James Conant’s Atomic Bomb Sketch? (1943)

by Alex Wellerstein, published May 25th, 2012

I had fun with the little visual mystery I posted last Friday, so here’s another one I’ve been chewing over for awhile.

Drawings of “official” atomic bomb designs are rare. (Where “official” means “created by people who actually build bombs.”) It’s the sort of thing which is generally kept close — what are released are generally extremely sanitized abstractions, which are then elaborated upon by people without security clearances (like John Coster-Mullen).

So I was somewhat surprised to find, buried in some files of the Office of Scientific Research and Development, this drawing which appears to have been made by none other than James B. Conant, then the President of Harvard University:

That looks an awful lot like the drawing of a gun-type nuclear weapon. But is it?

Conant, of course, was a major scientific administrator during the war. He was a chemist by training, and was no stranger to secret projects: during World War I, he had worked to develop lewisite for use in Europe while working at the “Mousetrap” facility in Cleveland, so called because once you went in, you were never supposed to come out.1 The chemical munitions that Conant worked on were never used in the war; the armistice came just before they were to be shipped out. During World War II, Conant was pals and colleagues with Vannevar Bush, head of the OSRD, and the two of them did quite a lot of work on early atomic development policy.

The context of the sketch is apparently a note from Conant to Bush, dated January 21, 1943 (with notes that it was amended March 10, 1943).  I say “apparently” because, while this follows the other sequentially in the file, it isn’t clear that they are attached or from the same period. (The handwriting is Conant’s though, which is something. Don’t read too much into the fact that the pages look different; one is just scanned in black and white, the other as grayscale.)2

The note itself is pretty hard to decode; it is in Conant’s nearly-impossible handwriting. The basic gist of it is that he is estimating how much enriched uranium they can product at Oak Ridge and what that implies about when a bomb would be ready (he seems to think one would ready by September 1944, and then later updates the note to push it back a bit).3

On the “drawing” page itself, there is a list (anything in italics is written by me, trying to make sense of his handwriting):

(1) Metallurgy
(2) cows [!?! see below]
(3) Development of technique for handling material in bulk
.                                          70-80, 90% of critical
(4) What cases are effective? [could this mean casings?]
(5) Further [???] [???] for cross section
(6) No. of neutrons for 49
(7) Capture + emission[?]  of neutr.                          (Bohr)
(8) Cross section of scattering[?]
(9) Firing problem
.                              length of time first mass stays in
(10) Source of neutrons Neutron source
(11) Effect of dilution
(12) Protection against thermal neutrons                (25)

To my eye, even with the ambiguity caused by his bad handwriting, it looks like a list of problems to tackle when thinking about designing a bomb the first time. What will the metallurgy of U-235 or plutonium be like? How will you shape these materials safely on a lathe? Was sorts of casings or reflectors will be best? How do you handle this stuff without getting totally irradiated? How many neutrons will plutonium emit per fission? How will you make a neutron initiator? What’s the engineering of the actual bomb assembly going to look like? And so on.

Except, of course, for “cows,” which I find inexplicable. It’s not a codename I’m familiar with. I am almost surely transcribing it wrong, but it looks a lot like “cows”:

Cows. Hmm. There were some cows involved in the Manhattan Project in a peripheral way, but I doubt he was thinking about that at this point. More likely is I’m making a garble of his handwriting again, but now that I’ve seen “cows,” I can’t stop seeing it. (Got a better guess? Let me know.)

Anyway, what it looks like to me is the result of either brainstorming or notes from a meeting that Conant was having, all of which seems to pertain to weapon design issues. So the idea that he might have sketched a crude gun-type design at the bottom of it isn’t fanciful in and of itself.

The drawing seems to show one “40 lb” piece of fissile material at the bottom of a gun barrel, with the cross section of a ring of the same stuff at the other end of it inside some sort of heavy neutron reflector or tamper. There are some other numbers nearby; it seems to say “10 meters, 30 ft.” Is that meant to be the length of the gun barrel? It would be pretty long, much longer than any of the actual bombs estimated for combat, but it might just be a back-of-the-envelope guess.

The bomb — if it is a bomb — that Conant has sketched out here doesn’t look much like Little Boy actually looked, but it doesn’t look wildly different than Thin Man, the plutonium gun-type bomb that was pursued before Little Boy.

Experimental bomb casings from the aborted “Thin Man” plutonium gun design. There are early “Fat Man” casings designs in the background.

The actual Little Boy weapon used (according to John Coster-Mullen) a cylindrical projectile that weighed around 85 lbs, and the “spike” that it was shot into (not the other way around) weighed 56 lbs, bringing it to a total of 141 lbs of fissile material, considerably more than is shown in this sketch. But still, the entire point of the list seems to be that they don’t know the details at that point.

The other possibility is that this isn’t a bomb at all, and that it is some kind of “tickling the dragon’s tail” criticality experiment. But that’s a much more boring conclusion.

Instead of pointing out how crude and inaccurate the drawing is, though, I’m still just amazed that it was hiding on that microfilm, waiting to be stumbled upon. It’s oh so rare to see bomb designs in “the wild,” and this one is considerably more “real” (in the sense of it being less conceptual and more of an engineering-style layout) that the only other declassified drawings from the same period I have seen (those in the Los Alamos Primer).

Did Harvard’s President sketch an atomic bomb on his notepad? I don’t know, but it’s a very real possibility, is it not? I wonder if any Harvard president since then — much less Harvard’s current President — has ever done such a thing.

  1. See James Hershberg’s James B. Conant: Harvard to Hiroshima and the Making of the Nuclear Age (New York: Knopf, 1993), chapter 3. []
  2. Citation: James B. Conant to Vannevar Bush (21 January 1943, amended 10 March 1943), Bush-Conant File Relating the Development of the Atomic Bomb, 1940-1945, Records of the Office of Scientific Research and Development, RG 227, microfilm publication M1392, National Archives and Records Administration, Washington, D.C., n.d. (ca. 1990), Roll 4, Target 3, Folder 21, “Miscellaneous Bush-Conant Material, May 1941-October 1944.” []
  3. Here’s an attempt by me to decode Conant’s handwriting. Anything I’ve put in italic means “I can’t read this.”

    Memo to V. Bush               Amended by JBC before [???] on March 10, 1945
    From J.B. Conant              Date Jan. 21, 1943

    The latest news from the electromagnetic front via Gen. Groves is (1) Tennessee Eastman are quite confident that process can be made to work. It now seems quite certain that each tank will yield from 50-300 mg per day.
    At  500 tanks that means 50-150 g per day.
    If priorities can be settled there is a chance this output can begin November 1, 1943 (First set Y tanks Aug 1). [Inserted note:  I ??? this now, March 10, 1943; a bomb will require 24 ???; 100 g a day begins ???, 1944. Will take till ??? 1, 1944 for amount! There is still a chance for a military effort in 44.]  This would yield first first [sic??] bomb Feb 1, 1944, at rate of 100 gm per day. This might mean first mean first military result July 1, 1944 allowing four months for developing bomb and manufacturing material for a second. I still believe barring miracles, best day is Sept 1, 1944 . The Chicago method might come along at the same point. So we have two chances of making that schedule. J.B.C.

    That’s not the world’s best transcription attempt (I loathe Conant’s handwriting, I should probably say), but you can get the gist of it, I think. “The Chicago method” refers to plutonium production. “Y” tanks refer to the electromagnetic method used at Y-12 in Oak Ridge. I’m open to any guesses as to better transcription attempts. Conant’s estimate for when they’d have a bomb ready was off by about six months, something I’m sure my German friends are undoubtedly thankful… []

20 Responses to “James Conant’s Atomic Bomb Sketch? (1943)”

  1. Will Thomas says:

    “Cores”? Also maybe “cases” again….

  2. Melanie says:


  3. Ash says:

    I would be very surprised if the “cows” referred to the Cow Puncher committee on which Oppenheimer asked Kitiakowsky etc. to “ride herd”.

    • Hmm, yeah, it’s too early for that. “Cow Puncher committee” was for implosion development in early 1945. But at least we are adding to our cow-themed Manhattan Project references!

      (The only other cow-themed thing I could think of is that cross-sections were measured in “barns,” as in “hit the broad side of a barn,” but since he was talking about cross-sections later it didn’t seem likely. And barns are not cows.)

  4. Moleman, Hans says:

    Cowling? To modify the drag characteristics of the case? As in a faring?

  5. cormac says:

    I think ‘cows’ might be a carelessly scrawled ‘cross’ as in cross-section – a key important parameter in this sort of stuff!

  6. Patrick says:

    “cores” is what I see…dammit.

  7. kme says:

    That diagram looks a lot like the one that ended up in a contemporary magazine – I think Life?

    • Do you mean this one? There is a definite similarity, now that you point it out, but I suspect it is just coincidental — it’s not an uncommon way to depict the gun-type idea.

      • kme says:

        Yes, that’s the one I was thinking of. Do you think that that could be a common depiction because of its early appearance in Life?

        • It’s possible — high profile drawings of nuclear weapon designs are often copied by others, since most of the people who need to draw nuclear weapons for public domain purposes (e.g. illustrating encyclopedias or books or articles) have any direct knowledge of bomb designs. So there is a lot of copying that goes on. (I have been working on an article about this for a long time — tracing the genealogies of these “secrets” was one of the things that got me interested in nuclear secrecy in the first place — but haven’t yet figured out what I ought to do with it.)

  8. james says:

    Maybe he was dislect and this is ‘crews’?

  9. Spruce says:

    I’m pretty sure if that’s ‘cores’. If you try to decipher pen strokes it takes to write that, it’s actually missing one vertical loop from ‘w’ if that were ‘cow’. Besides, it looks like Conant is ‘lazy writer’, who tends to take short cuts when writing fast (I have the same tendency), so leaving the loop from ‘e’ short at that point would be typical of that.

  10. Jonathan Thornburg says:

    Maybe it’s “costs”?

  11. […] transcription is Hershberg’s; the original is no doubt in Conant’s impossible handwriting. Any notes italicized in brackets are Hershberg’s, any not-italicized are […]

  12. Rettaw says:

    If you compare with that thing you’ve transcribed as “cases” it seems like whatever the end is it has the structure of “s-something-s”, probably “costs”

  13. Ben says:

    Speculatively, item 5 says “Further atomic measmt [for] cross section” ie measurements to determine the cross sections more accurately.