Oral History Transcript — Dr. M. King Hubbert
This transcript may not be quoted, reproduced or redistributed in whole or in part by any means except with the written permission of the American Institute of Physics.
This transcript is based on a tape-recorded interview deposited at the Center for History of Physics of the American Institute of Physics. The AIP's interviews have generally been transcribed from tape, edited by the interviewer for clarity, and then further edited by the interviewee. If this interview is important to you, you should consult earlier versions of the transcript or listen to the original tape. For many interviews, the AIP retains substantial files with further information about the interviewee and the interview itself. Please contact us for information about accessing these materials.
Please bear in mind that: 1) This material is a transcript of the spoken word rather than a literary product; 2) An interview must be read with the awareness that different people's memories about an event will often differ, and that memories can change with time for many reasons including subsequent experiences, interactions with others, and one's feelings about an event. Disclaimer: This transcript was scanned from a typescript, introducing occasional spelling errors. The original typescript is available.
See the catalog record for this interview and search for other interviews in our collection
M. King Hubbert; January 13, 1989
ABSTRACT: Born in Texas in 1903; influence of remote, rural environment on his upbringing and early education. Attended Weatherford Junior College until 1923; studies at University of Chicago, B.A. in 1926, M.A. in 1928, and Ph.D. (formally awarded) in 1937. Comments on courses, teachers and fellow students at Chicago, including J. Harlan Bretz and Rollin T. Chamberlin. Summer research at Amerada Petroleum Corporation (Oklahoma), Illinois State Geological Survey, and U.S. Geological Survey (USGS), late 1920s to early 1930s. First teaching position at Columbia University; research on ground-water motion; involvement in Technocracy Movement, 1930s. Marriage to Miriam Graddy Berry, 1938. Senior analyst on staff of Board of Economic Warfare, 1942-1943; deepening commitment to issue of natural resources. Thoughts on limited interactions between geologists and geophysicists; work in advisory committees on geophysics education, 1930s to 1940s. Theory of scale models, 1937; related research involving strength of solids. Career at Shell Oil Company and Shell Development Company, 1943-1964; directs research laboratory at Shell, perspectives on industry environment for scientific research. Lecture tours to geological, industrial, and policy groups, 1940s to 1960s; involvement in Atomic Energy Commission, National Academy of Sciences, National Research Council, advisory committees. Research with W. W. Rubey on overthrust faulting. Deepening interest in oil and natural gas reserves; responses from officials in petroleum corporations and federal government to his predictions of local, national, and worldwide reserves, 1950s to 1960s. Research geophysicist at USGS, 1964-1976, after retirement from Shell; studies of natural resources and conflicts over his conclusions involving other scientists at USGS. Visiting professorships at Stanford University, Johns Hopkins University, University of California, Berkeley, 1962-1977. Continued involvement in issue of geophysical education at American universities and in studies of natural resources, 1950s to 1970s.
Doel:You mentioned you had a few items that you wanted to include that we didn't discuss before?
Hubbert:Yes, we talked the last time about that second paper that I had written.
Hubbert:Well, the thing that I omitted to state was what was the problem involved, what was this about? The background of that was, first, we had this wire model that I talked about.
Hubbert:But Jack Mead who was this colleague of Leith's back in Wisconsin — and I think was the man who built the wire model in the first place — had a paper in about 1918. I don't remember the title. Anyway, what he did was this. In a rigid frame, he had stretched inside the open part of this frame two clamps holding a sheet of rubber, rectangular sheet of rubber. This was under tension. Now, on this sheet of rubber he did various things. But the thing of present greatest interest, he put a layer, a small layer, a millimeter or so, of paraffin. While it was still warm, he covered that over with a sheet of tinfoil, so this thing is stretched, and you had it shortened and the tinfoil folded. There's two things he did then. One of them was this straight rectilinear shortening, and folds at right angle to it. He moved this top brace horizontally, with respect to the other one, parallel, and he distorted this sheet rubber into a parallelogram, just a straight parallelogram. In that case, the folds were diagonal. Again the right angle was to the axis of principal shortening. Now, out of this same confusion, about the rotational versus the nonrotational strain and so on — he said, now, how can we tell, if we look at this operation mountain strain, how can you tell whether they were made by rotational strain or irrotational strain?
Doel:He claimed you couldn't tell.
Hubbert:Essentially. But in studying these fold patterns he found that in his second case, the individual folds tended to be more discontinuous than they were in the first case. In parallel, but they die out and another one would take over.
Hubbert:Now, he said, "That's more like the Appalachian Mountains. "The Appalachians have individual folds, die out, and others take over. So therefore, on the basis of this comparison, it seemed likely that the southern Appalachians were doing a rotational strain instead of irrotational strain. So that really was the basis for my doing those experiments and that paper. What I did was that I had this block with the modeling plaster in this. I cut the blocks diagonally, so you could press like so. Then we got those strain ellipses. They were 45 degrees and so on, less, to this axis. Now, that pleistocene was supposed to be analogous to, say, a body of sediments between two ranging buttresses like the southern Appalachians. What that really led up to was that if you had this shear motion between the two sides, the folds would be on echelon crosses about, instead of parallel to it. The fact that they were parallel indicated that that shear motion didn't take place. Therefore your principal stresses, your principal forces, were perpendicular to the axis. That was the object of that paper.
Doel:Did Rollin Chamberlin help at all in designing the experimental apparatus that you used or give you any advice?
Hubbert:He didn't even know I was doing it.
Hubbert:He had some grapevine knowledge, from this loudmouthed individual who was telling him that I was doing something or other. That got him kind of stirred up. In fact, he called me in and gave me hell one day on the basis of this rumor and misinformation. Why hadn't I told him about it? I said, "Well, I wasn't through, I wasn't finished. I was just trying to find out something. And I didn't have anything to discuss." Then I gave this paper, finally, before the Geological Society, and he was present. And I told you, he was almost apologetic to me.
Hubbert:He said could he help me! He hadn't been clear just what he meant — I said, "Well, it was a simple experiment, it's done, and it doesn't mean anything else." But then he printed the paper.
Doel:Yes. That was of course the second published paper that appeared in print.
Hubbert:That was the second, yes. Well, it was gracious, in a way, because he had a sore toe in this performance also because he was up to his ears in this thing, that I was really challenging him and was a little mean. And so after it was all over with, I didn't ask him to print the paper, but he volunteered to do it.
Doel:Were there many other students at Chicago who also used apparatus of that kind, who built it?
Hubbert:Well, there was one other. This one person did a whole series for his doctor's degree, a whole series of model experiments, but he had no theoretical understanding of what he was doing. It was purely empirical.
Doel:Who was that?
Hubbert:A man by the name of Theodore Link.
Hubbert:Theodore Link was a very competent oil geologist there. In fact, he'd been an oil geologist before and then he'd come back to school. But he'd done his laboratory experiments which were purely empirical, had no theoretical basis whatever, and he did his doctor's thesis on it. While I did this thing that I did for a very real theoretical reason, and for a very real problem in the literature. I didn't make that clear, just what the objects of these experiments were.
Doel:It's good to have that clarified and on tape.
Hubbert:Yes. I re-read the paper just before you came here, just to refresh my memory on it.
Doel:Did you have any first-hand experience with the Appalachians at that time, or had you gathered your information through the geological literature?
Hubbert:I had never seen the Appalachians, or, if I did see them… no, I'd never seen the Appalachians at that time. It was the literature.
Doel:You raised an interesting point in talking also about people going into the petroleum industry after Chicago. Where did the majority of graduate students find employment at that time?
Hubbert:Well, some went into teaching. Some worked for the Geological Surveys. A few of them went to the USGS. Some were in the Illinois Geological Survey. Others were in various schools teaching.
Doel:Right. Do you have a sense of what proportion?
Hubbert:I would say that probably the biggest single item — certainly one of the major items was going into oil geology. It was a good job, good pay and so on at the time and jobs were fairly plentiful. So quite a few of the students got jobs with the oil companies.
Doel:These are the PhD students?
Hubbert:Well, some of them PhDs, some of them bachelor's degrees or various states of graduate work.
Doel:That's another interesting point. Did most of the graduate students go on for their PhDs, or did many stop before then?
Hubbert:Well, there's a pretty clean break between undergraduate and graduate work. The undergraduates were local students. And most of them took geology just as a part of their general education. A few of those went on to graduate work in geology. But most of them did not. Most of them, it was just part of their education. Now, the graduate school, most of the students were there from the outside. Came in from other schools, or in some cases they'd been out working full time and came back to school, to do their graduate work.
Doel:You mentioned in the last interview that your first acquaintance with Columbia was when someone had come out and heard you lecturing in geophysics.
Hubbert:Yes, a man came through and I was giving this course I told you about. That was in the spring of 1930. The winter quarter of 1930. And this man was the head of the School of Mines at Columbia. He dropped by, and reported to the chairman of the department that he was looking around for somebody to teach geophysics.
Doel:What was his name?
Hubbert:I think it was Rade, T.T. Rade. And he, I think earlier than that, I think he had been the secretary of the AIME.
Doel:OK. How much expansion did he have in mind for geophysics? Was he only looking for one instructor, or was this part of a larger effort Columbia planned?
Hubbert:No. Just one. See, nobody in the academic world at that time, or very few, had any idea of what geophysics was about. What had happened was, the mining people had been doing magnetic mapping for 50 years or so, using magnetometers to find iron ores and that kind of thing. But the oil industry, by 1923, had brought in the first seismic crew — a German crew with mechanical instruments — and they were trying to find salt domes. These salt domes were more or less vertical cylinders a mile or more in diameter. Then the oil companies found some large oil fields and they turned up sediments around these plugs. So then they were certainly major targets of exploration on the Gulf Coast, and the clues in surface mapping were very very obscure. In some cases, yes, you had actual surface mounds.
Doel:You could see part of this structure?
Hubbert:Yes, these mounds were up above the level of the plain, in some cases as much as a hundred feet. But that's exceptional. There were many other deep salt domes that had no surface expression whatever. So the big problem in the oil industry was some means of finding these salt domes. So they brought in this German seismic crew, some company or other. Then the various companies began to develop their own seismic equipment. Their first seismic work was refraction, where you would shoot a shot here and record over here and for that distance the time ought to be so-and-so. But if there was a salt dome in between, with a higher velocity, then you got an anomaly in your travel time.
Doel:Right, then you could drag and locate it.
Hubbert:Shooting around a bunch of shots. You kind of shot a fan, or three-quarters of a fan, that delineated the salt dome. And the rivalry was very very keen for this. The oil companies had scouts, and the scout followed these crews around and reported back to his office what they were doing. Of course, if he could see them begin to do something unusual like fanning and so on, why, boy, that was hot, information that went immediately to his office. They sometimes rushed in and leased the land before the people who did the work could. It was a kind of a game. In fact, the scout was usually a very affable person, very helpful. The crew almost felt like the scout was one of them.
Doel:One of them?
Hubbert:The hired scout.
Hubbert:But they had tricks like this: here was this crew working. In case they've found something, and they don't want the scout to know about it — they want to shoot a fan on this — they all go out and spread out for their day's work, and the scout can only follow one of them. And whichever one the scout follows, that one will just keep going. He pulls the scout off on a wild goose chase, and then they do their fan shooting while the scout is out of sight.
Doel:That's interesting. Did those techniques spread very quickly through all the major oil companies?
Doel:That of course was the time when the estimates of expected reserves were fairly low, during the 1920s.
Hubbert:Oh yes. They had a surplus of oil, but there was a little bit of shortage in the twenties. Then they hit these Texas wells, and they just flooded themselves. And that was 1930.
Hubbert:But anyhow, with this refraction seismograph, they found dozens of salt domes by this means. One little story I can't resist bringing in here is irrelevant, except as an associated study. Did you ever know the name Ernst Cloos of Hopkins?
Hubbert:Well, Ernst Cloos was a student on this German crew.
Doel:Was he? That's interesting.
Hubbert:He came to Texas, and here they were in the broiling hot sun, unloading all this heavy equipment in big wooden boxes, from the ship onto the dock. The sun broiling down on these light skinned Germans, and so on. They were standing around and conversing with each other on the side, and one of the Negro dock workers came up and started a conversation with them in German.
Hubbert:They were astonished. "Where in the world did you learn German?" He said, "I am German." "What? You're a German?" He said, "When you've been here as long as I have, you'll look like this too." Texas had a large German colony back in the 1840s. There are all these German Texans now. And at that time, they were still speaking German. This Negro had grown up in a German community, where German was spoken.
Doel:I suspect he didn't convince them. He might have made them nervous. Did you come to know Ernst Cloos fairly well?
Hubbert:Yes, I knew him fairly well. I also knew his brother.
Doel:Yes, Hans Cloos. Right.
Doel:Yes. You had met Hans Cloos perhaps the first time in 1933?
Hubbert:I met him in Washington at the International Geological Congress, in 1933. He had an exhibit, oh, a big table about eight feet square or so, with these little models all over it. And I was looking at those. I'd seen the experiments. I knew the publishers and results for — what Chamberlin did — and things like that, when he was a young man.
Doel:On the strength of the materials?
Hubbert:Well, they were geological models. Willis had done the thing for the USGS way back in the 1890s. Modeling folding, like folded up Appalachian Mountains.
Hubbert:And all these things were done empirically, without any theoretical guidance, they would just put them together, plastic materials, some harder, some softer, and I was familiar with them. I had seen Chamberlin doing these things, and I knew of the published results of Willis' and another's papers. They weren't very good. So I looked at these little models. By golly, it looked like a piece of geology reduced right down to that size. It had small details and slips and fractures and all. Well, the stuff would harden at the time, I mean it was solid. I was looking at this thing, and Cloos came up and we got into a conversation. I marveled at the fidelity of it, the similarity of these things to real geological situations. "What material did you use?" He said, "Almost liquid clay."
Doel:That was a surprise to you at the time, wasn't it?
Hubbert:And so I told him, "By God, that's even worse from reality than these other things which we use." And then that night in my hotel room, I did a little bit of theoretical work on this, and I came to the conclusion that what Cloos had used had almost exactly the right properties.
Doel:Because of scaling.
Hubbert:Yes. Well, it boiled down to this, to take the very simplest case. The mass of a body proportional to its volume, and the volume is proportional to the length. Put this on a scale, so that if you have an original length 1 and you cut it down to L 2, the volume will be reduced by a ratio of L 1 over L 2 cubed. And over L 1 cubed. And what about the strength? Well, what about the forces? The forces will be proportional to the weight, and then the stress will be the weight by area. Well, the weight will be proportional to the volume, which will be L cubed or the Lambda ratio, Lambda cubed, and the force will be proportional to that also. But the stress is force per unit area and the area is proportional to L squared, and so you come out with the stresses proportional to Lambda. So that if you reduce an original say a hundred fold, you'd have to reduce its strength, specific strength of the materials by a hundredfold, if the density was the same. Well, if you took a block of rock, say, a thousand kilometers and you reduce it down to a model of a tabletop of say a meter — well, that would be a thousand kilometers, that would be a millionfold reduction. You'd have to reduce the strength of that material by a million. Well, that was what Cloos had done. Cloos had had a very simple model theory. It went something like this. If the original material could support a standing column so high, then if you had this reduced size, it could only support a column 1 Lambda of that height. And so for constant density, why it was a valid model theory. And I convinced myself for that, in my hotel room that night. That was a year or two before I ever got around to the idea of writing this. In fact, the idea of writing this paper was by direct assignment, you might say, from this interdisciplinary committee.
Doel:I was going to ask you about that later, but why don't we talk about that now if you like. You were involved in the interdisciplinary committee between physics and —
Hubbert:Well, the National Research Council.
Hubbert:They called it a borderland committee between geology, physics and chemistry.
Doel:Right. How did you get appointed to that committee?
Hubbert:Well, actually, the committee was set up in 1935 or so and I didn't even know of its existence. But some geological member of the committee had dropped out for some reason or other, and so I was invited to come in and fill out the second year.
Doel:Right, so this is 1936 you're coming on?
Hubbert:1936. When I got there and reviewed, and had the record of what they had discussed and considered the year before, I was struck by the fact that one of the things they had on the agenda was the theory of scale models.
Doel:Someone else had proposed this initially?
Hubbert:It had been proposed the year before by somebody. Well, I was intrigued by that, because I had a whole background interest in that subject. Finally when the committee wound up its affairs, which they did in the spring of, the end of the academic year that year, they had recommendations of things that ought to be done. One of them was that we needed a kind of a handbook of rock properties, mechanical and other properties of rocks. And so Francis Birch said, "OK, I'll tackle that one." And the scale model thing, I said, "Well, I'll take that." I don't remember what were the other things that the committee members were doing, but anyhow we took those two. And so then I settled down and wrote this paper. So the background on that is that it goes clear back to when I was a kid about 10 or 12 years old. I got into these curious things like a mouse could jump off the table to the floor and not break his legs. And a windmill. I made at that time a little toy windmill, made the rotors out of wood, and mounted them on a shaft and had this horizontal thing pivoted with a tail out here and this rotor on the other end. So here was a big farm windmill here. The rotor would go around like this, [Moving his fingers slowly] but the model would run like that. How come this thing turns faster than that one does? Why?
Doel:This again was something you were thinking about mostly on your own?
Hubbert:Of course. And there was another thing, when we built these dirt tanks that I told you about with my father. He did a lot of that; I was also a helper on part of that. You could build these dirt tanks and they would hold water perennially, but if I built a little dirt tank and poured a bucket of water in it, it would dry up in a matter of hours. And so how come? Your big one doesn't, your little one does. So things of that sort had been going around when I was young. In this big whole-year course in general geology, the advanced course raised questions of this sort. I told you about the question of how large a boulder could a stream move and so on. And all the way from sand grains up to great big boulders. Well, then, later on, when we got around to considering the dinosaurs, Bretz raised this question again. How big a land animal could you have and still be able to walk? And the evidence was that the dinosaurs had just about done it. Well, all this came together. About the same time, there was a volume that showed up in the geology library. It was a tome on hydraulic models, hydraulic laboratories, translated into the English from a German volume, with a description of a half a dozen German universities or German laboratories, hydraulic laboratories, in different institutions in Germany.
Doel:Do you remember the title or the author's name of that?
Hubbert:Well, I can book it up. It had various scale models in there.
Doel:OK, fine, we'll check on it.
Hubbert:Then what had happened was that a man by the name of John R. Freeman who was a very, very prominent man in the mechanical engineering circles—an impressive person, and he'd been over to the German universities and laboratories, and we had nothing like them in this country. He came back and promoted the translation of this volume into English, and published by the AIME. Not only that, but he had a couple of men on his staff and who said he was associated with some insurance company up in Hartford.
Hubbert:Yet he was a respected mechanical engineer, and he'd done it by himself. He was practically a self-educated man.
Doel:That's interesting. Of course Thomas Malone in later years at the AGU was also at an insurance company in Hartford.
Hubbert:All right, Thomas Malone. I'm not sure that he didn't come up under Freeman.
Hubbert:I think he did. Anyhow, Freeman had two of his men write papers on the various scale models as an appendix to this volume. Well, when I ran into that, which was two or three years before I had any idea to do this. In fact, this thing that I wrote was entirely unscheduled, it was just out of the blue. But I was intrigued by the fact that somebody else was thinking about these things. Not only that, they had theoretical structures. Well, I had some too. I'd done this kind of thinking, as to, no matter what the density of the body, the volume would go up as long as the two bodies were too much of a similar, the volume would go up as the cube of the ratio, up or down, depending which way you go. And also, the forces and stresses. So when this problem emerged, why, I could practically write that paper with minimum preparation after working on this for 20 years or so.
Doel:Right. This is when the committee report was assigned to you?
Hubbert:That's how I got into it, but I had all this background.
Doel:Do you recall talking with anybody in particular when you were preparing that paper and writing it? Percy Bridgman at Harvard, for instance?
Hubbert:No. I was pretty much on my own.
Hubbert:At Columbia, I didn't have very much intellectual companionship.
Doel:Right. We're going to talk about Columbia in a moment. But one other question that you just indirectly raised. Why was geophysics so strong in Germany in comparison to the United States? And was it Germany in particular among European countries?
Hubbert:Well, oh yes. The whole oil industry and the mining industry had magnetometers. The oil industry brought in these seismographs, and also they brought in the torsion balance. Now, do you know what a torsion balance is? It's a very sophisticated, very simple but very sophisticated instrument. They found they could find oil by these torsion balances. Then that was followed up by gravity measures about ten years later. That directly measured gravity instead of the gradients and so on.
Doel:Were those instruments also developed in Germany first?
Hubbert:The torsion balance was developed in Hungary by Baron von Edvers [Unclear]
Hubbert:Not only the instrument but the theory. A very elaborate and very sophisticated theory. He did the whole works. Developed the theory and built the instruments and actually used them, not necessarily oil exploration but for some geological work. Martin was with his cracking company in Amarillo, had gone over to Germany, and bought quite a few of these torsion balances and also some magnetometers and brought them back and set them up with field crews doing magnetic and gravitational work. That was in the middle or late twenties. Well, what I'm saying is that the geology departments, Chicago and the rest of them, were almost oblivious to any general notion of geophysics, meaning the physics of the earth. Seismology yes, but that was an individual situation, but not a comprehensive view.
Doel:Right. Even seismology tended to be at the government facilities or the research laboratories of the universities.
Hubbert:Yes. But even so, there were tomes, again largely in German, on seismology, but very little in English. Some, but not very much. And here the geologists were talking about these things, but they really had a very superficial knowledge or understanding of it. So the average outlook as to what was geophysics, as seen by a geology department, was these various gadgets for finding oil. Oil geology, that was the equivalent, magnetometers and seismographic instruments.
Doel:That's interesting. You're saying that it was viewed as an applied area of research?
Hubbert:Absolutely, that was all, that was synonymous. The word geophysics was synonymous with these techniques and operations in the minds of the geologists. Well, I was rebelling at Chicago against this whole damned thing, because here we had in the library these tomes on seismology and what not, in German, and nothing comparable to it in English or very little, nothing taught at the university, nothing whatever.
Doel:Right. I suspect I know the answer, but did anybody else in the department read those volumes and talk with you about them?
Hubbert:No. What I was rebelling against was that geophysics is not these instruments, it is not gadgets for finding oil, it's the physics of the earth. Whatever's running down hill is geophysics, mountain-making is geophysics, so geology was dealing with these things all the time, but without the physics. They were dealing with them as the geometry and the kinematics. They talked about physics, but usually their notions of stresses and things of that sort were mostly erroneous. One thing: in this matter of folding of things like the Appalachians, envisioned was sort of an amorphic push from the rear, and that was called an active force? So you wound up in those pictures with a one sided force that you introduce equilibrium into the situation. The other forces weren't recognized. The opposing ones. The fact that the body is not accelerating. And yet these other forces, their presence wasn't even recognized. The other thing was, and I go into that in this second paper, the tendency or the attempts to resolve stress problems by means of force resolution. It was very common to try to do a parallelogram force resolution on stresses. Well, it can't be done. You can resolve forces, but the stresses require the force over area, which is naturally at a different rate.
Hubbert:So, when these things were mentioned in textbooks, they were usually erroneous. It was usually a naive attempt to do something they didn't understand and physically erroneous. This was what I was stewing and fuming about while I was a graduate student at the University of Chicago, which led me to have the audacity to propose giving a course in geophysics myself. In that case, we were dealing with the whole earth and the earth deformation, with gravity. It's a question of the supportive mountains the post-glacial rise after you let go of the glacier, and all that kind of thing. So this course of mine was by no means complete. I put together what I had my hands on. But a considerable part of it was gravity, gravity and the isostasy problem, the Coast and Geodetic Survey problem. Torsion balance was brought in as sophisticated instrument, but not for dealing with potential gravitational measurements.
Doel:Did you offer that as only an undergraduate class, or was it also for graduate students?
Hubbert:This course was graduate.
Doel:How did you teach that? Did you have any written material that you felt comfortable enough to use?
Hubbert:Well, after I decided to do this, I spent all my time working on notes for this. There's a very interesting little story right there. I spent the summer of 1929 in my office in the geology building working on my notes on this course. And so everybody else in the department had all gone out on field work and one thing and another. I was the only staff member still there, except a paleobotanist.
Doel:So you had Rosenwald Hall to yourself.
Hubbert:All right. I also had various jobs on the side. They had a seismograph, and one of my jobs was tending this seismograph and changing the records, which is a daily chore. But it was also operated in conjunction with the Weather Bureau, and the central station of the Weather Bureau was also in Rosenwald Hall, so these two were related. They were under the same jurisdiction, approximately. I also was reading meters for the Weather Bureau and changing records for the seismograph and flying weather balloons off the physics building, and observing them.
Doel:That's interesting. They were launched directly from the campus.
Hubbert:The weather balloons?
Hubbert:Yes. And so those were some of my bread and butter chores. Well, I was at my desk and I had a German book Lehrbuch der Geophysik, edited by Beno Gutenberg. I was working on that. I had it spread out on my table. And there was a knock on my door, and the head of the weather station was outside. I opened the door, and he introduced me to a guest that had come by. It was Beno Gutenberg [Laughter].
Hubbert:And it was about 11 o'clock or so. I was a member of the Faculty Club — they called it Quadrangle Club there — and so I fortunately was able to take him to lunch. His uncle was with him. He had an uncle who lived in Chicago. He and his uncle were there and I took them to lunch at the Faculty Club. I may have taken von Neu [unclear] the Austrian paleobotanist. I think he may have joined us too. Well, that was the beginning of a lifelong friendship between me and Gutenberg. He was en route, this was his first trip to the United States.
Doel:Was he going directly to Caltech?
Hubbert:He was en route for his first visit to Caltech. Out of that visit, in fact, they pretty well agreed that he was the man they wanted, and they went ahead and hired him. So within a year, why, he was permanently at Caltech. But, while I didn't see him very often after that, we had occasional correspondence. I remember he sent me a weekly calendar of Caltech, and my papers on scale models were being reviewed.
Doel:That's fascinating. Do you remember what you talked with him about in that first meeting, any of the details?
Hubbert:Well, it was really a social occasion, affair. There was no opportunity to discuss technical questions.
Doel:He didn't talk for example about what he wanted to do at Caltech?
Hubbert:No. No. It was all very brief. Incidentally, I was the first scientist he'd met in the United States on that trip.
Hubbert:Everybody else, where he'd gone there was nobody home [Laughter].
Doel:Well, that's true of summertime in Chicago.
Hubbert:Well, I might back up a little farther on this. The Geological Society of America used to meet during the week between Christmas and New Year's. They traveled by train and that would allow people to come in from long distances to get to the meeting, and get back and not be losing time from the university. Most of the geodetic people were at universities and geological surveys at the time. So the meeting, probably 1928-'29 — it must have been, let's see, '29, I think so, yes. Well, again, a group of two or three people, two at least, dropped by Chicago University from Caltech. One was John Buwalda who was the head of the department. The other was Chester Stock, who was a paleontologist.
Doel:Those were the two main people in the department at that time.
Hubbert:OK. Well, this was the time when I was steamed up over this deficiency of the geological curriculum, as there was nothing on the physical aspects of the earth. So I had a session particularly with Stock, who was a very nice guy, and I was raising questions. I was interested in the possibility of coming out to Caltech. What were they doing in geophysics besides seismology and so on? Well, Stock told me at the time that they were considering an outstanding world figure to come to Caltech. That was probably — I'm sure it was at Christmas before — before Gutenberg came.
Doel:He didn't mention Gutenberg's name?
Hubbert:No. He didn't say who. And Gutenberg was coming out.
Doel:He was planning to edit it. OK. We're looking now at a copy of the HANDBUCH DER GEOPHYSIK. This edition here is published in 1931.
Hubbert:Well, it was coming out piecemeal, and there were plenty of volumes of it. But he got chased; he was promptly shut down by Hitler. Well, he really left before Hitler because he left there in 1929.
Hubbert:Or 1930. But things were getting very difficult.
Doel:This is what he had published in Germany prior to his leaving for the United States?
Hubbert:Yes, but he never got beyond these fragments. I don't know whether…is this an outline of the whole single volumes? I think it was supposed to be about a ten volume work.
Doel:This covers the chapters and the sections. Yes, this is just for this volume here.
Hubbert:Yes. Well, his plans were, I think for about a ten volume work, HANDBUCH DER GEOPHYSIK. It was coming out piecemeal in fragments. And this is one of the fragments.
Doel:What happened to the series after he left Germany?
Doel:That was it? No one else took it over?
Hubbert:Hitler came in.
Doel:Right. What university was he at in Germany?
Hubbert:University of — I don't know, does it say there?
Doel:It doesn't say. It was published in Berlin.
Hubbert:Let me see that.
Doel:We can check on that later. Aha, I see: Frankfurt. Did he ever talk to you later about the kind of research group that he had in Frankfurt?
Hubbert:Not very much. Actually I saw very little of him, although we kept in touch with each other by long distance. But I actually saw very little of him. I visited him whenever I was out at Caltech, had dinner with him and his wife at his house. But what I'm getting at is that my ideal of what geophysics is and ought to be in the university was this kind of thing. And the university's idea of it was extremely narrow, about torsion balance and magnetometers.
Doel:What impression did you get from talking with Stock and Buwalda? What were their views of geophysics?
Hubbert:Well, I had a very good conversation with Stock. I was pretty steamed up over the whole situation and I expressed it very forcefully. I'm not sure that I ever saw Stock again until the GSA met in Washington, the meeting after World War II, sometime in the forties. And Chester Stock rounded me up, and we had a considerable session. And he remembered this conversation that we'd had very vividly, and they were looking for a chairman of the department and were considering me.
Hubbert:All right, I was very interested. I agreed that I'd send him a bunch of credentials and papers and what not. So I got back and I did.
Doel:This was at a time when John Buwalda was still alive, or had he already passed on?
Hubbert:John Buwalda I think was gone by now, dead maybe. Chester Stock was acting chairman of the department, and so we had this pretty heavy session at the GSA meeting in the hotel, whether I'd do, whether I might take over as chairman of the department.
Doel:It was just the two of you, not Gutenberg?
Hubbert:No. All right. I would send him credentials, reprints, and biographical material. I was very interested in this possibility, so I did send the stuff. I waited for weeks and never got a reply. Finally I got a letter from my friend, an associate professor in geophysics at Caltech, and he said what an awful turmoil they'd been in since Chester Stock had died. It seems that Chester Stock died right after Christmas.
Doel:Right, yes. In fact, that was a greatest shock to the Caltech Division of Geology, because both had died within a fairly short time of one another.
Hubbert:A lot of these things tie together.
Doel:Do you remember what you talked about with Stock over plans for the department, ideas that you had in mind to do there?
Hubbert:No. But it undoubtedly would have been the kinds of things that I've been talking about for years.
Doel:At that time you were already with Shell, so it wouldn't have involved —
Hubbert:I was with Shell at the time.
Doel:That's a fascinating bit of information.
Hubbert:See, I'd first met him and I don't know that I'd seen him in between, since when he came back to Chicago in the winter of 1928. But he remembered me!
Doel:You must have made quite an impression on him.
Hubbert:Turning to another thing, somewhat similar. I've lectured various times at MIT. Different subjects.
Doel:This was after the Second World War?
Hubbert:Yes. So here about three years ago, I'd been through some of the troubles [asthma] I'm in now, and I didn't try to attend the annual meetings of the Academy too much. But I went down one day for a special program. What I listened to was mostly a series of committee reports on various things. One was given by a man I'd never heard of before, whose name was Howard Johnson. He was said to be the chairman of the board of—I don't know what they call it at MIT, it isn't trustees. It's an analogue of trustees but they don't call it that.
Hubbert:The Corporation, maybe that's it. Anyhow, he was chairman of it, and I'd never heard of him before, and I listened to his report. It had to do with some kind of committee dealing with the exchange of information internationally and so on. And then when this meeting broke up, I went out of the auditorium into the big lobby, the rotunda area, and sat down in a chair to rest a while. I was right near the archway which was the exit, and was sitting there when this man Howard Johnson came dashing out, obviously heading for his plane. He came along and I was sitting there, and he got about opposite me, put on the brakes, stopped and spoke to me. He said, "I heard you lecture at MIT and we haven't forgotten it yet!"
Doel:That's interesting. So he'd attended those lectures?
Hubbert:I didn't know what lecture he heard, so I wrote him a letter after that to inquire. I'd been there two or three times and I didn't know which one. It was the one dealing with mineral resources and energy resources and their social implications and the evolution of the world system that's tied up with these things. He'd specially gone to hear that, sat in on these lectures. The attendance was very small.
Doel:Do you remember approximately when that was?
Hubbert:1959 or so.
Doel:OK. But there was limited attendance? There weren't that many people there?
Hubbert:Well, the whole thing was in the earth sciences. It wasn't an all-university affair, all-institute affair, and so in that environment actually the attendance was a small handful. But this man, apparently he got wind of it and came over. I think he had been with the Sloane school or something, and he came over to hear it, so he told me.
Doel:It's good that you heard that.
Hubbert:It's interesting. He's coming out of here and I'd never heard of him before, and he stops [Laughter].
Doel:When you had talked with officials at Columbia, did either you or Columbia set any kind of conditions on what you would do? Were there any expectations for example that you would have your dissertation done?
Hubbert:Well, in a general sort of way, not specifically, it was to teach a course in geophysics. They also wanted me to take over structural geology. So I really had two assignments.
Doel:What had happened with structural geology? Had Lobeck been teaching that at a later time?
Hubbert:Yes, but the trouble with structural geology was just that. Everybody knew that it had to do with description. It had no physics to it. So, I simply elaborated the course I'd given at Chicago and extended it. But the difficulty was that the students. One in ten had never had a course in calculus and most of them had never had college physics. My God, what can you do?
Doel:Did they have much chance, when they were in the geology department, to take physics?
Hubbert:The geology department, as I learned, was a watertight little island of its own. They had about 12 professors give all different courses. They had about 50 graduate students. Only one person handled the undergraduates. It may have been Lobeck. Most of the rest of them were giving graduate courses. So imagine the situation. You've got 12 men giving courses and you've got about 50 students. In order to have students, every course has to be required.
Doel:That fills up two years.
Hubbert:And the students have no time for anything else. They're just stuffed and crammed with the required geology courses. So that the kind of thing that they needed, physics, mathematics, chemistry, all that kind of thing as background, they didn't have. Even if they wanted it they couldn't do anything about it because the administration would have been distinctly unfavorable.
Doel:You mean the university administration?
Hubbert:The department of geology. The department of geology was a little tight enclave, self-contained, and their only connection administratively with the outside university was with the School of Mines and also civil engineers. They had a course, one year course, for students that were civil and mining engineers, who came over to geology for this course. But there was practically no liaison with anybody else except mines and civil. So again, you had a lot of poorly trained students. In the ten years that I was there, I can't remember more than three or four students who'd ever had calculus, and most of them had never had college physics.
Doel:Was there interest on their part to try to learn physics?
Hubbert:Well, under their handicap, it would have been very difficult because they were having to take all these required courses in the department. Even if they had wanted to, they'd have had no encouragement. In fact, they would have been decidedly discouraged. In my judgment, the most scientifically oriented member of the faculty was Douglas Johnson. He handled — they called it geomorphology, but it was a contemporary, there as no processes or anything like that. So here he was giving these courses and seminars — they were elaborate and very able ones — on erosional phenomena of geology. They had a very good hydraulics laboratory and he never once sent a student over to it.
Hubbert:It's that kind of thing. Never once did any student ever go over to take this hydrology course. And yet you were dealing with those phenomena.
Doel:And there wasn't much personal contact either between the geology faculty and people in the physics department and in the chemistry department?
Hubbert:None. Practically none.
Doel:How much contact did you have with other departments?
Hubbert:Very little with physics. I didn't know anybody over there. My principal contacts were with people in engineering. And chemistry. Kind of late in the time, about 1935, the engineering school, chemical engineering, decided to put on a year-long course on theoretical chemistry for engineers. They had a professor who was the principal lecturer on that who was a very able man by the name of Jeddy, I think he was. I guess he was a chemical engineer but he may have been in the chemistry department. Then they had a chemical seminar given by Victor Lemaire. Well, I signed up as a visitor for this entire year's course. It was an advanced course in very fundamental chemistry, fundamental physical chemistry, and then chemical hydrodynamics. Well, I became very good friends with this fellow Lemaire. Some years later, in the Geological Society of America meeting — it must have been about 1945 or '46 I attended to a meeting, I think it was in Pittsburgh — and a couple of graduate students came up and introduced themselves to me. They were from Columbia University. It seems that they had decided that they needed chemical thermodynamics. They'd gone over to see Professor Lemaire who was in charge of the chemistry department. He looked them up and down and said, "Who are you folks? I've never seen you before. You're not chemists." "Well, we're geologists." "You know, I had a geologist in this course once. I think I'll let you in."[Laughter]
Doel:He meant you, of course.
Hubbert:Yes. One of those guys is a prominent professor at Yale now in the geology department.
Doel:Oh really? Who is this?
Hubbert:I can't remember his name.
Doel:We can put that on the tape later.
Hubbert:Yes. He is now professor in the geology department at Yale.
Doel:Did you know any of the other members of the chemistry department? Harold Urey was there at the time.
Hubbert:I don't recall meeting him. I met him subsequently, but at that time, I don't think I knew him. And I don't know how much of the time he was there because he moved hither and yon. He went to Chicago some time after that.
Doel:That's right. He left after the war to go to Chicago.
Hubbert:But I did know him at Chicago.
Doel:Yes, OK. How did it come about that you taught the class for the mining engineers?
Hubbert:Well, that was something that had been going on, this regular thing in the geology department. It was taught by a professor who left on rather short notice and took another job. During most of that decade I was working summers on the Illinois Geological Survey. So I'd come back from my field work in Illinois, and was told that the man who had been giving this course for civil and mining engineers had left, and you're taking over tomorrow morning!
Doel:Not much notice.
Hubbert:So I waded into this thing. In the first place, the students were strange. I knew the geology students but I didn't know these guys. Another interesting part of it was that they had had engineering. The Columbia Engineering School had two years of pre-engineering before the students were admitted to the Engineering School, and four years of engineering on top of that.
Doel:It was a six year program?
Hubbert:Yes. And generally these two years were fundamental mathematics, chemistry, physics and so on, as well as whatever else they may have had. Then they came to the Engineering School. These students that I had were freshman and sophomore engineers, which meant they were in their third and fourth year in college with this background. Well, I never had any students anywhere with that degree of technical training. And I was of course totally unprepared to give this course. So I took the position that geology is geology no matter what you do with it later, and so we will follow the general outline of general geology. But we will go into more detail where appropriate and where we can, appropriate to the background these guys have already. So it was kind of tentative, feeling your own way, what can you do with these fellows? Well, it worked out very nicely. I remember in that very first year, we got around finally to the part on geological phenomena. So I discussed the thing that I'd had buzzing around in the back of my head since 1931, when I began taking an interest in these things in my work in Illinois.
Doel:You were studying earth's electrical resistivity then.
Right. So I raised the question of what is the basic physical equation analogous to say electricity, with the flow of ground water? For the current obviously it could be either volume rate of flow or mass rate of flow. But what about forces? What about potential function, if there is one? I didn't know the answer to that. As I say, it was a kind of an idle curiosity type thing. But here I had a bunch of engineers, and I raised this question as to what goes on here? Is there a potential function, and if so what is it? So we did the usual thing, we made a guess. We guessed it was pressure. Well, it's easy to show that that was impossible. Look at the vertical loop between these two wells, pumping in one and out of the other. In a very slow rate of flow, you disturbed the hydrostatic state, well, approaching zero. But the flow pattern — as long as it flows at all — is invariant with respect to the rate. You know that in hydrostatics the equal pressure surfaces are horizontal to the pressure increases downward, so in this downward loop you plainly have pressure flowing from lower to higher pressure, the higher pressure coming back up. So it couldn't be pressure. Well, I did a little theoretical work on this in my office — maybe over the weekend, I don't remember. I went into this kind of reasoning about it. In electricity, the electrical potential is an energy-producing charge. Now the amount of work you do is to move the energy charge from a fixed defined state to this point. When it flows, you're flowing from higher to lower energy levels and you're discharging the energy, the potential energy and the heat, thermally.
Well, in the water case, you certainly are flowing. It must be from a higher to a lower energy level. It's a viscuous flow, so it is thermodynamically irreversible, and it is discharging the energy and heat continuously down the flow line, no matter what the direction is. Well, can we evaluate this energy per unit of mass in a fixed point? So in effect, I said, all right: we have a bucket of water over here. That's our reference state, along with the atmospheric pressure and the elevation of the surface of this water. So I take a unit mass of the water out of here, I lift it to the height that I suppose I'm dealing with a laboratory situation, so I lift it to the height of the point I'm interested in, and I do a gravitational work of GZ per unit mass. And I move it over here and I inject it into this point, from atmospheric pressure against the pressure of the point, and that's the PV work. The kinetic energy is negligible so that's the whole works. You see gravitational energy on the left, [draws diagram] and you see injection from the standard pressure to the pressure of the point. Now, that is a measure of the energy, the point. Does it have the desired properties? Well, another thing we can do with this is to tamp a manometer into this system at that point, and and let the water rise to a static equilibrium. When we do that, we have the pressure peak at the bottom of the manometer, and we have a free surface up here at a height H, see, above our reference level.
Now, the other route we can take to get in here, we can take our unit mass out of here, lift it up to the top of the manometer, and lower it down to that level and so work is GH. Then, I bring it down in here, and the force is zero. So I can bring it the rest of the way down for zero work. The total amount of work is GH. Now, is that GH the same as GSE? Of course, we don't know, but it turns out that it is. OK, so our manometer is in, an instrument that measures this energy per unit mass. H is the height and G is the factional portionality and GH is the energy per unit mass. All right. Now, let's go back to the classical experiment of the Frenchman D'Arcy, who flowed water through a sand.
Doel:Had you read that already before you began teaching that class?
Hubbert:I don't believe I had, but I got a lot out of the work.
Hubbert:It was called D'Arcy's Law. But all right, here you've got two manometers. If the water is static, the manometer stands at the same height. Therefore the energy at the two [unclear] is the same. But if you start the water flowing, say, downward, then the upper manometer stands higher than the lower one, and it's a measure of the energy above. The other one is a measure of the energy below. So out of this, if you start with the idea that the flow is proportional to the negative grading of the potential, the energy potential of mass, it turns out also that this is proportional to the manometer reading. And so then this satisfies D'Arcy's Law. Well, D'Arcy's original experiment was performed by downward flow. It would still be true if the flow is at some other angle than downward. Say, an arbitrary angle, thet to the horizontal. Or up, or vertical.
Doel:Right. D'Arcy's was an empirical determination.
Hubbert:D'Arcy was an experimental determination, and answered this one question: how does water flow through sand for the purposes of designing a sand filter? He wanted to know how to make a filter, for a given flow of water. That's all he was interested in. He got the answer, that the rate of flow was proportional to the difference in the manometers, top and bottom, across the sand bed, was directly proportional to the cross-section of area, inversely proportional to the distance between the two manometers. So that would be H 2 minus H 1 over L, would be DH by the gradient H. Divided by the area gives your flow per unit area. And so that becomes D'Arcy's Law in a differential form. Well, D'Arcy's colleagues called this D'Arcy's Law shortly after that and it was so named and so used by ground water people. Well, anyhow, the answer to the further question was that yes, this relation is true for any direction whatever, same as if it were electric current. But in terms of the pressure, if you're going down, you're flowing at a slow rate of flow, you're flowing from lower to higher pressure. If you were going up, you'd be flowing from higher to lower pressure. And if you were, say, inclined with two manometers, and the height of them was the same difference as their bases, the pressure would be constant. It would be flowing along an equal pressure surface. Anyhow, we worked that out. Just at that time I discovered a new book over in the Engineering Library, and that's this book by Muskat, The Flow of Homogeneous Fluids Through Porous Media.
Doel:You hadn't read any of Muskat's work before you made your own study?
Hubbert:I'd read sketchily, papers in JOURNAL OF APPLIED PHYSICS, but not critically. In fact, just that these guys were doing some interesting work. And then this book came out, and I thought I'd flip the pages of it. This is an exhaustive advanced treatise on the subject.
Doel:This is part of the McGraw Hill series.
Hubbert:Yes. And I recommended the book, and touted it to various people as being just that for the next year. Or two years. Then I think I told you about this conversation I had with a friend, or did I?
Doel:We don't have it on tape. We should.
Hubbert:Well, one of my fellow students at Chicago was a guy by the name of Deversy Crombie. And he wound up a member of the faculty of the geology department.
Hubbert:At Chicago. We'd shared an office together, and became very good friends there. We were both attending the International Union of Geology and Geophysics meeting here in Washington, during September 1939, the time when Hitler was invading Poland, that week.
Hubbert:Yes. Well, Crombie and I were staying at an inexpensive place that had formerly been a small men's club of some kind, bachelors lived there. It was a residential club, but by now it was a small hotel over on I St., I think it was. We were attending the thing and mostly attending the same sessions. We were having breakfast together and dinner together and everything together and during this time. Crombie has the most inflexible, one-track mind of anybody I've ever known, he can only think of one thing at a time and he's interested in it to the exclusion of everything else. He was working up a course he was going to give the next year at Chicago on petroleum, and he was going to use this Muskat book as his principal reference. So during that week or ten days, we had this course and we had D'Arcy's Law of this and D'Arcy's Law of that, three meals a day and between meals. I was getting a little tired of it. But I think there was one final day of the meeting, maybe a Monday. On Sunday morning we were over on the terrace, top terrace kind of cocktail bar or lunch bar or room or something at the Hotel looking west kind over in the direction of the White House. I remember we had a maybe a late breakfast and a couple of beers, just a lazy morning, and Bill was sounding off on D'Arcy's Law. Well, I had a diabolical urge. I said, "By the way, Bill, what is D'Arcy's Law?" He looked startled. I said, "I mean it, what is it?" Well, he rolled off some equation and I said, "Write that down. Let me look at it." The equation that he stated [begins writing] was vector V was equal to a minus small k divided by a mu gradient P. V is the volume of fluid crossing area per unit of time, normal to the flow lines, where mu is the dynamic viscosity of a liquid, and k is a fraction of proportionality in the equation. Gradient P is the gradient of pressure in three dimensional space. And I said, "Bill, that isn't D'Arcy's Law. Neither is it physically valid. Let's see what that equation says. It says, if the flow is perpendicular, equal pressure surfaces, and in the direction from higher to lower pressure."
Doel:Which you'd already discussed at length in your own class?
Hubbert:I hadn't discussed it with Crombie before.
Doel:That's right, but with your class?
Hubbert:Oh yes. OK. That was two years before. So, well, I can violate, I said, that equation experimentally with a table-top experiment any way you want to name. "I can take a tube," and I drew a picture of a tube with sand in it, "and inlet and outlet above and below, and I can let the fluid flow downward through the sand or up through the sand, with the manometers stuck in there, and so I can make it flow from higher or lower pressure, flow at higher pressure, flow when the pressure is constant in three dimensional space, and so on." Well, Crombie was no physicist but he was very, very deeply involved in statistics, but with no depth of understanding of that either. Crombie was a kind of a literal mind without too much understanding.
Doel:Had he been trained primarily in geology?
Hubbert:Yes. I'll give you a little background on that when I finish this particular item.
Hubbert:So I pointed out to him, I said, "This equation is physically erroneous. It violated any way you want to name with a simple table top experiment." And well, he was defensive, of course, and said he'd got it out of this book by Muskat. I said, "Bill, you surely must be mistaken. You must have misinterpreted it. This guy's a PhD from Caltech in physics. He and his colleagues have been carrying out laboratory experiments in fluid flow and published in the JOURNAL OF APPLIED PHYSICS the last six or eight years. And then he puts this book together out of his background. He couldn't possibly have written such an equation." Well, Bill said that that's where he got it. I said, "Well, maybe I'd better take another look at that book." \ I did. I went back and looked at it. And I found that that was a fundamental equation of the entire book! But without any understanding, any vestige of understanding of the physics of it. In the first place the equation was wrong, and he had no understanding of the fact that it was wrong. That was what prompted me then to write a little paper on this subject, primarily on just the things that we've been discussing. That was October. I started to work, to reading Muskat, reading the literature and doing my own work. I was on leave from the university at the time, and working entirely at home and entirely on this project.
Doel:Had you taken leave in order to do this work?
Hubbert:Well, I was in the process of leaving Columbia University, and this was my last year there anyway. I was kind of assigned this time to mop up. So what I was doing was really just lifting myself by my own boot straps and starting from zero, you might say. I drew up the first draft of this little paper, ten or fifteen pages. In so doing, all kinds of auxiliary intriguing questions arose, about so and so and so and so. By the time I worked that out, what needed to be put into the paper, it would be better if I didn't do the paper this way. So I started all over and approached it from a different angle. I marked that Roman I, put it over on this corner of the desk, and started over again. Then I got this far, with the addition, and I ran into something else. I went through the same problem all over again. By the time I got that straightened out, it would be better if I do it another way. So I marked that Roman II and started from scratch. I think the final draft that I had was Roman VII, which was 169 pages in print.
Well, that's evolution. Getting back to Crombie again for a moment: when we were occupying offices together, Crombie had been in some kind of a business situation for a year or two out of school. Then he had come back to school. Pettijohn had come there, and he was working under and with Pettijohn in sedimentology. One thing he got interested in was rounding of pebbles by a stream, the stream carrying these pebbles. So he set up an experiment in the basement of some kind of a tumbling barrel and he put cubes, marble cubes in this barrel. I don't know whether he had water in there, I guess he did, but anyhow the barrel tumbled. You took these cubes out and looked at them after a certain number of mileage of tumbling, and then you put them back and repeated this. So when you first looked at them, they'd kind of rounded the edges and corners but the faces were still flat.
A little farther along, the corners and edges were more round and the face had a circular area, it was flat, all the rest was curved, surfaces were curvilinear. And then finally at the end, this circular flat area got smaller and smaller until it finally disappeared. Well, Crombie told me these things, and was marveling at why were these edges and corners rounded off first? I said, "Bill, if your head were cubical, what would you knock the hide off more frequently?" And I took a pencil and said, "Why, if you drop a pencil or a pen, why does it always hit on one end? Never flat on the side?" And so then I gave him an elementary lecture on the theory of probability, that the probability that of its hitting flat on its side was zero, and the probability of its hitting any other angle depended upon the angle, that small angle you were looking at as compared with the totality. So I said, "Well, you better go over to the physics department, the mathematics department. They have a course in statistics there, you'd better go and take it." Well, he did. And that was his first connection with statistics. As I say, he went through and spent the rest of his life messing with statistics. But he never had any depth of understanding. It was all superficial. He was a tremendous worker. He published a great range of things. But he never had any depth. He never did understand this Muskat thing we started off with, because he didn't know any physics. He might have formally known some formulas, but he didn't understand them. And so that was the state that he was in when we had this powwow in September, 1939. So yes, he published all kinds of things statistically, later had a book about so big in statistics that he wrote. I have it somewhere in one of these cardboard boxes, put there when I moved. I never dug it out.
Doel:Was this the first occasion that you realized problems inherent in the Muskat book?
Hubbert:I didn't know there were any problems. I was telling everybody this is a great quantitative piece of work. All I'd done is flip the pages, and it had all the earmarks of authenticity. There's a little story involved there also. When I wrote this paper, well, I'd propositioned Chamberlin when I decided to write this paper.
Hubbert:Rollin Chamberlin. I said, "Look, I'm writing this paper. It's going to be bigger than your ordinary journal paper, maybe 50 pages. Do you think you can handle it? Or should I look somewhere else?" Well, he thought he could. How soon could he have it? I said, "Well, I don't know exactly, but two or three months, couple of months maybe." That was along about maybe November. And I was working. That was one case where I was 100 percent concentrated. The more I worked on this thing, the more concentrating it became. I'd start in in the morning, I'd work, maybe a short break for lunch, and I'd work till about 5 in the afternoon, very concentrated work, writing. One of these things where you woke up at night and solved problems during the night that you hadn't been able to solve during the daytime. And all these cumulating traps. Finally Chamberlin was getting very impatient, because he had to plan his journal for the rest of the year. It was getting on about March.
Doel:This is 1940 now?
Hubbert:Yes. So he demanded that I had to get this thing to him. Well, that's when I'd just finished writing Draft VII. I knew it was running over 50 pages but I hadn't had a lot of time to really check up on it. So I packed it in a typewriter paper ream box, and sent it off to him. I relaxed for the first time in weeks, and my wife and I were settled down for a couple of good scotches and sodas. The wheels started turning around and I found myself doing a little calculation, and finally it suddenly crashed through to me: my God! this thing's going to be over a hundred pages! Printed in the JOURNAL! (laughter). And I was embarrassed. I got off an airmail letter, immediate letter, and told him I was very sorry, I'd been working so intensely that I'd lost track of how big this thing was and didn't realize until after I'd mailed it that it was way beyond what I had said. The only thing I could suggest was that he just send it back and I'd see if the GSA could handle it. It was plainly too big for the JOURNAL. He wrote back and gave me hell but he didn't send it back.
Doel:He decided that he would publish it in the JOURNAL?
Hubbert:Well, it hadn't been decided, but he didn't send it back. I just told him, I said, "I'm sorry, it's out of bound, it's my fault, and the only thing I can suggest is send it back and I'll see if the GSA can handle it." What he did was that he turned it over to Carl Eckart in physics. Carl Eckart was the top theoretical physicist in the physics department.
Doel:You'd already had contact with him?
Hubbert:I'd called him on one of my projects. In graduate school I had about a year's work under Carl. I had senior college electricity and magnetism. I had vector analysis. And I had an advanced course in electromagentic theory. All under him.
Doel:Those were graduate seminars?
Hubbert:Well, one of them was undergraduate. The others were graduate. Carl Eckart and I had become very good friends, and so Chamberlin got in touch with him and sent it over. Well, Carl read this thing, and sent it back with an OK, but with a few pertinent comments that such and such item could be improved here and there and so on. So I did, I made minor revisions. Chamberlin then set the thing up in the JOURNAL in two parts. The regular issue of the JOURNAL had the year's index in it. One was Part 1, the other was Part 2 — anyway, the Part 2 had the index. Mine was run as a separate part of the JOURNAL, self-contained, had its own covers. Well, to jump on up to about 20 years, I was invited out to Scripps Institution about 20 years ago to give a series of lectures by Carl Eckhart on hydrodynamics. I had dinner with him and a woman who worked for Shell, a micropaleontologist from MIT, graduate work at MIT and so on, and now at Scripps. The three of us had dinner together. Her name was Frances Parker. And Carl told this story, I'd never heard it before, that he also had reviewed the manuscript of Muskat's book and had approved it.
Doel:Is that so?
Hubbert:And didn't see any of these things that I pointed out! (laughter).
Hubbert:Apparently he was a little jolted with this geology student came along and pointed this out in this book and he hadn't even seen it!
Doel:Was yours the first to point out any errors in the Muskat book?
Hubbert:Not quite. There was something else going on in parallel. There was a guy by the name of — I believe it's L. F. Richards, I'm not quite sure.
Doel:OK, we'll check that.
Hubbert:He was a physics student, took his PhD at Cornell in the thirties, maybe mid-thirties or so. He had a thesis on the flow of — this was mechanics or physics — the flow of water through sand. At this later date, he was with the research branch of the Department of Agriculture in soil, irrigation and so on. Well, this man had done a very able physical analysis of this problem, the equation of flow and so on. But I didn't know anything about it, I'd never seen his paper.
Doel:You didn't know about it at the time?
Hubbert:Never heard of him. I only heard of him after this, after my paper was published.
Doel:That was December of 1940 that your paper was published.
Hubbert:And later than that, one of the men in the groundwater branch pointed out to me this article by Richards which I'd never seen before. About the same time Richards was chairman of a committee of…I don't know whether it was the American Geophysical Union, or whether it was the JOURNAL OF IRRIGATION or something or other — I think it was the AGU—and he was the chairman of a committee whose assignment was to try to straighten out the nomenclature in this field. Things like — oh, the various terminology.
Doel:Primarily hydrology and ground water?
Hubbert:Well, this field of flowing through sands. You had the petroleum people using one set of terms, ground water people were using another, civil engineers using another, and irrigation people — the thing was a nomenclature mess. So he was the chairman of a committee to try to straighten out this nomenclature. And in that connection, he sent letters to people who were interested who might be able to comment usefully on this problem, and one of them was sent to me. And I believe, yes, by this time, Muskat had challenged me at a meeting at the AIME in May. I think I told you about that.
Doel:We don't have that on tape, though. We need to hear more about that meeting.
Hubbert:Well, in the chronology here, see, I published this paper in 1940. When it came out, it made the two principal camps involved fighting mad.
Hubbert:Furious. One was the petroleum people and the other was the ground water people.
Doel:Were the ground water people primarily in the Geological Survey?
Hubbert:In the Geological Survey. Not totally. And so, in February, 1941—I'm trying to remember whether it was '41 or '42, I think it was '41 — the AIME was having its annual meeting in New York, and at that time the petroleum part of the AIME was just a committee. It's now the American Association of Mining, Metallurgical and Petroleum Engineers. At that time petroleum was just a minor group in this organization. So the head of the petroleum committee, or whatever it was called, division maybe, had written me a letter inviting me to give a paper on this subject at their meeting.
Doel:This was after you published your paper in the JOURNAL OF GEOLOGY?
Hubbert:This was after my paper had been published, and that's the reason I was trying to remember whether it was '41 or '42. I believe it was '41.
Doel:OK, I'll be glad to check that for you.
Hubbert:Anyway, I was invited essentially to give a kind of synopsis of my published paper at this petroleum engineering meeting. Well, the meeting was an afternoon session, and I noticed Muskat in a restaurant nearby. I happened to see him and a friend of mine from Houston named Slotnik having lunch together.
Doel:Had you known Muskat before?
Hubbert:Yes. We had a meeting in Washington over this thing at the AGU meeting in the spring, in April, and a private meeting. A dinner meeting was arranged where I gave a synopsis of what I was working on to a number of people and Muskat was there. So all right. This meeting convened in the afternoon. I was the first one on the program, I think. So I gave a synopsis of what was contained in this published paper, which is what I was asked to do. Muskat got up and said that all I had done was really doing in a slightly different way what he had done more elaborately in this book of his, and that if I could show one single thing where my method of analysis would give the correct answer and his would give the wrong one, he would concede that I had done something original. I said, "Well, Mr. Chairman, that's a very fair proposition. And as a matter of fact, there are problems where my method gives quite incompatible answers or results from Muskat's. And I'll give you one example. I drew on the board an inclined surface, with coarse sand and fine sand. (starts a sketch).
Doel:Here you have two parallel layers of sand, but tilted.
Hubbert:I had a plane interface between two bodies of sand, one coarse, one fine, and the flow in number 1 obliquely to this interface, cutting it at an angle intermediate between vertical and tangential. So I said, "This is a problem that I think is the situation." Then he said with considerable condescension, if I'd only read his book a little more carefully I'd find that he'd treated that very problem in the later pages of his book. I said, "Well, Mr. Chairman, it's been over a year, nearly two years since I read this, but I have my reading notes with me, and this analysis occurred on page 401, and the equations given for the boundary here are that the normal component of flow in region 1 and region 2 is the same, and the tangential components of flow in region 1 and region 2 are given by the negative gradient of the pressure. That's the force P with respect to S. And that is physically erroneous." I said, "As far as the mathematical form is concerned, it gives you the tangent of the refraction. But the equation is physically erroneous. And you would have gotten the tangent along the refraction if you had assumed that the flow was proportional to the grade and the temperature. And you didn't measure the temperature, you assumed it!"
Doel:Muskat's formula doesn't give any physical insight into the process occurring.
Hubbert:Well, it is physically erroneous, because imagine the pressure you'd find if it wasn't what he had assumed. The real pressure is totally different. He assumes the pressure to fit the equation. And the equation is a mathematically valid form, but you'd have gotten the same thing if you'd assumed it was proportional to the gradient of the temperature, and then measured the temperature. Well, he became emotional and abusive, and beyond that there was no further rational discussion.
Doel:Were others who sided with him also at that meeting?
Hubbert:He didn't have any of his gang with him. There were other people present who were incensed by his performance. People I knew from Humboldt told me this later. A Shell man was also present at that meeting and they were incensed at Muskat's performance. Then almost in parallel with that — in fact it was in parallel — I got a manuscript from Chamberlin of a discussion that somebody had mailed in. It had been written before this meeting, but the argument was almost identical.
Doel:Was this by H. Krutter?
Hubbert:Yes, at Penn State. He was shocked that anybody could suggest that there was anything wrong with this Muskat book he'd used in his courses. He was sure that it was all OK, and if anybody saw anything wrong...things almost identical with Muskat's argument, which I interpreted to mean that Muskat had asked him to do it, take up the argument.
Doel:You'd never met Krutter at that point?
Hubbert:No. Never or subsequently. Anyhow, in reply to Krutter's paper, his discussion was published and also my reply which I wrote. I don't remember now what it was that we got off on all of this.
Doel:We were talking about the ground water work that you were doing.
Hubbert:Oh yes, we come back to Richards. That questionnaire that Richards sent out also contained—or at least in what he said to me, because I think he had read this discussion. I believe he had read the discussion I've described, because he enclosed in it the correspondence that he had had with Muskat on this same subject.
Hubbert:What happened was that Richards had sent this questionnaire to Muskat earlier. I guess I wasn't included in the chain of the original recipients of the questionnaire. He'd sent me this material with the correspondence in it because of this discussion. The correspondence with Muskat was very interesting, because Richards wrote out the correct equation of flow, including gravity and so on. It was physically correct. And Muskat had said, well, now, we'd always been worried about what to do about gravity, but we don't quite know what to do about it — in reply to a letter that is physically correct about the statement of this. He simply did not understand the physics of what he was talking about. Somewhere in my archives I had that correspondence.
Doel:It would be good to see that.
Hubbert:Maybe it's in a buried file somewhere, and God only knows whether I could lay hands on it or not if I searched. But anyhow, Richards did understand this thing, and he had in it finally, in fact he preceded me in writing a short but physically valid statement of these relations. Earlier than that historically, which I may have mentioned before, there was a big study carried out at the University of Wisconsin, back in the 1890s, on the flow.
Doel:On groundwater flow?
Doel:OK. I'd like to learn more about that.
Hubbert:That was in two parts. I mean, there were two facets to it. One was laboratory, and there was a man by the name of I think F.H. King in the School of Engineering at Wisconsin. He carried on a very elaborate set of laboratory experiments on flow of water through sand.
Doel:This seems early for this kind of work.
Hubbert:The 1890's, around 1896,'97, '98, somewhere along in there. Parallel with that, the dean of the graduate school, Charles Sumner Slichter, was a theoretical physicist, and he attempted to solve this whole problem mathematically, by mathematical analysis. Well, it was a heroic effort. I wonder if there's another copy of that. A book on the shelves over there. Then what he said was that the generality of ground water, that the pressure corresponds to things like gravitational potential, practical potential, and temperature and the more familiar physical equations, in this case the quantity of pressure. Delta squared p equals zero and so on. Well, he just simply jumped ahead. It violates the equations which are valid. It violates some of his own calculations which are valid, in individual cases. And yet he jumps to this generalization which isn't so. And that generalization had more influence subsequently than his valid work, to people who were theoretically inclined.
Doel:Charles Slichter didn't realize that error during his lifetime, did he?
Hubbert:Apparently not. But it was a boner, as I pointed out in the introduction to this book.
Doel:Right, we can bring that into the interview.
Hubbert:And yet this — you see, the subsequent, that is the recent theoretical people have frequently taken off from some generality like that. One of the erroneous take-offs is to assume that flow has a velocity potential — I believe Slichter said so, that P was the velocity potential. Well, now, what is the velocity potential? That goes back to classical hydrodynamics. And there are a bunch of theorems of classical hydrodynamics that are based on the flow of an ideal incompressible frictionless fluid, and that mathematical fluid has properties in singly connected space. You can't go around an island, but as long as your closed curve is within the space, that V is equal to the gradient of a scalar, well, well say V, minus gradient V. So, put the other way, V then is the interval of VDS along the path. And it's very easy to show, in fact that's one of the things that Muskat does: Muskat assumes the velocity of potential exists, and that this quality of P is it. Well, you can take a simple thing like, suppose that you have two sand channels, say, between two basins, the real potential of this first basin, the input one, and potential of the other one would have to be the same. Now, if it's the velocity potential, the velocity in one of these channels is coarse and the other one's fine, and it's very much greater, it can be ten times greater than the other one. Suppose it's twice. So your integral VDS from here to here, has to be the same as the other one. Well, it isn't. The distance is the same but V is not the same.
Hubbert:So that that would give you two different potentials at the other end of the line. You start at one end and do that integration. Or if you do it around a closed loop, you do not get zero. And the derivative of and potential around a closed loop has to be zero, by ds. So that's one of the more recent standard errors, that is, the last few decades, the author will take off on this assumption that the velocity potential exists. And one form it takes is to take this business: you've got the valid equation, there's a V or say, Q, a small q, which I prefer, is equal to minus say K gradient H. Now, that's a valid equation, where K is a permeability factor. That's not all the parameters: the property of sand is lumped into this one. Well, the velocity potential approach is let KH be the velocity potential, and then you say the Q is equal to the gradient of KH. The gradient of KH, the K gradient and H are not the same unless K is constant in three dimensional space. So the first equation is valid whether K is constant or variable. The second one is valid only provided it's homogeneous in space.
Doel:Which you're not often going to get.
Hubbert:Yes. So that's one of the standard errors, is the assumption, you take off, when you're dealing with a velocity potential. And that is just the straightforward assumption that you're dealing with the equation in terms of gradient P. Well, I think that's about all we'd better try to do — I'm pretty tired.
That's fine. Thank you very much for all we did cover.