This material is provided as a public service to support the student Space Settlement Contest. The views expressed herein are not necessarily those of NASA or any other government body.

John Holt * Technical debate *

John Holt

Author of How Children Fail; How children learn; What do I do on Monday?

It would take a book to discuss fully the many flaws and errors in Professor O'Neill's proposal for space colonies. Let me here mention a few.

First of all, the basic design of his cylinder is unworkable, for this reason If the cylinder is rotating at a speed sufflcient to produce centrifugal forces (nor gravity, as they carelessly call it) equal to 1g on the inner surface of the cylinder, then there will be equal or greater centrifugal forces on everything outside the cylinder which is rotating with it, ie. the movable vanes which will supposedly reflect sunlight into the inside. Folded in flat, these vanes will experience along their entire length a centrifugal force equal to 1g. Fully extended they will experience forces averaging about 4-5g. But these vanes are supported only at the hinge end. To see how far this is from being possible, we have only to ask what is the longest structure that we can make on the earth's surface, parallel to the surface, and supported or cantilevered at one end. Hardly more than fifty yards, if that. O'Neill asks us to believe that with present technology we can make equivalent structures many miles long. Clearly, we will not do this in ten years, or a hundred; the odds are great that we will never do it.

If, then, we were to build one of O'Neill's cylinders and start it spinning, it would not be long before the vanes would begin to bend, fan out, and soon break off. We must ask ourselves, how is it possible that the by now tens or hundreds of thousands of "scientists" who have read of this proposal have not pointed out this elementary flaw? Either they have not seen it, because they did not want to, or because they are, quite literally, out of touch with reality, have lost the feel of real things. Or they have seen and not spoken, perhaps from fear, perhaps so as not to impede a project that would be good for "Science." Either way, they have shown that we dare not take their word about the benefits, feasibility. or costs of such a project, despite their credentials they have shown themselves to be incompetent.

This defect in the design cannot be cured by supporting the vanes with cables. No cables are anywhere near strong enough to support many miles of their own length against such forces, to say nothing of the weight of the vanes. And if super cables, ten times the strength of any we have, should be invented, they could still not prevent these vanes, a mile or so wide, from sagging in the middle under the stress of centrifugal forces. Of course, the problem could be solved by having the reflecting vanes fixed in space, independent of the cylinder and not rotating with it. But there is no way with such an arrangement to have the day-night cycle which is the heart of O'NeiIl's plan.

Let me be clear at this point about what we can and cannot do. We can build cylinders in space as big as we want, if we do not try to put earth "gravity" and atmospheric pressure in them. Or, we can have earth "gravity" and pressure, if we keep the cylinders very small. But not with any technology we have now or are likely to have for a very long time, if ever, can we build the kind of cylinders O'Neill describes. We can make - have already made - human habitations in space. But they are, and must and will be, like submarines: they will never be like the surface of the earth.

This fundamental error of O'Neill & Co - hereafter ONACO, though we could as well say CRASCO (Crackpot Scientists & Co ) or even MASCO, is only the first of a great many. Among others:

It is impossible, using available technology, or any likely to be available in the next generation, to refine metals and do other heavy industrial operations in space, or on the moon's surface, and any technology we ever develop to do it will be enormously expensive. The reason is that in a vacuum there is no medium to carry away excess heat. A rough figure often given is that it takes about 50 tons of water (and no one has estimated how many tons of air) to make one ton of steel.

The task of this water is simply to take heat, first, away from the steel, and then, away from the steel mill, and to dump it into the heat sink of the earth environment. What weight of radiating surface, and how disposed, would be required to dissipate heat at this rate into a vacuum? (It is another careless mistake to speak of space as being "cold"; it is neither cold nor hot; objects in it may be cold or hot, depending on how much energy they receive and radiate.) The amount would be immense; at the temperatures at which humans can live and work, and at which presently existing machines are designed to operate. radiation is a very inefficient way to dissipate heat.

At any rate, the technologies to do such metal refining and shaping in a closed system surrounded by a vacuum do not exist, not in a pilot model, not on a laboratory scale, not on a drawing board, and probably not even in ONACO's imagination In like manner, the technologies do not exist that would, in O'Neill's airy (no pun) phrase, "unlock" oxygen from the rocks on the moon. On earth's surface we break oxygen loose from metals by heating the metals to very high temperatures and then giving the oxygen an abundant supply of carbon to react with instead. On the moon, where would we get the carbon? How would we then free the oxygen from the carbon? And, in what sort off containers and heat exchangers, and with what sort of pumps, would we contain, cool, and finally compress these huge amounts of enormously hot gas?

Nor do we have, even on drawing boards. the technology to build on the moon's surface the materials launcher which is another vital part of ONACO's plan. We do not have linear motors capable of applying to a tracked vehicle flle proposed accelerating force of 25+ g's and a decelerating force at 75+ g's, nor do we have speed measuring and controlling devices of the required sensitivity that would, and for long periods of time, withstand such forces, nor do we even have the vehicles themselves. Nor have we learned how to make a track level enough so that a vehicle could run smoothly on it at the required speed of about 4,000 miles per hour; the record on earth, and that over a very short distance, in a rocket powered sled, is only about 600 mph. Such a track would have to be heavily ballasted, not perched on flimsy supports as in ONACO's drawing. Nor have we any idea how these problems, difficult enough on earth, might be made more difficult by the moon's lighter gravity, different surface conditions, and, what is most important, enormous fluctuations in temperatures.

In these and many other respects ONACO have underestimated, by factors surely as great as ten and probably very much larger, the difficulty and expense of devising, building, testing, perfecting, and maintaining the devices needed to do the things they want to do. We do not yet know how to build a lunar habitat for even a half-dozen people; ONACO's plan will require a habitat that will house, and for long periods of time, hundreds and perhaps thousands.

In like manner, ONACO have grossly underestimated the weight of material that would be needed to build the space cylinder. From their words, and the artists' drawings, they seem to have no idea of the degree to which changing the scale of a problem changes the nature of the problem; they are like people who would try to build a 747 out of the same materials as a model airplane. Scale makes a great difference; to go from the 707 to the 747 we had to invent and make not only new metal alloys, but new machines to work those metals, among them forges many times larger than any that had existed.

Consider, for example the effect of atmospheric pressure on the end of the cylinder. It is very much like the problem of building, on the earth's surface, a water tank, to be suspended from its upper rim, and to hold a 32 foot depth of water. If such a tank was to be ten yards wide, we could build it with fairly conventional means heavy steel plates welded together. But if we try to imagine such a tank 50 yards wide, or 100 yards, and get some sense of the forces on the bottom, and their bending moment, we can see that much heavier construction, with massive stiffening beams would be required, but ONACO are talking about such a tank more than a mile wide! In like manner the side walls of ONACO's cylinder would be subject to immense forces, both atmospheric and "gravitational;" it would take enormous amounts of reinforcing beams, both longitudinal and annular, to prevent the cylinder from bulging out into something more like a sphere, or even a disc.

Scale is also important in the matter of the microwave projection of energy that is another central part of ONACO's plan. No doubt we can transmit power through microwave energy on a laboratory scale, but that is not at all the same thing as doing it on an industrial scale. The technology to do that does not yet exist. And if it did, or when it does, how big will be the target area on earth to which this energy is projected? More important, at such distances, through what kind of feedback mechanisms will the projector be kept on target? And more important yet, suppose these mechanisms break down - things do break down - and this energy beam, probably close to what we might call a Death Ray, starts to wander around on the earth's surface. What then?

As to the cylinders themselves, even if we could solve the vane problem and so get a cycle of day and night, which we can't, we would still not get an earth-like environment. What about rain? To get rain at night, the relative humidity would have to be very close to 100%, certainly far above any level of comfort. What about balance? The weight in the cylinder would have to be kept in balance, not only around the axis of rotation but also down the long axis. Otherwise the cylinder would begin to rotate eccentrically, or to wobble. There being no correcting forces the cylinders would be in unstable equilibrium, and these motions, once started, would tend to increase. For that matter, an object at L5 is itself in unstable equilibrium; there are no forces tending to keep it there, and since any movement will bring it into the gravitational fields of earth or moon, once it starts to move it will keep moving. And what about wind within the cylinders? Since, if the vanes could be made to work (which they can't), the end of the cylinder nearest the sun would always get much less sun than the far end, and would hence be colder, what sort of air currents might be set up? And how long would it take to establish a stable biosphere, and how would it be done? O'Neill's brief remarks about doing away with unwanted pests show that ONACO is thoroughly ignorant in this area. And even if some miracle technology of the distant future, as yet undreamed of, could solve all these problems, it would not produce an environment like the surface of the Earth. Living there would not be like living on Earth, only nicer; it would be like living in the inside of a big rotating cylinder with mirrors outside reflecting in sunlight. Who would choose to spend the rest of his life there? Not me, for sure. Only the starving and desperate - and for them, no such palatial accomodations would be needed.

Space is not Heaven. It is not even Disneyland. It is an environment as hostile and deadly as the core of a nuclear reactor or the inside of a tank of nerve gas. In time, we will probably learn how to move around in it a bit more and do a few more things in it. But Earth's major problems will have to be solved on Earth.

Technical debate

John Holt's above letter to us has been answered hotly by one T.A. Heppenheimer of the Center for Space Science in Fountain Valley, California, writing to someone named Cheston at Georgetown University in D.C. John Holt sent Heppenheimer's letter to us, along with his own (Holt's) counter-remarks, which were addressed originally to Senator Edward Kennedy. A peculiar form of private publishing, all this, but a fascinating debate. I have trimmed preambles and shuffled the retorts and counter-retorts together.

Heppenheimer begins:

Before proceeding to address the technical points, I believe it is first worth-while to consider points of theology. There are a number of places in the paper where the author makes assertions which can only be described as theological, as articles of faith. These are the statements that space is insuperably hostile, that it can be of no significant value to man, that we must solve our problems on Earth, and the like. There is an alternate position, which is equally theological. This is compounded of assertions such as "the earth is the cradle of man, but man cannot live in the cradle forever," or the "man's colonization of space is as significant as the colonization of the land by aquatic animals in the Cambrian Epoch," or that "man's imagination and daring can overcome any limits."

I feel the pessimistic theology is naive, and is irresponsible. In a time of challenge to the foundations of our industrial civilization, it ill-behooves us to dismiss major technologies out of hand. But non-theological pessimism is valuable if it leads us to examine carefully proposed solutions and to probe critically for weak points. The optimistic theology is also naive, and may lead us to underestimate the obstacles which obstruct a difficult project. But a type of optimism is useful in that it may lead us not to be daunted by initial difficulties, but instead to seek to apply ingenuity and resourcefulness so as to overcome difficult problems.

I personally feel that the most fruitful attitude is not one of optimism or of pessimism, but of what might be called "critical ingenuity." That is, one must seek to be critical, yet to buttress one's criticisms with solid technical reasoning drawn from a multitude of fields. At the same time, one must have sufficient command of the pertinent technical fields as to be able to recognize what problems are truly difficult, what problems will readily yield to intelligent design. Then, faced with such a true difficulty, one must seek to cut to the center of the problem, to lay bare the core of the difficulty, and to apply the pertinent sciences so as to propose a solution. With these comments, then, I will try to attend to the dozen or so chief objections which the author has made.

(1.) The mirrors of the cylinders. These are exposed to at most about 2g, not 4-5.1 . Structurally, they consist of lightweight supports for extremely lightweight reflectors, for example of aluminized mylar.2 There is no reason to build them as cantilever beams. on the contrary, one can easily arrange a system of tension-line supports to guy the reflectors.3 Nor is it necessary to fold them in and out each day.4 It may well be preferable to build the mirror support structures as fixed assemblages, on which mirror panels are mounted in the fashion of Venetian blinds.5. They would then be tilted so as to give partial or total illumination, or to illuminate only a part of the colony. In short, the problem of mirror design, so far from being an insuperable obstacle, is the sort of problem I would cheerfully assign to a sophomore course in strength of materials.6

Holt replies:

1. Heppenheimer is mistaken here. The centrifugal forces on the reflecting mirrors will depend on the design of the cylinders. Given certain designs, these forces might be as low as 2g; given others, they could be as high as 5g's or even higher. Most of O'Neil1's articles about space colonies have described, and the accompanying illustrations have shown, cylinders four or more times as long as wide with mirrors long enough to reflect sunlight into the full length of the cylinder. As the accompanying sketch shows, for such a cylinder in the forces acting on the reflecting mirror when in the fully open (i.e. 45 degrees) position will range from 1g at the end nearest the cylinder to 9g's at the extreme end, with a 5 g average. If the ratio of length to width of the cylinder is greater, these forces will be correspondingly greater.

2. Heppenheimer has missed the point here. What counts is not the lightness of the reflecting material, but its rigidity. Since all of these mirrored surfaces will experience "gravitational" forces greater than lg, they must be rigid enough to remain flat under these stresses, or they will be useless as reflecting mirrors. When we consider the size of these mirrors - for some cylinder designs they might be ten or more miles long and two miles wide - it is clear that the supporting structures which will be needed to give the necessary flatness will not be simple or light.

3. In the first place, none of the drawings and sketches which have accompanied O'Neill's articles to date have shown or indicated any such cables. In the second place, to support the kind of structures mentioned in No. 2 above, a veritable forest of cables would be necessary. In the third place, if we imagine the 2 mile x 8 mile cylinder that O'Neill often talks about, the cables to support the ends of the mirrors would have to be about eight miles long. If, as would be the case in such structures, the average force on such a cable was 5g's, we would require a cable strong enough to support, on the surface of the Earth, 40 miles of its own length, plus five times the weight of the much heavier mirror and supporting structure. The strongest cables we now have will support about 35 miles of their own length - but that's all. The four or five times stronger cables we need do not exist. It could be said in reply that there is no need to design the space colonies in this way. But this is the way that O'Neill, in article after article for a year and a half now, has proposed that they be designed. And I must ask again, if the scientific community cannot see, or seeing, will not speak publicly about a mistake as great as this, how much can we trust them to tell us about other mistakes?

4. No, it is not necessary. But this is what O'Neill was for a long time proposing, to give an Earth-like illusion of the sun rising and setting.

5. Yes, but if the mirrors are attached to the ship and rotate with it, the problem of the centrifugal forces remains. If the mirrors are not attached to the ship, a new and equally difficult problem arises - how to maintain their position with respect to each other. As for the Venetian blinds, that would probably work, but at the cost of much of the supposedly Earth's surface appearance of the environment.

6. Perhaps - as long as someone rather more skeptical, and with a surer feel for the reality of things, was there to correct the papers.

(2) At the top of page 3 is a comment, "But they are, and must and will be, like submarines. . ." I presume this means the colony internal design. Who sez? Sez you! Even a submarine (viz, the Beatles' Yellow Submarine) need not be like a submarine. It is a matter of architectural design and of interior layout.6a The prospective population densities are expected to be 7 similar to those of San Francisco or of other cities, and the pleasure of living will be enhanced by an absence of autos, highways, and urban noise.

6a. By this I mean that these colonies, whether small or large, spartan or luxurious, will look like what they are - artificial environments, large containers Boating in space (perhaps with some windows). They might even be as luxurious as the lobbies of Las Vegas hotels, or the insides of luxury ocean liners, which some people like. What they will not look like is the natural environment of the earth's surface.

7. This is sales talk, or an unrealistic hope. If and when such habitats are built, it is almost certain that most of them, like military barracks or troopships, will be crowded. More on this later.

(3) Chemical processing for extraction of metals, oxygen, and glass. Before addressing this critical problem, let us first consider the defining parameters within which a solution must be found. The difficulties attending this prospect, space ore-processing, must certainly be daunting. To begin, the ores are not rich concentrations of oxides or other simple compounds, such as we find on Earth. Rather, we will deal with typical lunar materials such as anorthosite, plagioclase, ilmenite, forsterite, and the like. These typically involve complex chemical compounds, and concentrations rather lower than we are accustomed to dealing with.

Then, it is entirely true that we have no free availability of air or water, or of cheap carbon for reduction. Recycling of these materials will be essential.8 The problem of heat disposal will also be critical.9 But while we must be aware of these constraints, we also must realize we are operating under conditions which in other respects are more favorable than on Earth.

8. "Re-cycling?" Where are these materials to come from in the first place?

9. Very true. And, as I said in my earlier letter, here we have no experience whatever to guide us. We do not know how to dissipate very large quantities of heat without a cooling medium. Yet this is a problem that must be solved before any construction of colonies can begin.

The chief of these is the economics of materials production. On Earth, for example, it may be required to produce metals at fifty cents a pound. This is about right for steel, and somewhat lower than the cost of aluminum, I believe. But at the colonies, far higher costs are tolerable. The reason is that the metals are not to be produced for sale as raw ingots, but rather are to be used in space for construction of power satellites, and of similar projects producing a very high economic return.10.

For example, let us suppose that the prime chemical plant costs $60 billion over a twenty-year period, for its design, development, construction, establishment in space, and operation. In this time it produces one million tons of raw metals and two million tons of byproduct oxygen, most of which is used for rocket propellant. The net cost of these products then is $10 per pound, which by terrestrial standards is uneconomic. But the alternative, in space, is to ferry them up by rocket transport, at $100 a pound or higher.11. (This is the transport cost to L5.) Moreover, the power satellites built at the colony may by then be generating a revenue of $30 billion a year, thus amortizing the debt - and then some.

10. & 11. Here I must refer you to my point No.39. For reasons I point out, there will not be a very high economic return, if any at all. And the costs against which these costs should be compared are not the costs of sending this material into space from the earth's surface, but the cost of obtaining a comparable amount of energy from the sun (or wind, tides, etc.) on the surface of the earth. In addition to being aware of these economics, we should also be aware of the opportunities for integration of the chemical processing with the rest of the colony. The colony's thermal design may be so arranged that waste heat from the ore-processing would be used for internal space heating 12 Then, the entire surface of the colony would function as a radiator.

12. It is almost certain that the waste heat from the metals refining process will be far greater than the internal space heating requirements of the colony.

Having said this, let us consider specifics. What sort of methods for ore-processing appear of interest? We have considered methods for the extraction and refining of aluminum, of titanium, and of glass; of these, the smelting of aluminum is indicative of the processes. We begin with lunar anorthosite. This is melted and quenched 13 to produce a glassy solid. This substance is treated with sulfuric acid 14 to extract the aluminum in the form of its sulfate. The sulfate is then treated with chlorine 15 and carbon dioxide. The former is electrolyzed, yielding aluminum. The carbon dioxide is put through the Bosch process, to recover the carbon and to produce as a by-product, oxygen 16 Also, the sulfuric acid is recovered through acid reformation. 17

13. "Quenched." How? And with what? And how obtained?

14. How obtained? Is sulfur plentiful on the surface of the moon? What will be required to refine it?

15. Again, how obtained? Same questions as above.

16. What are the raw materials requirements of this process?

17. Again, what other materials are needed to carry out this process? My point is that to establish in space a smaller scale counterpart of our materials refining and processing industries on earth, a great many raw materials must be produced - not just aluminum, oxygen, glass, and a few others, but a host of metals, chemicals, etc. These processes are interlocked: to make A, we require B and C; to make B and C, we require E, F. and G - but also a great deal of A. To get the thing going at all would require that we lift out of earth's gravity and into space very large quantities of materials - hardly less than many tens of thousands of tons.

We have considered the individual steps required for ore refining 18 through such processes. The temperatures required fire typically of a few hundred degrees Kelvin where caustic chemicals are present, and up to 2000 degrees K for the melting. The chemical technologies involve such long-established methods as treatment with sulfuric acid or with chlorine; indeed, the carbochlorination and electrolysis steps represent a process patented by Alcoa, for use with low-grade ores.19 Certainly, we do not anticipate a need for any technologies as advanced as the hexafluoride methods which were developed so successfully for uranium isotopic separation, over thirty years ago.

18. Aluminum only. But an industrial base cannot be made from aluminum alone. The key metal is steel. What about steel refining? And what about chromium, molybdenum, tungsten, vanadium, and other metals needed to produce modern steel alloys? And what about tin, lead, copper? Do these exist generally on the moon's surface, and in what sort of alloys, and in what concentrations? And what kind of refining processes will we need to extract them? In taking the example of aluminum, Heppenheimer has picked the easiest case. What about the hard ones?

19. Yes; but these processes are not carried out in a totally enclosed space surrounded by a vacuum. We do not know how to do this.

We have estimated the overall systems requirements for a production capacity of 150 tons per day of aluminum.20 We find the plant mass required is 7600 tons. The chemical inventory is 650 tons; the processing equipment is 3500 tons. Powerplant mass is 2800 tons. We estimate the required energy as 76 megawatts for process heating, 115 megawatts as electricity. Of the latter, 70 is for electrolysis and 40 for carbon reforming (Bosch process). Some 600 tons is required for the space radiators, with area of some 100,000 square meters.21 The associated radiator temperature is 6000 K or less which is quite conservative for proposed space radiators.

20. It is not clear whether Heppenheimer thinks this is all the capacity that would be needed, or whether he has just picked this figure out of the air for purposes of illustration. If the former, the figure is absurdly small, as I will later show. And in any case, what this plant would be producing would be raw aluminum. Later, Heppenheimer talks of making much of this into cable. What would be the materials and power requirements for the factories to do that? Or to make the other forms of finished aluminum that would have to be used?

21 This is very speculative. In any case, there would almost certainly have to be some kind of closed circulating coolant (like the water-system in an auto engine to carry heat from very high temperature areas to the radiators.

I feel that this work represents a useful preliminary effort to assess the requirements for ore processing. This difficult problem certainly represents one of the critical areas in which a modest amount of research is expected to pay off in greatly enhanced understanding of the issues involved.22 However, this work certainly has given an indication of the types of issues with which we will be concerned. It is far, far too early to say we have achieved anything like full understanding of these issues. But we are quite prepared to propose and to defend answers to the types of questions asked in the letter.

22. This is very modest and tentative language, very far removed from the language of O'Neill in his articles. He does not talk about understanding issues involved. He says we know how to do this, right now. I will return to this point again.

(4) Lunar mass-driver. I am particularly pleased to have the opportunity to address this issue, since it is one on which I have spent a great deal of effort. So far from this mass-driver being beyond the state of the art, I would be quite delighted to point out how it may be built with existing lasers, tracking systems, alignment controls, cryogenics, and the like.

The linear synchronous motor, proposed for the acceleration drive, is an application of classical electric engineering. It certainly is much less complex than such commonplace devices as computers, particle accelerators, or similar electronic systems. It has been extensively studied for its possible application to high-speed ground transport. 23 It is irrelevant to state that such linear motors as we require have not yet been built, for the development of such motors appears to be a straightforward exercise in systems engineering. 24

23. One would think from these words that linear electric motors were in common use, their problems well understood, their bugs ironed out, as is generally true of computers and Heppenheimer's other examples. Such is not the case. The linear electric motor is in a state of early research and development. To my knowledge, it exists, at least as a form of vehicle propulsion, only on a few miles of test track in a few countries. No practical, operating, installations exist, and none are projected for something like another ten Years. We have to ask, if it seems likely to take ten years before we have a linear electric railroad on the surface of the earth, operating at perhaps 200 mph, with acceleration and decelerating forces of perhaps 1/2g, how long will it take us to develop and perfect a 4,000 mph railroad, with accelerating forces of 25g's and decelerating of 75g's, on the surface of the moon? I have no doubt that given enough time and money, it could be done someday. But it certainly can't be done, as O'Neill has repeatedly suggested, within the next five or ten years.

24. This is advertising agency talk, not scientific talk. One might say the same of the development from the Wright Brothers' airplane to the 747. Such talk ignores the relevant factors of time and cost.

The accelerating vehicles ("buckets'') are envisioned as being built around small, powerful cryogenic magnets such as are well understood by physicists.25 The velocity measurement systems proposed involved laser doppler from fixed locations, with no delicate hardware carried aboard the buckets; 26 in our estimates of the achievable accuracies, we have used the state-of-the-art performance of existing mode-locked lasers. 27 The problem of track smoothness is largely overcome by arranging for the buckets to be suspended above the track, by means of magnetic levitation; again we here propose to rely on technologies developed for ground transportation. 28

25. As it happens, they are well understood by me. For any who may not know, cryogenic magnets are magnets that operate at extremely low temperatures, not far above absolute zero. They are kept at these temperatures by liquid nitrogen or helium (perhaps other gases), liquefied by complicated and expensive processes, confined under high pressure, and heavily insulated from the outside environment. Such magnets are expensive, cumbersome, and fragile. They exist now in the protected environment of laboratories, not along the edges of a railroad track on the moon, where ground surface temperatures may vary as much as three or four hundred degrees. The distance, in time and money, from today's cryogenic magnets to the lunar railroad O'Neill proposes is comparable to the distance between the radio and airplanes of, say the 1920's and the color TV and airplanes of today.

26. I stand corrected on this point. I take it that Heppenheimer is talking about a gadget comparable to that which the lurking highway patrolman measures the velocity of oncoming cars. The difference is that instead of talking about an accuracy of perhaps one percent, we will be talking about an accuracy of something like one-thousandth or ten-thousandth of one percent - again, in an environment of wildly fluctuating temperatures.

27. Where do they exist? What velocities are they measuring? Under what conditions? Does Heppenheimer claim that, using existing equipment we could regulate the speed of an earth's surface vehicle to the above-stated degree of accuracy? Where has it been done?

28. Here again we take promise for performance. I have followed with some interest the developments in this area. The facts are that the companies, mostly German, that are doing research and development on magnetically suspended trains, are running into serious problems, so much so that within the last year the city of Toronto which had a contract with one of these companies to develop for them a magnetic-suspension system of transportation, has canceled the contract. It is a serious error of fact to speak of these technologies as "developed."

The principal requirement for accurate track alignment arises from the fact that slight misalignments will give rise to vibrations, transmitted to the payload on the bucket, thus preventing release and launch with desired accuracy. The track thus must be aligned to high accuracy immediately prior to release. This is to be done by supporting the track upon screwjack actuators. An alignment reference is provided by lasers; detection of track misalignments is provided by means of track-mounted Fresnel zone plates.29 The specifications of the alignment system have been taken as those of the existing alignment system of the Stanford Linear Accelerator Center.30 We have used methods of classical control theory to estimate the associated launch errors and miss distances; we find that following a flight of 40,000 kilometers, the launched payloads should arrive within a circle of 100 meters diameter.31

29. Once again, we have the difference between what can be done under closely controlled, optimum laboratory type conditions, and what would have to be done in a much more difficult, variable, and uncontrolled environment. The Stanford Linear Accelerator is, to my knowledge, underground, carefully shielded from shocks, temperature changes, and other disturbances in the environment. Moreover, this accelerator is, in effect, a railroad only for atomic particles; it could accurately be said to carry no load at all. The proposed materials launcher would be a railroad carrying loaded cars, weighing at least several tons, moving at speeds up to 4,000 mph. To maintain an equivalent degree of alignment and rigidity on such a railroad is not a task already accomplished, but a wholly new task.

30. Included above.

31. This 40,000 kilometer flight will be a spiral path around the moon. Neither O'Neill nor Heppenheimer say how high above the moon's surface will be these loads of lunar ore when they are "caught." To keep them within a circle of 100 meters in diameter will certainly require a very precise control of velocity, greater by a factor of 1,000 (I have read) than the control we have so far achieved over our space rockets.

Thus, the mass-driver appears to be well-understood, 32 in terms of its major features and requirements.

32. What this means is, "We think we know how we might go about trying to build one." O'Neill says we know how to build one right now. We might note here that the maximum speeds of railroad trains in active commercial service, even in those countries (Japan, France) that take railroads seriously and spend money on them, has increased only about 40 - 50 miles per hour in the last fifty years. Some of the problem is with wind resistance, which would not occur on the moon, but much has to do with the difficulty of making a sufficiently level and smooth track - and the requirements of the 4,000 mph moon railroad would be far more stringent in this respect.

(6) Lunar habitat. NASA has conducted systems studies for the definition of habitats housing several hundred men, and there is a large literature of studies for smaller lunar bases. 33 Indeed, had the U.S. space program continued at the pace of the mid-to-late 1960's (the pace of John F. Kennedy), then by now we might well be on the way to building such habitats.

33. A study is not a habitat. No doubt we had studies for the F-111 or the C-5, and no doubt the studies showed that they would be wonderful airplanes. Events proved otherwise. Building self-sustaining habitats on the moon is to a very large degree a much newer and more uncertain enterprise than designing a new airplane.

(7) Design of large cylinders; structural considerations. The assertion is made that it would be very difficult to devise suitable structures for pressure vessels several kilometers in diameter. This statement misses the point. It is not true 34 that ONACO set the goal of a mile-wide cylinder and then tried to define appropriate structural supports. Instead, ONACO began with specific, rather conservative assumptions 35 as to the technologies to be used. These were principally the bridge-building structural designs similar to those employed in the construction of suspension bridges. ONACO then undertook to solve the problem: with such designs, how large a cylinder could one build? To their great surprise, they found that cylinders a mile or more in diameter, and tens of miles long, would be possible.36 Thus, ONACO do not envision "massive stiffening beams"; rather, the emphasis is upon tension cables of conventional design. This involves the production in space of large quantities of wire and cable, rather than the fabrication of immense and massive structural members.37

34. I did not say it was.

35. That depends on who defines "conservative."

36. The proper comment here is GIGO, a computer maxim standing for "Garbage In, Garbage Out." Whatever O'Neill may have fed into his computers, what has come out is nonsense. The previously mentioned figure of 150 tons of aluminum ore per day shows that O'Neill has enormously underestimated the amount of materials that would be needed to build the kind of cylinders he talks about. It can be shown quite simply that if it takes a certain weight of materials to build a container of a certain size, to hold a certain pressure of gas, if we hold the shape and the pressure constant, the weight of materials needed will vary with the volume - double the volume, double the weight of materials. Or, in other words, the weight of materials increases as the cube of the linear dimensions. If we ask, how much material would we need to build, to use one of O'Neill's favorite shapes, a cylinder 100 feet long and 25 feet in diameter, to hold (with appropriate safety factor) a pressure of one atmosphere, a ton seems an optimistic answer, and two or three tons much more likely. But let us say a ton. Consider now a cylinder 2 miles in diameter and 8 miles long, not by any means the largest that O'Neill has proposed. Its linear dimensions are 400 times greater. This means, not taking into account the weight of its associated machinery, or soil, air, and water, or stabilizing ballast, but considering only the shell itself, we would need 400^3, or 64 million tons of material, to build it. For that little 150 ton/day plant, that would be something over 1,000 years worth of output.

37. So what. What is critical is the total weight of material. I realize now, too that Heppenheimer, when estimating the weight of his 150 ton/day aluminum plant, did not take into account the weight of the dwellings required for the people working there or the weight of the system required to raise the food that these people would eat, or the weight of the systems required to recycle their wastes. And this leads me to wonder, for every worker engaged in primary industrial production, whether of aluminum, steel, sulfuric acid, chlorine, hydrogen, oxygen, etc.,etc., how many more would be needed to maintain and repair these factories, dwellings, vehicles, "farms", etc., or to supply all the needs of the people ? How many people would we need in space to operate a factory that required 100 workers?

(8) Power beams from the power satellites and safety assurance. The power beam is to be formed by means of a phased-array antenna involving microwave generators (Amplitrons) and ferrite core phaseshifters. These exist and can be mass-produced by the electronics industry. 38 The beam will illuminate a target area at the Earth of some 50 square kilometers. Considerations of physical optics prevent the focussing of the beam into a smaller area.

38. No doubt they exist and can be mass-produced. But they have never been put together and used in this way and for this purpose, and least of all in space. We will have to learn, by trial and error, how to do that.

The associated power densities will approach a kilowatt per square meter at the center of the beam. This is somewhat less than the intensity of sunlight: 39 birds and animals will in no way be fried or cooked. However, they are likely to find the beam uncomfortably warm, thus avoiding it.

39. Here we reach the most extraordinary paragraph of Dr. Heppenheimer's letter. If his words mean what they appear to mean, and it is hard to see how they could mean anything else, we are seriously being asked to consider spending something like 100 billion dollars so that, from a Point in distant space, we may beam to 50 sq. kilometers of the earth's surface somewhat less energy than the unaided sun (at least on sunny days) regularly delivers to the same area. This surpasses belief. And it totally destroys the myth of the eventual cost-effectiveness of such a project. If we were getting from space what we could not get anywhere else, there might be some grounds for paying these enormous prices for it. But the relevant comparison must now be against the cost of developing solar and sun-related (wind, etc.) energy on the surface of the earth. At this point the absurdity and wastefulness of this proposed project becomes clear.

Excuse me, John. As you must know, the microwave energy continues to arrive in all weathers and all night. It is far more easily converted to electric current than sunlight - the inefficient (and waste heat producing part) of the conversion having been done in space. The steady supply eliminates the need for storage, which remains the major problem with solar conversion to electricity on the Earth's surface.

The beam will be formed using a pilot signal from the ground to provide a phase reference. This phase reference will serve to control the phase-shifters so as to form a tightly-focussed beam from the outputs of the individual Amplitrons. The pilot signal, in turn, will be run off power obtained at a fixed site (such as the nominal target area center) from the microwave beam itself. If the beam wanders off-target, this pilot signal will lose its power and effectively shut off. Then lacking a phase control, the phaseshifters will fail to provide a coherent beam. Instead, the individual Amplitrons will radiate into the entire forward hemisphere of the antenna. The beam will spread out, and at any point on Earth the signal intensity will drop by some ten orders of magnitude to levels such as are used in communications.40

40. I stand corrected here; I suppose such a control system could be made to work reliably. But No. 39, makes the question irrelevant; if we are not beaming energy to earth in enormously high concentrations, there is no economic reason or justification for doing it at all. Thus, it is not the case that we will rely on techniques which can fail and produce disaster. We will not build a power beam transmitter which could swing about wildly, the beam being possibly quite hard to turn off or to control. Instead we propose that it will be somewhat difficult to form the beam, and that a continuous control will be required merely to keep it properly focussed.

I am amused by the comment, "No doubt we can transmit power through microwave energy on a laboratory scale, but that is not at all the same thing as doing it on an industrial scale. The technology to do that does not exist.'' I suppose one would have made a similar comment about transmitting human voice or music through microwave energy, in the year 1910.41 Actually, we are somewhat better off than that. The development of radio was immensely advanced through the invention of the vacuum tube, but we already have all the inventions 42 we need to do the job.

41. It would have been true. I return again to the point made in my earlier letter and many times in this one, that O'Neill and his supporters enormously underestimate the costs, in both time and money, of turning laboratory-scale inventions into industrial-scale technologies - costs certain to be far greater in the unfamiliar, radically different, and highly dangerous environment of space.

42. As above.

(9) Problems of rain and of dynamic balance in large cylinders. The largest cylinders will have atmospheres of radially varying density as well as clouds, so that rain may well be obtained much as it is on Earth.43 If that is inconvenient, we will use sprinklers and give everyone a constant blue sky.44 It may seem that a total lack of rain would be unnatural and even frightening. But to residents of southern California or Arizona, it is everyday reality. It all depends what you are accustomed to.

43. The conditions which generally produce rain on the earth's surface - variations in terrain, very large air masses of different temperatures, presence of microscopic dust particles (apparently needed for raindrop formation) etc. will not exist in the proposed space cylinders. By keeping the air at something close to 100% relative humidity it might be possible to cause a little rain to fall when "night" cooled down the cylinder. But this would not be a pleasant environment to live in.

44. For years scientists have written that the blueness of our "sky" is caused by the diffraction of sunlight passing our upper atmosphere. No comparable conditions will exist in a space cylinder. People looking up will see one of two things: a) other parts of the cylinder on which people are living, or b) windows, through which they will see the sun (if it is being reflected in), perhaps some stars and otherwise the black of space. In any case it will be clear to all that they are not in an environment similar to earth's surface, but inside a large cylinder (or sphere, or whatever).

As for dynamic balancing, that is readily accomplished by pumping quantities of water between holding tanks strategically placed, 45 Of course, it may be that the cylinders will be so massive that the mass-shifts will be small indeed. If we think of the rocking of an aircraft carrier as the men go about their duties, we may have the idea.46

45. As an old submariner, let me say that the balancing of a body in unstable equilibrium is not "readily" accomplished. But ballast tanks, as Heppenheimer suggests, might well do the job, though the mechanisms to do this would be large and complicated.

46. Yes, but if the sun were being reflected into the colony, that sun would be rocking in the "Sky," and it shadows would be moving back and forth on the ground. Some might find this interesting or pleasant, others not; it would surly be unlike earth.

(10) L5. L5 is a point of stable equilibrium, and not unstable equilibrium, in the restricted three-body problem. When one considers perturbations due to the Sun, it is found that there exists a stable orbit about L5. If the colony is moved slightly, it will not depart from this orbit, but will instead remain close to it.47

47. Perhaps; I will have to have this confirmed by skeptical astronomers. For the moment, it seems that if L5 is a point where the gravitational fields of earth and moon cancel each other out, any movement toward either earth or moon would lead to further movement in that direction, there being no correcting or opposing force. The effect of these forces might be very slight, so that we could say of a 64 million ton cylinder that it would take many thousands or tens of thousands of years before it finally reached the earth. Still, it would be rather hard for those on earth when it did get there.

(11) The last page or so of the paper. These points are largely theological,48 reflecting bias or intuitive dislike, rather than any semblance of reasoned assessment. It may make the writer feel good to turn up his nose and say, Ugh, I would never want to live there! But this is a poor basis for policy. 49 Certainly, one should not seek to deny others the possibility of what to them may be an important and exciting new type of life, merely because one would not himself choose that life. In this country, we provide for some people to live aboard aircraft carriers, others at Army bases, and still others along the Alaska pipeline, not from esthetic judgments as to whether we ourselves might like such a life, but from considerations of national needs.50 It can scarcely be denied that large numbers of people will freely volunteer to live in space, even under austere conditions, when this becomes possible.51 If it is in the national interest that they do so, then esthetic judgments lose much of their force. 52

48. Again, "theological." My objections to this project are variously ethical, moral, philosophical, political, and economic. (I might add that, according to Gerard Piel, publisher of Scientific American, many scientists themselves oppose this project on moral grounds). To call such objections "theological" is imprecise, and has in it more than a whiff of Dr. Strangelove, or hard-nosed talk about "megadeaths" or "credible first strike capability" or "acceptable risks." And this may be the point to note that in all of O'Neill's and Heppenheimer's talk about space colonies there is no mention of risks. The risks would in fact be enormous. We have already lost three lives in space, and almost three more; the Russians have lost at least three. This is a death rate of something over 6%. But our ventures into space have been very modest, and surrounded by the most elaborate and expensive precautions. It seems altogether reasonable to assume that if we begin complicated mining and industrial operations on the moon and in space, our casualty rate will be even higher, perhaps much higher.

49. Not at all. "Do not do unto others as you would not have them do unto you" is a very good basis for policy, which would have spared us one major and recent national disaster.

50. Here the iron hand begins to slip a bit out of the velvet glove. So we "provide," do we, for people to live on aircraft earners, army bases, and in other unpleasant and dangerous places. How do we do that? How in fact do we filI the ranks of the Armed Forces? To a large degree, poverty, unemployment, and boredom do the job; when that is not enough, we conscript what we need, without worrying much about "esthetic judgments." The model of an aircraft carrier is not a bad one, to give us an idea of what space colonies will be like. A few people in them will live like Admirals; most will live like enlisted men. The space colony, as O'Neill himself has suggested in a recent (Jan. 18, 1976) article in the New York Times Magazine, will be an ultimate company town, and history can tell us a good deal about company towns - particularly those on which the workers were not free to leave. If "national needs" (i.e. the interests of powerful pressure groups) dictate, we will find plenty of poor people to draft for the menial and dangerous work of colonizing space.

51. I do deny it - unless, of course, they have been told terrible lies about what life and work in space is really like. I expect that this will happen, and in fact is happening, and it is one of my ethical and moral reasons for opposing this project.

52. Oh, indeed they do!

To sum up: We are not claiming that we have found the solution to the energy crisis, or that we have discovered the future human destiny. We are not claiming the right to unlimited funding, 53 or for space colonization to be declared a major national priority. 54

We cannot responsibly propose that it shall be national policy to undertake space colonization, at a cost of $100 billion. 55 But we are quite prepared to propose that it should be national policy to support the study of space colonization, at a cost of some $1 million. 56 Our proposals are in many ways new, and to some they may be disturbing. But if these further studies confirm present indications as to the feasibility and economic merit of space colonization, then on that basis we may lay this as an issue upon the national agenda.

53. - 55. This is disingenuous. O'NeilI, in his articles in the mass media, has said over and over that we can have a space colony operating in fifteen years for 100 billion dollars, and that we should do so. There is nothing tentative or modest about his way of saying this; he is actively lobbying for such a project. Thus, from his article In the N.Y. Times Magazine of Sunday, Jan. 18, 1976, we have: . . . it appears on the basis of technology being developed for the space shuttle that construction of a high-orbital facility could begin within 7 to 10 years and that it could be completed in 15 to 25 years.... The levels of technology required to do all of this have already been achieved. (Emphsis mine) . . . If the concept is realized as soon as is technically possible, something like the following ''letter from a space colonist" might be written as early as the l990's. (There follows a fictional "letter" which is pure sales promotion, speaking as it does only of delights, saying nothing of difficulties, let alone dangers).

56. A case might be made for spending $1 million for such a study, as long as people representing my point of view - call it "pessimistic theology" if you like - are a part of the study. But we need adversary proceedings here. We need, not just people who say, "How can we figure out how to do this?", but people saying, "This is not worth doing even if we could do it." The almost certain danger is that the million dollar study will lead to another, and another; to a five-million dollar study, then a ten, and so on. Once this bandwagon, this permanent WPA for the Aerospace industry, gets rolling, it will be very difficult to stop it, and the further it rolls, the more difficult, as more and more people have a vested interest in keeping it going.

Let me close by referring once again to my statement that Earth's problems must be solved on Earth. This is not a statement of theology, or even philosophy, but a matter of hard politics and economics. To colonize space to the extent that O'Neill has proposed, to a point where it would make a difference to our population problem, will take, in all probability, fifty to a hundred years or more. These must be years of peace and relative economic stability and prosperity. It would take nothing larger than a minor war or major depression to put a stop to the project, and a major war would of course stop it indefinitely. Beyond that, it is hard to imagine a time in the next fifty or more years when a given amount of money, spent on dealing with Earth's problems, on Earth, would not bring vastly greater benefits than a similar amount spent in space. Thus it would have taken, and would now take, probably less than a billion dollars of research and development in the direct conversion of solar energy to electricity to make that form of power economically competitive over most parts of the world. We have the needed collectors, and need only find a way to mass produce them. In the same way, less than a billion dollars worth of research and development in wind power would probably be enough to make New England, and other comparably windy sections of the world, self-sufficient in energy. And it would take much less than a billion dollars worth of research and development of the kind the New Alchemists are doing at Woods Hole to give to large numbers of the world's poor the means, with very little capital and materials ready to hand, to double or triple their food production. Had we done such research earlier, we would now be energy self-sufficient, instead of at the mercy of the oil producing countries; we would be prospering, instead of being in a deep and probably lasting depression; and we would be in a position, as we are not now, to make a direct, immediate, and important contribution toward dealing with the problems of the world's poor. All this we can still do, and in a short space of time, if we choose. To choose instead to spend billions of dollars in the way that O'Neill, Heppenheimer, and many others suggest, seems to me in the highest degree impractical, wasteful, and immoral. Someday, in a world where mankind has learned how to put a stop to war and to feed all the world's people, to try to colonize space might be a practicable, fitting, and even worthy enterprise. But not now.

Note: Heppenheimer has done a book, Colonies in Space (1977. 224 pp.), $12.95 from Stackpole Books, Box 1831, Harrisburg', PA 17105. Holt also is doing a Space Colonies book. -SB

Investment space