Nature Does Not Abhor a Vacuum

Blaise Pascal

Translated by I. H. B. Spiers and A. G. H. Spiers¹

Annotated by Edward Grant

CONCLUSION OF THE TWO PRECEDING TREATISES

I have recorded in the preceding Treatise all the general effects which have been heretofore ascribed to nature’s effort to avoid a vacuum, and have shown that it is utterly wrong to attribute them to that imaginary cause. I have demonstrated, on the contrary, by absolutely convincing arguments and experiments that the weight of the mass of the air is their real and only cause. Consequently, it is now certain that nature nowhere produces any effects in order to avoid a vacuum.

It is not difficult to demonstrate, furthermore, that nature does not abhor a vacuum at all. This manner of speaking is improper, since created nature, which is the nature under consideration, isnot animated, and can have no passions. Such language is in fact metaphorical, and means nothing more than that nature makes the same efforts to avoid a vacuum as if she abhorred it. Those who use this phrase mean that it is the same thing to say that nature abhors a vacuum as to say that nature makes great efforts to prevent a vacuum. Now, since I have shown that nature does nothing at all to avoid a vacuum, the conclusion is that nature does not abhor it. To carry out the metaphor: just as we say of a man that a thing is indifferent to him when his actions never betray any movement of desire for, or of aversion to, that thing, so should we say of nature that it is supremely indifferent to a vacuum, since it never does anything either to seek or to avoid it. (I am here still using the word "vacuum" to mean a space empty of all bodies which our senses can apprehend.)²

It is perfectly true (and this is what misled the ancients) that water rises in a pump when the air has no access to it, that a vacuum would result if the water did not follow the piston, and that the water ceases to rise as soon as any cracks develop by which the air can get in to fill the pump. Thus it looks as though the water rose merely for the purpose of preventing a vacuum, since it rises only when otherwise there would be one.

Similarly it is a fact that a bellows is hard to open when its apertures are so carefully sealed that no air can enter it; and it is true that its opening would produce a vacuum. This resistance ceases when air can enter to fill the bellows, and since it is met with only when a vacuum would otherwise result, it seems to be due to nothing else than the fear of a vacuum.³

Finally, it is a fact that all bodies in general make great efforts to follow one another and to keep together whenever their separation, and nothing else, would produce a vacuum between them. This is why it has been inferred that this close adhesion is due to the fear of a vacuum.

To reveal the weakness of this reasoning the following example will serve. When a bellows is placed in water in the manner we have often described, with its nozzle at the end of a tube assumed to be twenty feet long which projects out of the water into the air, and with all its side apertures sealed so as to exclude the air, everyone knows that it is hard to open, and the more so the greater the amount of water above it; whereas if the vents in one of the wings are unsealed so that they admit the water freely, the resistance disappears.

If one wished to reason out this effect like the others, he might say: When the side vents are closed and when, therefore, if the bellows is to be opened, the air must enter through the tube, there is difficulty in opening it; but when water instead of air can enter to fill it, the resistance ceases. Therefore, since there is resistance only to the entrance of air, the resistance arises from an abhorrence of the air.

There is no one who will not laugh at this inference, seeing there may well be another cause of the resistance. It is evident indeed that the bellows cannot be opened without raising the water, because the water that would be pushed aside in the act of opening cannot enter the body of the bellows, and, compelled to find room for itself elsewhere, raises the whole body of the water and causes the resistance. This does not occur when the bellows has vents through which the water can enter; for whether it is opened or shut the water neither rises nor falls in consequence, since water enters the bellows just as fast as it is pushed aside, and thus offers no resistance to its opening. All this is clear, and consequently we must believe that the bellows cannot be opened without two results: first, the air does really enter, or second, the level of the water is raised. It is the latter action that causes the resistance and with this the former has nothing to do, although it occurs simultaneously.

Let us give the same explanation for the difficulty experienced in opening in the air a bellows sealed on all sides. If it were forced open two things would occur: first, a vacuum would really be formed; second, the whole mass of air would be raised and upheld. It is the latter action that causes the resistance felt; the former has nothing to do with it. This resistance also increases or diminishes in proportion to the weight of the air, as I have shown.

The same facts explain the resistance to separation offered by all bodies between which there is a vacuum: air cannot filter in, otherwise there would be no vacuum, and this being so, they cannot be separated except by raising and upholding the whole mass of the air. It is this which occasions the resistance.

Such then is the real cause of the adhesion of bodies between which there exists a possible vacuum. It was for a long time not understood because erroneous opinions were entertained which were discarded only by degrees. There have been three different periods during which different opinions of this character were held; and these involved three generally prevailing errors which made it absolutely impossible to understand the cause of this adhesion of bodies.

The first is that, in nearly all times, it was believed that the air has no weight; for the ancients said so,⁴ and their professed disciples followed them blindly. They would have remained forever wedded to that theory had not keener thinkers rescued them by the force of experimental evidence;⁵ for it was impossible to believe that the weight of the air causes such adherence, so long as it was held that air had no weight.

The second error lay in the belief that the elements have no weight in themselves,⁶ for the sole reason that the weight of water is not felt by those who are in it, that a bucket full of water immersed in the water is easy to lift so long as it stays there, and that its weight begins to be felt only when it is lifted out. As if these effects could not be due to another cause—or rather, as if this one was not wholly beyond all probability! For there is no sense in believing that water in a bucket has weight when out of the water, but has no weight left after it is poured back into the well; that it loses its weight when mixing with the rest, and recovers its weight when lifted above the surface. Strange are the means men employ in order to cloak their ignorance! Because they could not understand why the weight of water is not felt, and were loath to confess their ignorance, they declared it had no weight, for the satisfaction of their vanity and to the ruin of truth. Their views prevailed; and, of course, the weight of the air could not be accepted as the cause of these effects so long as this vain imagining had currency. Even had it been known that air has weight, the claim would still have been made that it has no weight when contained within itself, and consequently the belief would have persisted that it can effect nothing by its weight. That is why I have shown in the Equilibrium of Liquids that water weighs the same within itself as outside, and I have explained there why, in spite of that weight, a bucket is not hard to raise while it is in the water and its weight is scarcely felt. And in the Treatise on the Weight of the Mass of the Air I have given the same demonstration in the case of the air, to clear up all doubts.

The third error is different in kind. It does not appear in connection with [the weight of] the air, but in connection with the effects which were ascribed to the abhorrence of a vacuum. Concerning these the most erroneous theories were entertained. For it had been imagined that a pump raises water not only to ten or twenty feet, which is true enough, but still farther—to a height of fifty, one hundred, or one thousand feet, or as high as you please, without limit. Likewise the belief was held that it is not only difficult, but actually impossible, to separate two polished bodies in close contact;⁷ that not even an angel, or any created force, could do so, with hundreds of exaggerations which I scorn to repeat. And so with the rest.

This is an error of observation so ancient that it cannot be traced back to its source. Heron himself, who is one of the oldest and best of the authors who have written on the raising of water, states as a positive and uncontrovertible fact that the water of a river may be made to pass over a mountain ridge and to flow into the valley beyond, provided this valley be somewhat lower down, by means of a siphon placed on the summit with its legs stretching along the slopes, one into the river and the other on the farther side; and he asserts that the water will rise from the river over the mountain and drop down again into the other valley, however high the ridge between may be.⁸

All writers on the subject have said the same thing; and even at the present time our fountain builders guarantee that they can make suction pumps which will raise water as much as sixty feet if it be desired.

Neither Heron, nor those other writers, nor the artisans, and still less the natural philosophers, can have carried their tests very far; for had they tried to draw water to the height of forty feet, they would have failed. They had only seen suction pumps and siphons six, ten, or twelve feet high, which worked beautifully; and in all the experiments they had occasion to make, had observed no case in which water failed to rise. They never imagined, consequently, that there was a limit beyond which water behaved otherwise. They conceived that the facts they had noticed were the results of an invariable natural necessity; and since they believed that water rose by an invincible abhorrence of a vacuum, they concluded that as it rose at first, so it would continue to rise without limit, applying their interpretation of what they did observe to what they did not observe and declaring both statements to be equally true.

So positively was this believed that philosophers have made it one of the most general principles of their science and the foundation of the treatises on the vacuum. It is and has been didactically asserted every day in all the schoolrooms in the world, ever since books were written. Everyone has firmly believed it, and it has remained uncontradicted down to our own time.

This fact perhaps may open the eyes of those who dare not doubt an opinion which has always been universally entertained; for simple workmen have been able to prove in this instance that all the great men we call philosophers were wrong. Galileo declares in his dialogues⁹ that Italian plumbers taught him that water rises in pumps only to a certain height: whereupon he himself confirmed the statement as others did also, afterward, first in Italy and later in France, by using quicksilver, which is easier to handle but provides merely several other ways of making the same demonstration.

Before men gained that knowledge, there was no incentive to prove that the weight of the air was the cause of water rising in pumps; since, the weight of the air being limited, it could not produce an unlimited effect.

But all these experiments were insufficient to show that the air does produce those effects: they had rid us of one error but left us in another. They taught us, to be sure, that water rises only to a certain height, but they did not teach us that it rises higher in low-lying places. On the contrary, the belief was held that it always rises to the same height, in every place on the earth. And since the weight of the air never entered anybody’s head, it was vaguely thought that the nature of the pump was such that it lifted water to a limited height and no further. Indeed Galileo took that to be the natural height of a pump, and called it la altessa limitatissima. How indeed could it have been imagined that that height was different in different places? Certainly, it would seem improbable. Yet that last error again put out of the question the proof that the weight of the air causes these effects; since, because this weight would be greater at the foot than at the top of a mountain, its effects, obviously, would be proportionately greater there.¹⁰

That is why I decided that the proof could be obtained only by experimenting in two places, one some four or five hundred fathoms above the other. I chose for my purpose the Puy de Dôme mountain in Auvergne, for the reasons that I have set forth in a little paper which I printed as early as the year 1648, immediately after the experiment had proved successful.¹¹

This experiment revealed that fact that water rises in pumps to very different heights, according to the variation of altitudes and weathers, but is always in proportion to the weight of the air. It perfected our knowledge of these effects, and put an end to all doubting; it showed their real cause, which was not abhorrence of a vacuum, and shed on the subject all the light that could be wished for.

Try now to explain otherwise than by the weight of the air why suction pumps do not raise water so high by one-quarter on the top of Puy de Dôme in Auvergne as at Dieppe; why the same siphon lifts water and draws it over at Dieppe and not at Paris; why two polished bodies in close contact are easier to separate on a steeple than on the street level; why a completely sealed bellows is easier to open on a house-top than in the yard below; why, when the air is more heavily charged with vapors, the piston of a syringe is harder to withdraw;¹² and lastly why all these effects are invariably proportional to the weight of the air, as effects are to their cause.

Does nature abhor a vacuum more in the highlands than in the lowlands? In damp weather more than in fine? Is not its abhorrence the same on a steeple, in an attic, and in the yard? Let all the disciples of Aristotle collect the profoundest writings of their master and of his commentators in order to account for these things by abhorrence of a vacuum if they can. If they cannot, let them learn that experiment is the true master that one must follow in Physics; that the experiment made on mountains has overthrown the universal belief in nature’s abhorrence of a vacuum, and given the world the knowledge, never more to be lost, that nature has no abhorrence of a vacuum, nor does anything to avoid it; and that the weight of the mass of the air is the true cause of all the effects hitherto ascribed to that imaginary cause.

^1. Reprinted with the kind permission of the Columbia University Press from the Physical Treatises of Pascal: The Equilibrium of Liquids and the Weight of the Mass of the Air, translated by I. H. B. Spiers and A. G. H. Spiers, with Introduction and Notes by Frederick Barry (New York: Columbia University Press, 1937), pp. 67–75. The present selection, translated from the original French, constitutes the complete conclusion in a volume by Pascal edited and published in 1663, one year after his death, by his brother-in-law F. Perier. The two treatises in the volume, referred to in the heading of the conclusion, are titled "On the Equilibrium of Liquids" and "On the Weight of the Mass of Air." As Frederick Barry remarks in a foreword (pp. v–vi), "to him we owe the first conclusive proof of the pressure of the atmosphere, and the final banishment from the minds of natural philosophers of that ancient and persistent conception of horror vacui which had long inhibited investigation in this field; a complete mechanical correlation of all the diverse phenomena of fluid equilibrium, without recourse to that or any other imaginary conception; and finally, the establishment of detailed analogies between the effects of pressure in liquids and in air which affected the unification of hydrostatics and aerostatics as one deductively organized and logically coherent discipline." In the preceding chapter of this source book the selections from John Buridan and Marsilius of Inghen typically represent the medieval concept ion of "abhorrence of a vacuum." Even Nicholas of Autrecourt, who argued for minute, interstitial vacua, reaffirmed nature’s abhorrence of a separate or extended vacuum. Pascal’s exposition is so simple and clear that it requires little comment.

^2. The usual assumption made by Aristotle and his followers was that a vacuum, by definition, was absolutely devoid of matter. However, in discussing imaginary void space beyond the cosmos it was assumed that, although devoid of corporeal entities, it was wholly occupied by God or spirit (see Selection 73, introduction). Pascal, here, seems to allow for the possibility that nonperceptible, or non-sensible, matter might, in some way, occupy, or partially occupy, a vacuum.

^3. Compare to Buridan’s discussion of the bellows in the preceding selection.

^4. On the basis of an experiment in which an inflated and uninflated bladder were found to be of equal weight, the Greek commentator, Simplicius, in his Commentary on Aristotle’s De caelo, concluded that air has no weight in air but apparently believed that it would have weight in the natural place of fire. Aristotle (De caelo, 31lb.6–10), however, on the basis of the same experiment insisted that an inflated bladder weighs more than an empty one and concluded that air would have weight in its own natural place as well as in the natural place of fire. The measurements were too delicate for the instruments available and the issue could not be resolved in this manner. See M. R. Cohen and I. E. Drabkin, A Source Book in Greek Science (Cambridge, Mass.: Harvard University Press, 1948), pp. 247–248.

^5. Frederick Barry observes (pp. 27–28, n.3): "The famous Milanese mathematician and philosopher of nature, Girolamo Cardano (1501–1576) appears to have been the first to demonstrate that air has weight and to attempt a determination of its specific gravity. After him, and before it was known how to produce a vacuum, many experiments were made to establish his principal conclusion as indubitable fact. The best of these were carried out by Galileo and described in his Discorsie dimonstrazioni mathematiche (1638), in the dialogue of the First Day. With a syringe he forced into a properly valved bottle, previously weighed, an amount of air which at ordinary pressure would have been of two or three times its volume, and determined with precision an actual increase of weight; then he allowed the excess air to escape without loss through a tube which led into another bottle completely filled with water and so constructed that the water displaced by the entering air could be caught in a suitable weighed vessel, the increase in the weight of which, finally, he determined. Galileo remarked that if air were specifically light, the first flask when filled with condensed air should be lighter than before, not heavier; that air consequently had weight (as, indeed, previous experience had already proved); and that since the volume of the water displaced was that of the excess air introduced, in its normal condition, the ratio of the determined weight of this air to that of the displaced water was the specific gravity sought. In a second, more elegant experiment, he forced water, instead of air, into the valved flask first used and thus compressed the air within it; weighed it in this condition; then opened the valve, thus allowing to escape a volume of air which, in its normal condition, had previously occupied the space now filled by the water; and finally weighed again. In this case, the increase in weight after the water had b een introduced was the weight of this water, the subsequent loss in weight being that of the same volume of air in its normal condition. See the English translation of the Discorsi by Henry Crew and Alfonso de Salvio, Dialogues Concerning Two New Sciences by Galileo Galilei (New York: Macmillan, 1914)."

^6. Pascal is in error here, for Aristotle insisted that earth, water, and air have weight in their own natural places (De caelo, 311b.6–10; fire is an exception, for it was deemed to be weightless or absolutely light). See note 4.

^7. The problem of two plane surfaces in contact was discussed by Galileo (see Selection 53.4 and n. 29 thereto).

^8. Barry remarks (pp. 71–72, n. 3) that the modern Greek edition of Hero’s Pneumatica and the editions available to Pascal contain no such explicit statement and that Pascal either misread Hero or relied on hearsay. "On the other hand, Heron, in discussing the siphon. . . makes no mention of any limitation to its efficacy; neither does Cardan (De subtilitate, 1560: 1.3.364a); and Galileo’s discussion of pumps in the Mathematical Discourses of 1638. . . implies that it was then still taken for granted by the philosophers, if not by all workmen, that a siphon of any height would operate."

^9. See the First Day of Galileo’s Dialogues Concerning Two New Sciences in the translation of Crew and de Salvio, pp. 16–17.

^10. Barry supplies the following note (pp. 73–74. n, 5):

"Pascal here refers, in the first place, to ideas current up to the time of Galileo, which were consistent with the theory of a limitless horror vacui; then to Galileo’s observations on the pump which, though they permitted the retention of the general notion of a horror vacui, nevertheless proved that it was a finite and measurable force. But Galileo did not dogmatically assert, as Pascal implies, that it was an invariable force: it was contrary to his habit of thought, in the absence of experimental evidence, to do anything of the sort; indeed, for lack of sufficient knowledge, he left the phenomenon unexplained. Pascal’s omission of all reference to the several physicists who, after Galileo, quite definitely conceived that this force was external and due in fact to atmospheric pressure—among whom were Mersenne and Descartes, whose ideas were almost certainly known to him—and particulary his failure to mention Torricelli, who verified this theory, leaves a false impression too favorable to himself. It was not his work that invalidated the conception of horror vacui: and his contradictory hypotheses were not only suggested but were almost completely verified by his predecessors. It is sufficient honor to him that he so brilliantly completed their work.

^11. Perier included this paper in a later part of the same volume which contained the conclusion reproduced here. It describes Pascal’s most famous experiment.

^12. In an earlier note (pp. 51–52, n. 1) Barry explains, "although the density of moist air is less than that of dry air under the same conditions, barometric pressures at the same altitude and like temperatures are usually greater when the air is moist; primarily on account of its greater compressibility at ordinary temperatures in this condition, since this may occasion an increase of density that overcompensates, at the moderate altitudes where measurements are usually made, the decrease due to the mere presence of water vapor—especially when, as is usual, the upper air is dry. Perier observed this effect in a long series of observations made with the mercury barometer at one station in 1649–51. . .; and Pascal, in his experiment with the bellows. . .not only noted it, but made it the basis of several acute meteorological inferences. . . . In 1753 Bouguer remarked especially the strikingly greater difference of pressure between two fixed altitudes when the air is moist—an informing fact which had previously been missed by Perier for lack of a sufficient number of observations; but it was not until 1784 that Deluc, as the result of a long series of measurements made under widely variable conditions and carefully compared, finally explained the phenomenon satisfactorily. See Deluc: Recherches sur les modifications de l’atmosphere, nouvelle ed. 1784, III, 715, 716.