would need, or if there was any temperature that would work. But, in any event, the temperature would be higher than the fission bomb could provide. But, if that weren’t a stick of deuterium, it might be a stick of deuterium and tritium, and that’s a quite different style of object. A fission bomb could provide a temperature that would be likely to start that sort of thing burning. Of course, that wouldn’t make a hydrogen bomb because tritium was so expensive, and a hydrogen bomb was exciting because it was so cheap. Deuterium, you can get from the ocean.

Rhodes: Yes, right.

Mark: Tritium, you can’t.

Rhodes: So, when Ulam in his autobiography describes those calculations, he kept adding more tritium to the formula, as it were, to get the temperature lower, and it still wouldn’t propagate.

Mark: That still didn’t answer the question whether it would propagate or not.

Rhodes: Yeah, but it did say there was going to need to be quite a lot of tritium, right?

Mark: It did say there’d have to be some amount of tritium. Edward promised people in Washington that they’d get by with a certain amount. He had no particular basis for the amount that he mentioned, except that it didn’t appall them. He chose an amount that had that property. It didn’t necessarily have the property of starting up a Super.

Rhodes: Someone described the George shot to me as piping radiation to a small DT pallet.

Mark: That’s right.

Rhodes: And, that, presumably, might be the ignition system.

Mark: That was getting some thermonuclear reaction going using a fission bomb for a start.

Rhodes: Yeah. But, with the secondary start in the form of the DT.

Mark: DT, it was thought, would burn at the temperatures that a fission bomb could provide. Herb York doesn’t describe the thing to that sort of extent that you could picture it. He did refer to the fact that it got a small amount of DT burning.

Rhodes: Yeah, yeah.

Mark: And, there was a monster of a fission bomb. It was uninteresting. If you had to build super bombs that way, you would probably decide you didn’t want them. But, that wasn’t the frame of mind we were in. We had never had DT burning before, and it was very exciting to find out if one could persuade it to burn. One knew that it could be made to burn, but would this gadget that we build do it? For that reason, you had to test it, and it did it, and so you were encouraged to say, “Well, at least we’re doing things fairly well, maybe correctly.” Because, what we described is apparently pretty close to what happened.

Rhodes: If you were using radiation for that, was that a deliberate choice, rather than using the whole thermal—

Mark: Oh, it’s the choice that’s forced on you. You don’t decide whether you’ll use radiation or use something else. It’s a property of material that the kind of temperatures that you’re talking about that the radiation is there, and is the most certain, fastest moving thing that progresses from where the fission bomb is to where you might have your DT pellet. And, if you decided you didn’t want the radiation, you couldn’t really strain it out. It would force itself on your attention.

Rhodes: Yes. Was that at all suggested then of the Teller-Ulam idea?

Mark: Huh?

Rhodes: Was that shot, that design, at all suggestive of the Teller-Ulam idea? Did it contribute it to it eventually?

Mark: Well, look, it suggested, it’s true that the Teller-Ulam idea involved radiation. It’s true that Ulam though that you could use the signals that proceeded through material, hydrodynamic shock, that’s certainly there, too. It doesn’t move as rapidly as radiation does. It’s harder to control and direct, and so if you’re trying to exercise Ulam’s idea of using hydrodynamic shock, you’d find, by God, the radiation gets there first, so let’s use that instead.

Teller’s then supposed to have proposed that it’d be better to use radiation than to rely on material, signals carried by material. Well, he was right. It was simpler, but you couldn’t have avoided it, had you sat down to design the thing, asking now what’s the material shock doing, where is it, how fast does it move. You’d say, dear God, the radiation is going faster, it’s there, so let’s concentrate on that.

So, it was hardly an important circumstance that Teller thought of radiation, whereas Ulam thought of the signal that would traverse material. The fact that Edward thought of radiation was natural, because he had been involved in much more detailed work on the George shot than had Ulam. And, in the George shot, one was looking all the time at radiation as the sort of natural emanation from the immediately adjacent fission region, exploding fission region. There’s material shocks there, too.

In fact, that’s what caused all the consternation at Hiroshima, the material shock started from the bomb and swept on through the air. But, if you’re dealing with that sort of phenomenon, the close-in state of affairs in the immediate environs of an exploding fission region, you can’t escape the fact that there’ll be material shocks, of course, and there’ll be radiation, of course. If the bomb is very efficient, the radiation will be dominant. If the bomb is at low efficiency, the radiation will be minor. The temperature will be lower.

Rhodes: Would it have been possible to build a Mike with a 1945 fission weapon? Did the weapon have to be improved or the fission device have to be improved in order to get to that point?

Mark: Yes and no. When this Ulam-Teller breakthrough was first pictured, we didn’t have an improved fission bomb. Before we went ahead with the experiment with the Mike shot, we did.

Rhodes: So, the levitated composite core hadn’t come in as of ’51?

Mark: Oh, a composite core isn’t—

Rhodes: Is not an improvement?

Mark: Not necessarily. The improvement that was significant was to have the bomb smaller and lighter and more efficient. But, one could’ve done it with a Trinity bomb, or one could’ve done something of that type with a Trinity bomb. But, that Mike shot, as you have read, was a monstrous affair, 67,000 pounds or something, and it would’ve been bigger that if it was a Trinity bomb, quite a bit bigger.

Rhodes: Well, Harold Agnew, I think, tells a story about changing the core at the last minute, too, to get a little more yield.

Mark: I think the reason was not so much to increase the yield as it was to decrease the chance that it pre-detonate and not give a proper yield.

Rhodes: Oh.

Mark: The last minute, well, it was just about the last minute, but the shot was in November. The last minute started in August.

Rhodes: The way he tells it, it happened the day before, you know.

Mark: Yeah. Well, that was almost correct in that this was going on in the Pacific, whereas the change had to be proposed here and people like [Jay] Wechsler had to be involved in accomplishing the change.

Rhodes: Was the amount of tritium in the Mike shot quite small?

Mark: Yes, it was quite small. Let me put it this way, there was tritium in it. It was put in I’d say to strengthen the design, make things more certain, push them further in a favorable direction. It was later discovered that things would’ve worked all right without any. So, in things that were built later, didn’t have it.

Rhodes: Because, they went to lithium, or even without the lithium, they wouldn’t have needed it?

Mark: No, we just understood, a bit more confidence.

Mark: It was a radically different sort of package. The fact that one was using a fission bomb to compress the thermonuclear fuel and by compress one means just make it a great deal higher density than one had ever considered in connection with the classical Super. It changed the nature and pattern of events that were involved, so that your statement sounds as if there was a connection between the two, and there wasn’t much. It was a radically different situation, and the argument that you might need tritium and maybe tritium wouldn’t be enough supplied to the classical Super, wasn’t a point of discussion in this case.

Rhodes: You chose to go with a cryogenic design rather than with lithium. Why?

Mark: You insist upon there being a logical and well-thought-out explanation for everything, which is very often not the case. Why?

Well, the cryogenic design, which liquid deuterium certainly was, was different in another way, which is more important than whether it was cryogenic or not. Thinking of the proposition to burn deuterium with the cryogenic design, you had nothing but deuterium, the material whose behavior it was necessary to think of was an infinite medium of deuterons. With lithium deuteride, you had this inert atom and its electrons of the lithium as well as deuterons. And, you had to allow for its presence. There were as many lithium atoms as there were deuterons and the thing we were interested in happening was that deuteron reaction, deuteron with deuteron.

There wasn’t a prominent reaction of a deuteron with a lithium. Lithium was inert in that respect. In fact, it was a diluent, and certainly a complication. And, the cryogenic pattern avoided that complication. It introduced some physical complications in construction and handling, but those one knew how to deal with. They had nothing to do with any of the thermonuclear behavior. The fact that you put in a dewar, and the fact that you had to have a cooling system didn’t affect the behavior of the thermonuclear fuel, whereas the lithium did. So, the lithium was a diluent and a complicating item.

Now, you’ve heard, of course, of the great virtue of lithium in that it provides you with a free source of tritons. That’s only really true of lithium-6. We didn’t have large quantities of separated lithium isotopes. We set out to get them and by 1954 we had them. We could’ve had them earlier if we’d known enough to go after them.

Rhodes: Known enough in what sense? You knew there was, it was the reaction.

Mark: Well, the description of the process, the burning process of pure deuterium is much simpler than the description of the burning process with either lithium-6 or normal lithium deuteride. The description of the compression of liquid deuterium is simpler than of the compound, the deuteride. To avoid discussing the lithium seemed like a virtue. It was a fact then, and later when we came to appreciate this more clearly, that although it was a diluent, the lithium-6 was actually a shot in the arm for the behavior of that sort of situation. And, we didn’t have to have been afraid that we would have trouble computing the compression of lithium deuteride.

It turned out that we were capable of doing that fairly well. The fact that the burning reaction was more complicated, we were capable of describing that all right, too. But, every departure from the simplest picture seemed like something to avoid.  Until one found out that things worked better than we’d expected, and then you didn’t need to worry about the fact that something that might interfere, might not interfere enough to bother you.

Rhodes: Yeah. We talked to Harold Agnew in San Diego a few days ago, and he mentioned that after a while it became clear that you could use ordinary lithium, that it would make lithium-6. Is that how it worked? He didn’t explain the reaction and I haven’t looked it up, so.

Mark: No, there’s a little lithium-6 in ordinary lithium, six percent.

Rhodes: Oh, so it’s really not enriched, yeah.

Mark: The thing that we learned only after a while was that whereas neutrons, whether they’re fast or slow, get absorbed by lithium-6 and produce a triton, which is immediately then available within the fuel as an enriching feature of the fuel, makes it more burnable because of the deuterium and tritium combination than if the triton weren’t there. So I said the lithium was a diluent, just as lithium it is a diluent, but lithium-6 more than replaces the trouble caused by its presence by exchanging so easily into a triton.

What we didn’t appreciate at first was that lithium-7, the other isotope, the more frequent isotope also provides tritons, but only for neutrons of fairly high energy. Lithium-6 provides it for all neutrons, high energy or ones that have settled down to the temperature of the material you’ve got. The neutrons that get produced in connection with the burning of deuterium include a component of high-energy neutrons, and that group makes more tritons with lithium-7 than we had anticipated. So, lithium-7 diluting and cluttering things up and complicating them as a positive aspect was stronger that we’d suspected, and consequently, one could use it in places where we’d supposed you might not.

Rhodes: Does this explain the extra yield in the Bravo shot, this reaction?

Mark: That failure to understand the real behavior of lithium-7 certainly has a considerable amount, maybe the major amount to do with the Bravo shot going off better than we’d expected.

Rhodes: I spoke at some length with some of the Russian scientists about their Joe 4, the bomb they called their first hydrogen bomb. You probably have had this conversation directly or indirectly with Yulii Khariton, Yuri Smirnov, those are two people I’ve had contact with. They are, of course, very concerned to claim that that was a hydrogen bomb, whether a “true” hydrogen bomb or not.

Mark: Well, look, some Americans, included [Hans] Bethe, I guess, has said that Joe 4 wasn’t a true hydrogen bomb. I can understand that complaint, whereas Khariton, ‘Hariton’, whatever his name is, claims credit for its being a bona fide hydrogen bomb. I have to agree with the Russians on that score. After all, Joe 4 was a physical object of about the same size as the Trinity bomb.

They trucked it off in a plane, dropped it, and it gave thirty times as much yield, or twenty times as much yield. It was a tremendous augmentation of the yield of Joe 1 or Trinity. And, that came through because of thermonuclear reactions. Now, the extra yield was indeed provided by fissions, not by D-D reactions. It was provided by fast neutrons causing fission in U-238, but, they were thermonuclear neutrons. Gave that extra yield.

Well, what was the hydrogen bomb? The hydrogen bomb was a bomb with a lot more yield. Okay, the Russian bomb had that characteristic. It got it without adding large amounts of fissile material. That was the feature of the hydrogen bomb, was to get the extra yield without having to use up extra U-235. Well, the Russian bomb had that feature.

The hydrogen bomb also had the feature that it might give ten megatons or so. The Russian bomb didn’t do that, and indeed, you couldn’t have pushed that pattern of things to give ten megatons without getting a size that was unusable. They managed to get 400 kilotons without going to an unreasonable or even a heavier size. And, they did it by using thermonuclear reactions. Want to call that a hydrogen bomb? Well, why not?

Rhodes: They put the fusion yield of that device at about 15 to 20%, which is comparable as I understand it, to many of our hydrogen bombs with uranium blankets.

Mark: Sure.

Rhodes: They say, and this I just learned confidentially from one of them, that they tested that up to one megaton, that design, successfully.

Mark: I’m not sure that’s true. They were prepared to push it to a megaton. Did they test it?

Rhodes: Sorry, I may be misquoting. I’m trying to remember what Yuri Smirnov told me, that it could have gone, I think he told me it could have gone up to a megaton.

Mark: Now, who was doing the telling of this?

Rhodes: Yuri Smirnov, who is one of the designers of the big bomb.

Mark: One of the designers—

Rhodes: Of the big one, the 58-megaton one. He worked with [Andrei] Sakharov. He didn’t work on Joe 4, I don’t believe, I think that was before his time. But, he’s kind of a historian of science for them.

Mark: Was he the one who wrote to the bulletin?

Rhodes: Yes, with Khariton.

Mark: Khariton? Okay. Well, my understanding without me saying I know this, and I haven’t talked to any, on these points at all, is that they were pretty well-pleased with Joe 4, and thought of it as a hydrogen bomb. And, the reason they thought of it as such was what we’ve just been talking about.

They used thermonuclear neutrons to provide a good part of the yield, and didn’t get it by goosing it up with a lot of extra fissile material. And, they didn’t get it by increasing its size appreciably. So, they were entitled to be pleased with it, and to call it a hydrogen bomb, because they didn’t have any doubt that you could make a big explosion if you put in lots of U-235. They knew that, they just didn’t want to do it.

Rhodes: Is that design—

Mark: It’s not the hydrogen bomb in the sense that it’s got a potential for arbitrarily large yield.

Rhodes: Right, yeah.

Mark: It’s not a hydrogen bomb in the sense that most of the energy comes from thermonuclear reactions.

Rhodes: Good.

Mark: So, Bethe says it’s not a true hydrogen bomb. Well, I think that’s swallowing—arguing over a gnat, really. Who knows? If you were hit by the bomb, you wouldn’t know whether it was hydrogen or not.

Rhodes: True.

Mark: Now, the team was told to get a weaponized version of Joe 4, and this has gone through in Sakharov’s memoirs. They were told to work on weaponized Joe 4. Unfortunately, they were annoyed to be told that, because they knew it would be a deuce of a lot better to spend their time on the third idea, which went beyond Joe 4.

In fact, it was the Ulam-Teller breakthrough, and they thought that was much more promising. But, they were under directions to build a bomb of the Joe 4 style for the next test. I’ve forgotten the number of the test they did, and it was that that I think either was a megaton, or could’ve been a megaton, and it was tested. It was tested just before the first true hydrogen bomb—

Rhodes: Oh, yes, yes that’s right.

Mark: In November ’55.

Rhodes: That’s right, yeah.

Mark: But, I don’t remember the yield of that shot. And, I think it did what they were counting on, except that having done that they rejected the thing immediately.

Rhodes: Was it comparable to our Alarm Clock?

Mark: Oh, the first one was identical to the Alarm Clock. Romanov says that in an article in Sakharov Remembered, calls it the layer cake, says that the pattern was similar to one invented by Teller and called the Alarm Clock. And, that’s correct.

The Alarm Clock as an idea came up here in 1946, end of the Summer of 1946, and mainly Rittenmeyer worked on it. He worked on it for at least a year or a little more. He had some help. But, just why, we come back to one of these things which you might want to say, “Why did you do that? Why didn’t you do this other thing which would make more sense?”

The Alarm Clock came on the scene, Teller’s excited and enthusiastic proposal, as a way to build a hydrogen bomb. The hydrogen bomb, however, carried with it the sort of irrelevant trappings of—it wasn’t a hydrogen bomb unless it gave a megaton or multi-megatons, and it wasn’t a hydrogen bomb unless it had the potential to increase the yield indefinitely.

So, instead of approaching it as it seems as if the Russians approached Joe 4, by saying, “What can you do along with this new idea in the Trinity-size system?” which is what the Russians seem to have done, and they got a half a megaton. What got asked here was how big would it have to be to give a megaton or ten megatons or something or other? And the answer came back was it’d have to be too big.

Now, of course, it didn’t have to be bigger than the Trinity bomb, but it did if it had to carry these trappings. Semantics. You see, why didn’t you realize that you were screwing things up by taking that approach? I think we realize now, at least it’s my conviction that we did screw it up by taking that approach. The hydrogen bomb had been engraved in mind to have certain features, and therefore, you tried to develop something those features. And, that’s something which looked as if it would work in principle, was too big to be of interest, ten tons or something like that.

Rhodes: Why did it get big so fast? If you can get half a megaton for a Trinity-size, why does it suddenly become gigantic? I think Herb York said to me, “Well, it would’ve almost had to be as big as Mike.”

Mark: I can’t be sure I can explain that. And, Rittenmeyer, I don’t think, remembers what thoughts he had in mind at the time. The Alarm Clock, as it was introduced and as it was worked on, was a system that you got, you made a number of layers and you wanted to have it started in such a way that it would burn through and continue the burning through a number of layers. I think Joe 4 only had one layer. Didn’t ask the question, “What do we have to do in order to have it burn through a layer and then ignite another set of layers?” That makes it expand pretty rapidly.

We gave up the Alarm Clock, because we didn’t have anything which had the properties that were in mind in connection with it, that had a weight or a cost that had a practical interest. The weight was too big. If you were going to start the thing off in such a way that you went through one layer and that, diverting of that incited the burning of the next layer and that incited the burning, and you had a propagating system, you needed to start with a fission bomb of possibly a megaton size all by itself.

Now, at that time, we had a perfectly clear way of making twenty kilotons, period. So, we dropped it. And, we never put it to a test in the form which had we happened to think of, it could’ve been interesting in the sense that Joe 4 was.

Rhodes: There was a later device that was also called Alarm Clock. Was there any connection between them?

Mark: Once we had the idea of using a fission bomb to compress a thermonuclear assembly, we could construct or picture an Alarm Clock style of assembly with the layers that I speak of, two, four, seven. And, we could compress that sort of object, and we did try a thing which had those characteristics.

Rhodes: I see, so there was a connection, yeah, I wondered.

Mark: And, they worked perfectly well, and it could’ve been the basis of a hydrogen bomb arsenal had we chosen. Maybe Jay Wechsler is responsible for not having taken that road. The Alarm Clock was spherical assembly with a set of layers, however many you decide you can afford. And, it’s now to be in a case with a fission bomb that’s going to produce the energy to compress this mess. You’ve got to support all that inside a bomb case. That’s to be competitive with a cylinder that you also support inside a bomb case, and also have a fission bomb to handle the first stage. But, you can easily see how you’re going to support a cylinder inside a case, just make the ends a little long and stick them in brackets in the cylinder in the case.

With a sphere, you’ve got to support it preferably with a weightless filling. You can’t protrude, you don’t have a sphere anymore if you put big knobs on the end to hold it up. What end? There isn’t any end to a sphere.

Rhodes: That sounds like Jay.

Mark: That’s the part that Jay was much more clearly aware of than I, and so he didn’t like the Alarm Clock. The business of having it mounted in such a way that it would take the jolting that you had in a plane ride, or the accelerations you had just from dropping the bomb made that pattern very awkward. It was dropped for that reason.

Rhodes: Couldn’t it have been a cylindrical implosion system?

Mark: Huh?

Rhodes: Couldn’t you have done that with cylindrical implosion system?

Mark: No, we were depending in the Alarm Clock on a hot center that spread.

Rhodes: Oh, oh, oh, okay.

Mark: Rather than a hot point that went on down the line.

Rhodes: Was that Soviet Joe 4 test disturbing at Los Alamos? Did it cause disturbance here in terms of where we were and where they were?

Mark: Well, I don’t remember it in those terms. I remember our being very intensely involved in trying to reconstruct Joe 4, figure out what they had done on the basis of the debris evidence that we had. And, that was quite an effort, intensive effort, involved Bethe and [Enrico] Fermi both over a considerable period and a bunch of the rest of us sitting around the table. 

Rhodes: Did those discussions of the Joe 4 fallout lead to right conclusions?

Mark: Oh, yeah. They managed to speak of an object physically quite similar to what Joe 4 must have been. They didn’t lead us to want to emulate it.

Rhodes: No. When I was in Moscow, I talked with some physicists from Arzamas and I remembered a comment you had made about needing to wait for calculations until there was some computers in your history of the thermonuclear program here, right after the war. You talked about, in fact, I think Teller at some point said in a paper, “We need to delay this work for a couple of years until we have adequate digital computers.”

Mark: Yeah.

Rhodes: And, I asked one of them what did they do, knowing how poor their digital computers still are. And, he said, “We just used theoretical physicists to do these calculations.”

Mark: Well, you see, they never, that remark of Teller’s, which was made, I think, in 1947 related to both the classical Super and the Alarm Clock. Well, Teller said, “Defer intensive work for a couple of years, until calculations can be made.”

Well, the calculations, the computer that was supposed to come along in a couple of years came along in more like five. And, the calculations of this progress of the Super was also made about five years after that time, five or six, and indicated not so much proper behavior as trouble.

Also, by the time the calculation was possible, we had the staged pattern in mind, and that didn’t require—oh, we used tremendous amounts of computer time, but it didn’t require it to the same extent, to establish feasibility or operability. It required it to check details, to compare whether twenty-five centimeters was better than twenty, or things like that. And, you’re quite sure that either one of them would work, but you wondered which would come out better if you were to compare the two. That’s why you really needed computers. So, they just used theoretical physicists, but they never did, they never did go ahead with the Zeldovich pattern.

Rhodes: Right, yeah.

Mark: Now, you mentioned the people that you spoke to, you apparently didn’t run into or didn’t meet with Yuri Trutnev?

Rhodes: No, I did not get to meet with him. He was out at Arzamas, so they wouldn’t let me go to Arzamas at that time. So, instead, they offered to bring some guys into Moscow, but they turned out to be—what can I say, they weren’t their leading men.

Mark: Oh, well, Khariton was a leading man.

Rhodes: I didn’t get to talk to him personally.

Mark: You did not.

Rhodes: I sent him questions by email occasionally, but I didn’t get to personally interview him, no.

Mark: So, you did talk to Smirnov.

Rhodes: We talk a lot, also by email. I talked personally to another one who worked with Sakharov named Adamski, Victor Adamski. He has a piece in the Sakharov Remembered book. I talked with Lev Altshuler, who also was involved in the early work.

Mark: Okay.

Rhodes: Why was the Mike device so physically complicated in terms of all the different layers and materials and so forth? How did you arrive at that configuration? Was it just—

Mark: I’m not quite sure what you’re referring to.

Rhodes: Well, I’ve heard, I’ve heard comments like there nineteen different layers of materials inside the device, and it’s—

Mark: Where did you get that information?

Rhodes: Who said that to me? Bill Van Dorn, I think. Someone who was here, I guess, working on a history at one point.

Mark: Well, if you got it from an historian, I’m not surprised. 

Rhodes: But, it was a rather complicated device, was it not, just in terms of structures?

Mark: Well, not really. It had a uranium container for liquid deuterium. Inside that container, apart from liquid deuterium, there was a structure with just one, there was a central column which ran the length of the cylinder. It wasn’t frightfully complicated. Now, then there were arrangements for a dewar, and that led to several layers of material to provide the heat shield, to keep the heat loss from the cryogenic system to a minimum.

But, that wasn’t really a complication. That was a structure, people knew how to made dewars, if you were going to have liquid deuterium in a heavy object, you know from decades ago how to make a dewar to go with that. If it’s bigger than any dewar you’ve ever made before, it presents you with a problem to make sure there’s no leaks. That will present you with a problem. If the dewar’s also got to put up with some jiggling and shaking and bumping, you have a problem. But, of a classical sort, not of a sort that relates to the behavior of the device or as a novelty that adheres to a nuclear bomb construction.

Rhodes: The radiation splashing down from the fission device had to go through the dewar walls?

Mark: No. I think the dewar wall was inside the uranium wall that contained the thermonuclear features.

Rhodes: Ah, okay.

Mark: No, I’m really puzzled by the origin of that comment you made. It’s likely to have some complications if you think of trying to mount a container of liquid deuterium and a large thing surrounding it, and to mount that away from the wall, inside of which the radiation is to be free to flow. You better pay a lot of attention to how you do this mounting. You must have enough solid contacts of metal so that a heavy weight is suspended and held in the position you want, but you mustn’t have any unnecessary metal contacts between the cold and the hot regions, or else you’ll complicate your cryogenic requirements, which are already pretty demanding.

Rhodes: By the time of the Mike shot, had you started work on future weapons designs? I gather that you moved pretty quickly after that, too.

Mark: By the time of the Mike shot?

Rhodes: Yeah.

Mark: Well, we had several designs quite well along. If you take a cue from the Castle program, which followed about fifteen months later, am I right? First of March, after first of November, however many months that is, a year and a bit. And, we had, what, four or five devices out in the Pacific by March of ’54, having had to wait until after November of ’52 to confirm that there was nothing wrong with the November of ’52 design, that that way of doing things came off all right.

The only option we had was to do it similarly with the next round of devices. If we had had a large warning that you mustn’t do whatever it was you did in November for Mike, you mustn’t do things that way, we would have to have started over with those devices. But, the fact was that Mike came out all right. The plans for the later devices could proceed and stay proceeding on the assumption that if we make it this way it will be like the other and presumably work as well, or work.

Rhodes: Was the primary factor then that led to small hydrogen bombs reducing the primary?

Mark: Was the which?

Rhodes: The bombs got smaller through the years.

Mark: Very much.

Rhodes: Yes, was that primarily a matter of making the primary smaller, the fission bomb smaller?

Mark: Well, that was certainly a feature. They couldn’t have gotten smaller in diameter without the fission bomb diameter being brought down.

Rhodes: Was that the controlling feature?

Mark: Well, it may have been controlling, but I’m not sure that that’s the same as driving. Very strong considerations that pushed in the same direction were the change in the carriers. When Mike was fired and the Castle bombs, the carrier for any bombs was a big plane. There may have been missiles coming in sight, but there weren’t any missiles in action. Am I right?

Rhodes: Right, yeah.

Mark: I think so. Well, already the Castle bombs were smaller than Mike. They got that way partly by having been able to skip the cryogenic feature. I think maybe we began to have a smaller primary we could call on. I’ve forgotten what plane was being aimed at, thirty-seven, forty-seven, not the twenty-nine anymore. But, the Castle bombs were all portable by some existing or coming plane. Some of them were pretty monstrous, too, 40,000 pounds. Air Force was never daunted by a challenge to carry 40,000 pounds, but they snapped at the idea of carrying twenty-five instead pretty rapidly. 40,000 pounds was a threat to the wing structure of some of the planes that they had in sight.

However, by the time you came to talk of carrying things in missiles, and that followed very quickly after Castle, you didn’t want even 20,000 pounds anymore, but you wanted to get down to 1,000. That was a very strongly motivating item in getting the bombs smaller.

Rhodes: To want to and to be able to are two different things. You must have found ways to make these large mechanisms very small and elegant and much more efficient. Did they get much more efficient also?

Mark: Well, you’ve heard that the booster was an important development in connection with the weapons.

Rhodes: Sure.

Mark: Now, that didn’t come onto the scene until about 1956 or ‘8.

Rhodes: That was the Item test, wasn’t it?

Mark: Well, the Item was in ’51. That didn’t produce a boosted weapon. That produced proof that the boosted principle could be used. It was quite a bit after that before we had managed to adapt that to go in manageable weapons. Item couldn’t have been a weapon. It was an experiment on a tower. To be a weapon, it had to be more rugged, smaller, and there were other features that had to be changed before we had a boosted weapon, as contrasted with a proof of the principle of boosting. It took us about five years to get that finally checked through. Now, that had a lot to do with reducing the size of the trigger or the primary, and increasing the yield at a given size. And, those were both required in order to proceed with the kinds of things which would go in the sort of ICBMs, which were coming in sight.

Copyright 1983 Richard Rhodes. This transcript may not be quoted, reproduced, or redistributed in whole or in part by any means except with the written permission of Richard Rhodes. Exclusive rights granted to Atomic Heritage Foundation.