The barometer is an instrument used to measure the weight of the atmosphere and its variations, and which marks the changes in the weather. See Atmosphere and Weather.

This word is made up of βάρος, weight , and of μέτρον, measure . The barometer is usually confused, however inappropriately, with the baroscope; this latter, however, as the word indicates, only marks the changes in the weight of the atmosphere; the barometer not only marks these changes, but measures them as well. See Baroscope.

The barometer and its uses are based on the experience of Torricelli, and is thus named by its inventor Torricelli. One takes a tube of glass filled with mercury, of which tube one end is hermetically sealed, and the other, open end is plunged into a small vessel filled with mercury; when the weight of the atmosphere diminishes, the surface of the mercury at the lower end, on which the air presses, is less under pressure; thus, the mercury within the tube descends; and if on the contrary the weight of the air increases, the mercury rises; for the column of mercury suspended in the tube is always equal in weight to the weight of the atmosphere that presses from above, as is demonstrated in the article Torricelli. [1]

In this explanation we suppose that the pressure of the air comes solely from its weight, which consists of the upper parts on the lower parts. However it is certain that several causes combine to alter the pressure of the air: in general the immediate cause of the pressure of an elastic fluid like the air, is the elastic nature of that fluid, and not its weight. One should not then attribute the suspension of the mercury in the barometer to the weight of the air, even though this weight is the principal cause of the air pressure. In effect, the mercury of the barometer holds itself just as well in a completely sealed room as in the open air, because the air in this room, while it does not bear the weight of the atmosphere, is compressed in the same manner as if it did. If the air remains of the same weight, and the compression of its parts increases or diminishes through some incidental means, then the mercury will descend or rise in the barometer , although the weight of the air has not been increased. Traité des fluides, Paris 1744, p. 61. [2]

There are different sorts of barometer , of which we will list here the principal types.

Common barometer . The construction of the common barometer is as follows. A tube of glass, hermetically sealed at its upper end, is filled with mercury, the diameter being around 1/10 of an inch and its length being at least 31 inches; one fills this tube in such a manner that no air remains mixed with the mercury, and so that no other matter attaches itself to the walls of the tube. To do this successfully, one can use a glass funnel ending in a very narrow tube, and fill the tube by means of this funnel.

One can then get rid of the air bubbles in two other ways: the most common is to fill the whole tube with quicksilver, except for about an inch that is left full of air, one then closes the opening of the tube with a finger, reverses the tube, and by allowing the bubble to move up the tube, it gathers all the tiny imperceptible air bubbles, after which one fills the tube. Mussch. ess. de Phys . [3]

The other method consists of heating a tube that is nearly full over a brazier covered with ashes; this is turned continuously, and the heat expanding the air bubbles makes them escape through the open end.

When the tube has been filled to the rim, one closes the opening tightly with a finger, in such a way that no air can be introduced between the finger and the mercury; then, one plunges the tube into a vessel filled with mercury, but in such a way that the tube does not touch the bottom of the container: at the distance of 28 inches from the surface of the mercury, one attaches two bands divided into 3 inches, and these inches are subdivided into a certain number of smaller parts; finally one attaches the tube to a wooden board, to prevent it from breaking: one leaves uncovered the vessel into which the tube has been plunged, or if one wishes one covers it, so that no dust can enter, and the barometer is completed.

In place of plunging the tube into a vessel, one can often achieve a satisfactory result by curving the end, in such a way that the tube has two vertical branches, of which one is much shorter than the other, and ends in a sort of very large funnel, which is found filled with mercury, on whose surface the atmosphere presses, and knows to raise or lower the mercury in the tube in a manner that is much more sensitive, when the variation of the weight of the atmosphere is greater. This is the simple or ordinary barometer . See Plate Pneumat. figure 1.

One has often sought whether it would be possible to make the variations of barometers more sensitive, with the aim of being able to measure the pressure of the atmosphere with greater precision; this has led to a great number of barometers of different constructions, such as the wheel barometer , the diagonal barometer , the horizontal barometer , etc .

Descartes, and followed by Huyghens, used a tube AB , ( figure 2. ) closed at A , and having a part CD wider than the rest; half of the section CD , the same as the upper part of the tube, is filled with water; and the other half of CD , the same as the lower part of the tube, is filled with mercury. It is true that in this type of barometer , the suspended column being wider, renders the variation more detectable: but the air held in the water evaporating by degrees, fills the empty space in the upper part of the tube, and thereby renders the machine defective. Hence Huyghens imagined it would be better to place the mercury and the water in the barometer , in the following manner: ADG ( figure 3. ) is a curved tube hermetically sealed at A , and open at G ; the cylindrical vessels BC and FE , are the same, and about 29 inches apart from one another; the diameter of the tube is about a twelfth of an inch; [4] that of each vessel is about 1½ inches, and their depth is around an inch: the tube is filled with mercury, which is suspended between the vessel FE and the vessel BC , the space remaining up to A empty of air and mercury: finally one adds plain water mixed with a sixth part of aqua regia [5] (so that the water does not freeze) in the tube EFG , in such a manner that it partly counterbalances the mercury CDF . Hence, when the mercury rises in tube AD , above the level of the mercury held in FE , this mercury in rising reaches equilibrium with the atmosphere; if the pressure of the atmosphere increases, the column of mercury increases, consequently the water will fall; if the atmosphere presses less, the column of mercury will fall, and the water will rise. In this way this barometer indicates the smallest variations in the air far better, than the common barometer ; for in place of two inches, the liquid can vary much more; this comes as much from the circumference of the cylinders in relation to the tubes, as from the weight of the water, which is less than that of the mercury, and hence whose variations are consequently more noticeable; for 14 inches of water are equivalent to one inch of mercury. By increasing the diameters of the cylinders, the variation is even more detectable. However, there is still the inconvenience, that the water will evaporate, and render ineffective the variations, although one can to some extent prevent evaporation by putting a drop of oil of sweet almonds on the surface of the water.

But this drop of oil produces another inconvenience, for it will attach itself to the walls of the tube, and as a consequence after the water has passed it, and sometimes broken it down, it makes the tube opaque.

The greatest failing of all is caused by heat and cold, which makes the fluid in tube EFG act like that in the bulb, and the tube of a thermometer. In effect, this fluid is rarified by heat, and condensed by cold; whence it follows that the height of the water varies by the heat alone, and consequently makes the mercury vary, with the effect that the variations in this type of barometer are more often the effect of heat rather than the pressure of air.

Some have recently tried to make these barometers simpler, by substituting spirit of wine for the water, and spheres for the cylinders, but spirit of wine is very subject to evaporation and expansion from the heat; and therefore changing the cylinders into the shape of pears, prevents the making of true scales. For the rest it is apparent that the movement of this barometer is opposite to that of the ordinary barometer ; while the mercury falls in the latter, the water and the spirit of wine rise in the other, and reciprocally. Mussch .

Thus the shortcomings to which this barometer can be subject, have obliged some others to have recourse to the horizontal or rectangular barometer ABCD ( figure. 4. ). This barometer is formed in such a manner that the branch BC is vertical, and the branch CD is horizontal. It is joined at the end of the perpendicular branch to a vessel AB , and and the variations are marked on the horizontal branch CD : thus the interval or space of variation may be extended so that one may see it; furthermore the smaller the tube BCD is in relation to vessel AB , the more the variations of the mercury in tube AB , will cause the mercury in part CD to vary; and as a consequence the smallest variations will be very detectable. The diameter of tube CD being given, it will be easy to determine the diameter of vessel AB , so that the parts of the horizontal scale in tube DC , corresponding to the parts of the scale of vessel AB , may be as wide as one would wish, and having among themselves the same proportion as the parts of the scale in vessel AB , so that the diameter of the vessel is to that of the tube in the ratio of the square of the parts of their scales; likewise, the diameters of CD and AB having been given, as well as the height of the mercury in the vessel, the height of the mercury in the tube is found by this proportion; as the square of the diameter of the vessel is to the square of the diameter of the tube, thus the parts of the scale of the mercury in the tube, are to the corresponding parts of the scale of the mercury in the vessel. [6]

The construction of this barometer , like that of the barometer of Huyghens, is based on a theorem of Hydrostatics; which is, that fluids that have the same base, weigh in proportion to their perpendicular height, and not from the quantity of their matter; thus the same weight of the atmosphere supports the quicksilver with which tube ACD and vessel AB are filled, as it would have supported the mercury in the single tube ABC . See Hydrostatics. This barometer also has great defects.

For, in the first place, air introduces itself between the particles of mercury in tube CD , and divides them when the tube is too wide. To remedy this inconvenience, one makes the part CD only one-twelfth of an inch in diameter, or even less, one takes care that this little tube is new and very clean, and one uses mercury that has been well purged, with the aid of fire, of all the air it contained; despite all this, the mercury comes out with the weather within from air that comes in, which very often produces some separation between the parts of mercury, so that it moves from D towards C , or at least it forms itself into small globules, which stop here and there in the anterior part of the tube that is empty.

There is another, much more considerable defect, which arises from the major friction of the mercury against the glass, and which prevents this barometer from being much more sensitive than the ordinary barometer . In effect, competent observers have assured us they have often noticed that if the mercury rises or falls by a half of a twelfth of an inch, or an entire twelfth of an inch in the ordinary barometer , it still remains in the same place in the tube CD ; but if the variation increases in the ordinary barometer , a very large movement occurs in the tube CD , such that the operation of this barometer is much less controlled than that of the ordinary barometer . Mussch .

These reasons are why many people prefer the diagonal barometer , in which the space for variation is much larger than in the common barometer , and of which they believe the variations are more regular than those of the others. For this purpose Chevalier Morland [7] has conceived of an inclined tube BEC . ( figure 5. ) for it is evident that if mercury rises to the same height in an upright barometer , and in a curved barometer , the variations will be much more detectable in the inclined tube B E C , than if the tube were vertical, and being the more detectable, the more the tube is inclined, because the mercury, in order to rise, for example, one-twelfth of an inch on the perpendicular, will have three- or four-twelfths of an inch or even more to run along the length of this tube. This discovery is however subject to several inconveniences; for the surface of the mercury in the tube B E , is not parallel to the horizon, but it is convex and inclined; hence with this in place, it is difficult to know at what point one should fix the height of the mercury. In addition the angle at B , makes the surface of the tube very rough at that point, and the unevenness of the surface producing a resistance to the fall or the rise of the mercury, the variations of this barometer are not as prompt as they should be. This last inconvenience is even greater, that the tube B E C makes a large angle in B ; thus the sensitivity, that is to say, the variations of this barometer , is then compensated by their slowness. Mussch .

Wheel barometer : this is an invention of Dr. Hook, [8] that makes the variations in the air more detectable; it consists of a common vertical barometer , to which one adds two weights A and B ( figure 5. ) hung on a pulley, of which one is open to the air, and the other resting on the surface of the mercury in the tube, rises and falls with it. Weight A communicates its movement to the pulley, and this pulley has around its pivot a long needle L K , which shows on a large circle graduated M N O P , the variations in the height of the mercury in the barometer . In addition, the barometer tube is topped by a large globe A B , and the small ball B , which is open to the air, is nearly equal in weight to ball A . As the globe A B has a much greater diameter than that of the tube, an inconsiderable lowering of the mercury in this globe, can make the mercury in tube F A rise, to the height of three inches. Now let us suppose that the entire circumference of pulley F D is three inches; it will therefore make one turn while the mercury rises or falls by three inches, so that the needle L K will also make one turn; and if the diameter of the circle M N O P is one foot, the mercury will not be able to fall or rise by three inches, unless the needle moves by around three feet. This barometer shows well enough the considerable variations in the height of the mercury: but as soon as the mercury falls or rises in tube A F , and it has consequently only just begun to become a little convex or a little concave, the small ball A does not have enough movement to turn the pulley S D a little, because this pulley is subject to some friction on its axis: this prevents it from perceiving insignificant variations in the height of the mercury: but when the pulley begins to move, its movement is greater than it should then be. Here, without doubt, is an inconvenience that cannot be remedied without a great deal of trouble. This barometer is yet subject to other inconveniences that have been noted in the Philosophical Transactions, n. 185. p. 241. such that no-one uses it. Mussch .

Conical barometer : this is a machine more curious than useful. It consists of a conical tube set vertically, of which the upper end, and which is the smaller, is hermetically closed. This barometer has no vessel or basin, its conical shape supplanting here, provided that the lower end of this tube has a very small diameter: for then the mercury supports itself in the tube, being held up by the particles of the air, as by a solid piston or a base. When this tube is filled, if the mercury supports itself there, its weight is equal to the weight of the atmosphere; and if the atmosphere varies, the mercury will rise or descend. Thus when the weight of the atmosphere increases, the mercury is pushed into the narrowest part of the tube; and by this means the column is extended, and its weight is increased. To the contrary, when the atmosphere decreases, the mercury lowers itself in the widest part of the tube; and by this means its column is shorter, and consequently its pressure is weakened.

To make this more intelligible, let us suppose that this barometer is represented by tube A B ( figure 6. ) which is conical, and that this tube being reversed, is filled with thirty inches of mercury from A to C ; and as the variation of the mercury in this barometer is from thirty to twenty-seven inches, let us suppose that the same quantity of mercury A C in the lower part of tube D B , has height D B of twenty-seven inches; it is then certain that while the mercury will find itself in the ordinary barometer at the height of thirty inches, the mercury in tube A B will occupy space A C ; and when the mercury will be in the barometer at a height of twenty-seven inches, the mercury in the tube will occupy space D B ; thus the variation of the mercury in the barometer will be from A to D , which is a space of nearly thirty inches, while this variation will be only three inches in an ordinary barometer . This barometer is the invention of M. Amontons. [9] Mussch .

The inconvenience of this barometer is that is that to prevent the mercury and the air from changing place, and mixing together, the internal diameter of the tube must be very small; and this small size makes the friction of the fluid so notable, that it can prevent the fluid from moving freely: thus this instrument is of no use whatsoever except to Mariners who do not look at it so closely, and who have used it for thirty-five years, because it is so handy. In effect, it is enough to invert it when one wishes to store it; and when one wants to know the air pressure, it is enough to take the tube in the hand, and hold it vertically. To prevent the mercury from coming out of the bottom, as could occur during violent movements of a vessel, one places beneath the tube, near to B , a bit of cotton through which air can freely pass; and if by some accident a little mercury should happen to fall from tube A D , it is enough to turn the tube over; and what had fallen out will in this way, rejoin the column. There is yet another barometer used by Mariners. This barometer which was also invented by Dr. Hook, so as to be used at sea, where the rolling of the vessel renders others impracticable, is nothing other than a double thermometer , or two tubes half-filled with spirits of wine, of which one is hermetically sealed at both ends, and encloses a certain amount of air; and the other is closed at one end, and open at the other. Hence the air, as one knows, acts upon the spirits of wine, and makes it rise for two reasons; by its own gravity, as in Torricelli’s tube; and by its heat, as in the thermometer . If the two tubes are then divided by degrees, in such a fashion that they matched one another at the time that the air was contained therein, it follows that when they match one another again, the pressure of the air will be the same as it was when the air was sealed in. If in the thermometer that is open to the air, the liquid is higher, in considering during the same weather how the other rises or falls through the operation of heat or cold, one will see that the air is heavier; to the contrary, when the open thermometer is lower by comparison to the other, the air is lighter than during the weather in which the instrument was divided into degrees. But it must be remembered that the condensation and rarefaction of the air, on which above all this machine is based, does not depend solely on the weight of the atmosphere, but is also caused by the action of heat and cold. This is why this instrument cannot be called a barometer , but more an instrument that indicates changes in the air. See Manometer.

However this instrument is regarded as being very good for knowing if the weather will be bad, just as with changes in the winds, and the coming of cold. Philos. Transact. nº. 429, p. 133 .

The static barometer , as used by Boyle, [10] Otto de Guericke, [11] etc. is defective, as much by the action of heat, as because it is not very precise and not handy: it consists of a fairly large glass bottle, held in balance by a copper weight, in balance pans that are very light: these two bodies being of equal weight, but of unequal volume, if the medium or fluid in which they weight equally is changed, the change in their weights will follow; in the sense that if the air becomes heavier, the larger body will become lighter in appearance, because it will lose more of its weight than the smaller, which is the denser: but if the medium is lighter, then the larger body will overcome the smaller.

Phenomena of the barometer . These phenomena are different, and authors are no more in agreement on their causes, except on the use to which one can put them to predict changes in the weather. At the top of Mount Snowdon in England, which is 1240 toises [12] high, [13] Doctor Halley found the mercury three and eight-tenths of an inch lower than at the foot; [14] from which it appears that the mercury falls 1/10 inch for every thirty toises. Derham made parallel experiments on the height of the mercury at the peak and at the foot of this mountain, [15] and believed that a perpendicular rise of 32 toises was required, for the mercury to fall by 1/10 of an inch: whereby this author believed that one could derive not only the height of the atmosphere, but also a method by which to measure the heights of mountains. Following this author, if the height of the mercury at this low level is 30 inches, then at an elevation of 1000 feet, it would be 28 9/10 inches; at 2000 feet, 27 86/100; at 3000 feet, 26 85/100; at 4000, 25 85/100; at 5000, 24 93/100; at one mile, 24 67/100; at two miles, 20 29/100; at five miles, 11 23/100; at ten miles, 4 24/100; at fifteen miles, 1 60/100; at twenty miles, 0 95/100; at thirty miles, 8/100; at forty miles, 12/1000. But one supposes in this calculation that the atmosphere is everywhere of more or less the same density, and that if one divides it into portions of equal height, the weight of these portions is almost the same, which is very far from being true; for the atmosphere becomes continuously less dense to the extent by which one moves away from the earth, and thus the same quantity of air occupies a greater and greater volume. It is because of this that if one divides the atmosphere into a series of bands all of equal height, these bands will weigh the less, the further they are from the center of the earth. M. Mariotte, in his essay On the nature of the air , [16] gives a calculation of the height of the atmosphere, based on observations of the barometer made on mountain tops. This calculation is based on the principle, that the air condenses because of the weight it bears; the author derived 15 leagues as the height of the atmosphere, which is also quite close to the result derived by M. de la Hire on the theory of twilight. [17] M. Mariotte [18] also adds to his calculation an essay on the method for determining by the same principles the height of mountains: but today, one generally regards all these methods, as more curious than reliable and useful. See Atmosphere.

One has found that the greatest height of the barometer in London, was 30 3/8 inches, and its largest dip was to 28 inches; at the Paris observatory, its greatest height was 28 4/10 inches, and its least was 26 4/10 based on the Paris foot, which is 9/144 larger than that of London: these observations accord with those made by M. Wolf at Hall in Saxony. At Algiers the mercury rises to 30 2/10 or 3/10 inches on account of the north wind, inasmuch as this wind is often accompanied by rain and high winds. It is true that there has been one occasion on which the height of the mercury greatly exceeded these numbers; the mercury being perfectly purified and suspended in a tube, in the manner of Torricelli, rose to a height of 75 feet, although at the slightest shaking it fell to the normal height. This phenomenon has caused no little embarrassment as long as there has been the question of discovering its cause. Here is the explanation given by M. Musschenbroek in his Essays on Physics . Once one has purged the mercury of the air it contained, it becomes a much denser body than it was when air was mixed in with it: this mercury can then also attach itself very tightly to the surface of the glass; which makes its particles stay suspended there: and as these particles attract one another very strongly, they support their neighboring particles, and by this means the mercury remains suspended to a very great height: but if one shakes the tube, then the particles of mercury that were next to the glass become detached, and everything falls down. One can see in the cited work the more detailed explanation of this singular phenomenon, and the refutation of all the other hypotheses that have been imagined to explain this.

M. Boyle notes that the phenomena of the barometer are so variable, that it is extremely difficult to provide general rules about its rising, or its falling. It however seems that there is a general rule, that when the winds blow from low to high, the mercury is lower; but this is not always true. The illustrious M. Halley has given us the following observations. In calm weather, when it should rain, the mercury is commonly low; and it rises when the weather should be fine. When there should be high winds accompanied by rains, the mercury falls more or less low, depending on the wind that blows. All things equal, the highest level of the mercury comes when the wind blows from the east, or from the north-east. After the wind has blown violently, the mercury that during the time the wind blew was very low, rises rapidly. During calm weather, during which it freezes, the mercury holds itself high. In places most exposed to the north, the mercury undergoes greater variation than in exposed places in the south: in Naples it rarely varies by more than an inch; while at Upminster it varies by 2 5/10 inches, and at Petersburg by 3 31/100. Phil. Transact. nº 434, p 402 . Within and near the tropics, the mercury varies only little, or not at all.

Doctor Beal remarks, that all things being equal, the mercury is higher in the winter than in the summer, and ordinarily in the morning than at midday; that it is also so during a period of fine weather a bit more than before, or after, or when it rains; and that it ordinarily falls lower after rain more than as before: if it should happen that it rises after it has rained, this is ordinarily an indication of fine weather. However, great changes in the air do not happen without the barometer varying detectably.

In connection with the use of the barometer , an able Physician has noted that with its help, we recover the knowledge that animals have, and that we have lost, because our bodies are never exposed to the air like theirs: and because we give ourselves over to intemperance, and we corrupt the sensitive nature of our organs. By reference to the predictions of barometers , the above-cited M. Halley has found that the rising of the mercury predicts fine weather after a storm, and that the wind will blow from the east or the north-east; that its fall marks that there will be winds from the south or the west with rain, or it precedes tempestuous winds, or that both will happen; and that during the storm, if the mercury rises, it is a sign that the storm will soon pass.

M. Patrick notes that in summer the fall of the mercury announces thunder; and when the storm arrives right after the drop in the mercury, it is rarely of any length: the same thing is seen with good weather, that it comes immediately after the mercury rises. And then Derham comparing his observations with those that Scheuczer made at Zurich, on barometers , notes that during the course of the year the mercury varies more at Zurich, sometimes by one and even two inches; and he concludes from this that Zurich is nearly 1/14 of an English mile higher than Upminster. He therefore finds a remarkable agreement between the observations made at Zurich and his own; one of the barometers following very nearly the same variations as the other: however this agreement is not as perfect as that of barometers in places that are closer to one another, such as those of London, of Paris, etc . [19]

Causes of the phenomena of the barometer . The hypotheses by which one has sought to explain the phenomena of the barometer are almost infinite. It is true that the weight of the atmosphere is generally regarded as the principal cause of the movements of the barometer , and the changes in the air as the incidental cause; but this opinion is not universally held. One learned author, for example, regards the changes in the barometer , as being caused by cold and heat: he says he has often observed that during storms, and when the mercury is low, it divides itself and pushes up particles, which he calls kinds of membranes or skins; and he maintains that every time the mercury falls, it is more or less disengaged from these membranes: and that in this movement the bits of mercury are rejoined together; and it is for this reason that it falls; and that also then there escape small particles of air, that were held in the mercury: and that in raising themselves to the upper part of the tube, they force the mercury to fall, the columns being shortened by the exit of these particles; and by their position in the upper part of the tube: this is why, he adds, the mercury rises in very cold weather to the same height as it does in very hot weather, between the two tropics, because it is in its natural state: and it falls during intermediate levels of heat and cold; because it is tightened, and its parts are repressed and compressed together. But this feeling does not produce a very likely reason for the phenomena.

The variations of the atmosphere should be regarded as the cause of those of the barometer : but it is not easy to determine whence come these variations in the atmosphere, and hence it is difficult to find a single principle in nature to which one can relate variations that are so great and so irregular. It is probable that the winds that blow from this or that place cause them, just as do the vapors and exhalations of the earth: changes in the air in neighboring regions, and even the ebb and flow that the moon causes in the air, could contribute equally.

This final cause should certainly be included among those that produce the variations in the barometer : but its effects should not be very considerable in this respect; although the action of the moon raises the waters of the Ocean to a very great height. Here is the reason for this difference: let us suppose that the water in the open sea rises to a height of 60 feet from the action of the moon: when one replaces the Ocean with the atmosphere or such other fluid as one wishes, it is certain that it should rise to nearly the same height, as the atmosphere being less dense than the Ocean, it will have, in truth, a lesser mass to move, but also the force that agitates this mass in drawing together each of its parts, will also be less for the same reason. The air will then climb 60 feet while rising, and will descend below its natural height by a space of 60 feet, that is to say it will vary in height by 120 feet in all. And yet, mercury being 11000 times heavier than air, a variation of 120 feet in a column of air, will not move the mercury by more than around two-tenths of an inch. This is near the amount by which one finds it should rise at the equator, on the supposition that the east wind blows there at 8 feet a second. As there is an infinity of other causes that make the barometer vary, it is not surprising that no-one has distinguished the small variation that the action of the sun and the moon can produce there in raising or lowering the columns of the atmosphere. However it is to be hoped that observers pay attention to this in due course. Rech. sur les vents. Paris 1746. [20]

The learned Halley believes that the winds and the exhalations suffice to produce the variations in the barometer : and following this opinion he has provided a probable explanation: we will give the substance of his remarks on this subject. 1º. It is, he says, the winds that alter the weight of the air in a particular country, and that, whether in bringing together and accumulating a great quantity of air, and in thus burdening the atmosphere in one place more than in another, which happens when two winds blow at the same time from two opposed points; or whether by raising one part of the air, and thereby relieving the atmosphere of one part of its weight, and giving to it the means by which to extend itself more: or finally whether by reducing and sustaining, so to say, a part of the perpendicular pressure of the atmosphere, which happens every time a single wind blows with violence in a single direction: so that as one has experienced that a violent gust of wind, even if artificial, makes the atmosphere lighter, and consequently makes the mercury fall in a tube that is near the place where the wind is blowing, and even in a tube that is at a certain distance. See Philosop. Transactions nº 292 .

2º. The nitrous and cold parts, and even the condensed air in the northern countries, and driven in another place, burden the atmosphere and raise its pressure.

3º. The dry and heavy exhalations of the earth add to the weight of the atmosphere and its elastic force, just like we see the specific heaviness of the menses being augmented by the dissolution of salts and of metals.

4º. The air having been rendered heavier and stronger by the causes we have just reported, becomes more capable of supporting vapors, which being closely mingled with them and floating over them, make the weather fine and calm; to the contrary the air being rendered lighter by the causes opposed to those we have just discussed, becomes unable to support the vapors with which it is loaded, which then coming to throw themselves down, collect in clouds, which in turn become drops of rain. This being so, it appears evident enough that the same factors that increased the weight of the air, and made it more capable of holding up the mercury in the barometer , similarly brought about the good weather and the heat, and that the same cause that makes the air lighter and less able to hold up the mercury, produces clouds and rain: thus, 1º. when the air is very light and and the mercury of the barometer is the lowest, the clouds are very low and move very quickly; and when after the rain the clouds dissipate and the air becoming calm and serene is purged of its vapors, it appears extremely clear, and one can see objects at a considerable distance.

2º. When the air is heavier and the mercury is high in the tube, the weather is calm, although sometimes it can be a little overcast, because the vapors are equally dispersed; if a few clouds then appear, these clouds are high and move slowly; and when the air is very dense and very heavy, the earth is normally shrouded in small thick clouds, which appear to be formed there from the densest exhalations, that the lower air is capable of bearing, which the upper portions of the air could never do, as they are too light for this.

3º. Thus, this is why in England, for example, the mercury is at the highest degree in the coldest weather when the wind is north or north-east, as then there are two winds that blow at the same time, and from two points almost opposite one another; as there is a constant south-west wind, which blows in the Atlantic Ocean at the latitude corresponding to that of England; to which one can add that the north wind brings cold and thick air from the northern regions.

4º. In the northern regions the variation of the mercury is more noticeable than in the south, the winds being more frequent, more violent, more variable and more opposed to one another in the northern countries than in the southern countries.

Finally, it follows from this that within the tropics the variation of the mercury is very little, because the winds there are very moderate, and they ordinarily blow in the same direction.

This hypothesis, while it may appear capable of explaining many movements of the barometer , is not however shielded from all criticism, for 1º. if the wind were the only agent that produced these alterations, there would not be any detectable alteration if there were no wind, and there would never be any wind without a detectable variation in the mercury, which is something counter to experience.

2º. If the wind were the sole agent, the alterations in the height of the mercury would be different in different places on the earth, according to whether or not the wind blew there; thus, that which one tube would lose in London, would be regained by another in Paris, or in Zurich, etc . but according to many Physicists, one notices the contrary; for in all the observations made up to the present, the barometers in different places, they say, rise and fall at the same time, such that there has to have been an equal change in the absolute weight of the atmosphere, that led to the variations. Is this fact really true?

In the end, in omitting every other objection, the fall of the mercury before rain, and its rise after rain, appear inexplicable by this hypothesis; for by supposing there are two contrary winds that chase the columns of air above London, all they could do, would be to cut a certain part of the air above London: consequently it could happen that the mercury falls, but there is no apparent reason for why the rain follows. It is true that vapors can descend, but only to the point that they are in air with the same specific weight as themselves, and having reached the point, they remain there without descending any more. Leibnitz has attempted to repair the defects in this hypothesis, and to provide a new one. He therefore claims that a body plunged into a fluid, only has weight in this fluid when it is supported; so that when it ceases to be so, that is when it falls, its weight ceases to form part of that of the fluid, which by this means becomes lighter. Thus, he adds, the aqueous vapors, when they are supported by the air, increase their weight: but when they fall, they cease to weigh with it, and the weight of the air is reduced; the mercury then falls, and it rains. But Leibnitz’s principle is false, as has been shown by the experiments of Doctor Desaguliers. Furthermore, in supposing that the vapors are forced to descend because of their condensation, and cease to weigh with the atmosphere, they would descend until they arrived at the part of the atmosphere, that is of the same specific weight as themselves, and, thus as we have already said with regard to M. Halley, they would stay suspended as before. If the mercury falls, it would be only during the time of the descent of these vapors; for once these vapors are fixed and in a state of calm, the first weight will be reborn, so to say, or if it does not return, at least the rain would not follow the fall of the mercury.

Some authors, in order to explain these variations, have imagined the following hypothesis. Let one suppose a number of tiny vesicles of water floating in part of the atmosphere, and on a determined part of the surface of the terrestrial globe; for example, on A B, figure 2 1 ; if the higher vesicles are condensed by the cold of the higher regions, their specific gravity will increase and they will descend; the horizontal level 1, for example, will descend to 2, 2 to 3, etc . there meeting other vesicles that have not yet been precipitated, they gather together and change themselves into larger vesicles, as they follow the laws of attraction.

If we choose the wind as the agent, let us suppose that it blows horizontally or obliquely: in the first case vesicles from 8 will be driven against those from 9, those against 10, etc. ; in the second case vesicle 7 will be driven against 4, 8 against 3, etc . by this means the particles will increase and form new, larger vesicles than before, so that their number, which before was, if one wishes, a million, will then be reduced, for example to 100,000.

But the same combination that reduces their number, in some manner augments their specific weight; that is to say, there is more matter beneath equal surfaces: this is easily proven by geometrical principles; for, in the augmentation of the mass of homogeneous bodies, that of the surface is not as large as that of the solid body: that of the first is like the square of the diameter; and that of the other, like its cube.

Hence while the same quantity of matter finds itself under a smaller surface, it should lose less of its weight from the resistance of its surroundings: for it is evident that a body that moves in a fluid, loses some of its weight through the friction of its components against those of the fluid. And yet, this friction is evidently due to its surface; this is why, as the surface becomes less in proportion to the mass, the resistance does also: consequently the vesicles, whose weight, before the combination, was equal to the resistance of its surroundings, finding this resistance lessening, will descend with a speed that is proportionate to the real diminishment of their surface.

When they descend and when they arrive in denser parts of the atmosphere, for example to points 4, 5, etc. , their mass and their surface are augmented by new conjunctions, and thus by new and constant augmentations, they become more and more capable of overcoming the resistance of their surroundings, and to continue their fall through all the levels of the air until they teach the ground, their mass then being excessively swollen, and form raindrops.

Now during the descent of the vapors, one must consider how the barometer is affected by their descent. Before any of the vesicles begins to descend, whether by the action of the cold, or that of the wind, they all swim in the part of the atmosphere A B D C , and all weigh towards the center E . And yet, each of them respectively resting in one part of their surroundings, which has the same specific weight, will lose a portion of its weight equal to that of a part of its surroundings that has the same volume; that is to say, that each one will lose all its weight, but then that weight that they have lost, will be transferred to their surroundings, which will press down on the earth A B with its own weight together with that of these vesicles. Let us suppose then that this joint pressure acts on the mercury raised in the barometer to thirty inches: through the conjunction of the vesicles, done as we have described above, their surface, and consequently their friction, is reduced: this is why they will communicate less of their weight to the air, that is to say, a lesser part than their total weight; and hence they will descend with a speed that is proportional to what remains of their weight, as one has just said. And as the vesicles cannot act on the surface of the earth A B except through the action of the air, their action on the ground will be reduced in the same proportion as their action within their surroundings; from which it is evident that the surface of the earth A B , will always be pressured less than before; and moreover the vesicles will retain whatever of their weight they have not at all transmitted to their surroundings; furthermore they will speed up their own descent; that is to say, that the rate of descent of the vesicles will always be growing: in effect, when the vesicles descend, the mass grows continuously, and by contrast the resistance of their surroundings and the pressure on the ground diminish, and the mercury will consequently fall during the whole time of their descent. From that it is easy to conceive that once the vesicles have begun to fall, they continue to do so; that the mercury will fall at the same time; and that it will both continue, and stop, at the same time as the vesicles do so.

One can present an objection to this system; this is, that the vesicles being set in motion, and bumping into the particles in their surroundings, encounter considerable resistance in the inertial force of their surroundings, by which their descent would be slowed, and the pressure of the atmosphere restored. One can add that the additional pressure will be greater in proportion to the speed of the descent of the vesicles, a strong impulse being required to overcome the inertial force of nearby particles in the surroundings.

But those who are of the opinion that we report, believe they can overcome this objection by means of reason and experience: for, they say, although the inertial force of the air can be very weak because of its low density, we see that in water, which is a medium that is both very dense and non-elastic, a piece of lead, in descending through this liquid, weighs considerably less than when it is at rest. However this fact is negated by M. Musschenbroek. Essays de Physique, § 234.

We believe we had to present at length this explanation that, however ingenious, does not have nearly enough of the precision one could desire. But in so difficult a matter, there was virtually nothing else we could do, than to lay out what the philosophers have thought. See an erudite dissertation, of M. de Mairan, on this subject, Bordeaux 1715. [21] See also Musschenbroek. This author regards, with reason, the predictions of the barometer as not being very certain.

According to M. Musschenbroek, here is the best way to make an ordinary or common barometer ; these types of barometers being the best of all, as he affirms. First, one must take very pure mercury, and be quite sure that it has not been tampered with; one must then pass it through a very clean skin, and place it in a new, glazed pot, which one covers with a top that fits well. One then places this sealed pot on a fire of very clean coal, and brings the mercury to the boil: it then becomes volatile, but is contained by the cover that has been placed over it. By thus boiling the mercury, one purifies it of any water and air held in it. One must have newly made glass tubes, which one uses for barometers ; and so that they may not be either dirty inside, or filled with air, one must take care to have these hermetically sealed at either end in the glassworks, before they are shipped. When one wishes to fill them, one can open one end with a file, and at this time keep them next to an oblong fire, to make them equally hot, and even very hot, so that the humidity and the air that are held within the walls, detach and dissipate. If one neglects to take this precaution, the air will attach itself with such force, that it cannot be chased out by the mercury one places in the tube, but remains suspended in several places. To be even more successful in purging the air from the tube, one can do worse than to attach to a fine wire a scrap of kid or leather, and shape this like the piston in a pump, that passes down the tube from top to bottom, and from bottom to top several times, to remove the air it contains. By this means, the boiling mercury will then dissipate the air, by driving it from the hot tube. One then forms from a large barometer tube a small glass funnel, and in lengthening it one reduces it to a capillary tube, that should be a little longer than the tube one has to fill. One must first clean the upper part of this little funnel very well, and make it very dry and very hot by exposing it to the fire: one then slides it into the tube of the barometer , so that it reaches the very bottom, and then one pours the boiling mercury into this little funnel, which must be very hot, so that the heat of the mercury keeps it from separating into pieces. Once one has poured in the mercury, it falls to the bottom, fills the tube, and slowly rises thereafter. One must take care to pour it into the funnel without any interruption, so that the mercury always falls without stopping, and so that air cannot get into it. Once the tube is full, one gently withdraws the little funnel. In this manner one can fill the tube as exactly as possible, and it will then appear brown in color along its whole length, and without the least tiny air bubble. If one has no sealed tubes, one must first, before filling the tube one will use, clean the inside well, by washing it with well rectified spirits of wine, and attaching to the bottom of a brass wire a little scrap of leather in the manner of a pump piston, which one pushes frequently through the tube to remove the air, which without this would not fail to remain suspended there. After having thus cleaned this tube, one must dry it before the fire, and heat it.

Translator’s Notes

1. Evangelista Torricelli (1608-1647), Italian mathematician, credited with the invention of the first mercury barometer, in 1643.

2. Jean-Baptiste le Rond d’Alembert (1744) Traité de l’Équilibre et du Mouvement des Fluides Pour servir de suite au Traité de Dynamique. Par M. D’Alembert, de l’Académie Royale des Sciences. A Paris, chez David, l’aîné, Libraire, rue Saint Jacques, à la Plume d’or.

3. Petrus van Musschenbroek (1739) Essai de physique avec Une description de nouvelles sortes de machines pneumatiques et un recueil d’experiences par J.V.M. / traduit du Hollandois par Pierre Massuet. Leyden: Samuel Luchtman.

4. “... est d’environ une ligne”. The ligne was a pre-metric French measure of length, defined by the Petit Larousse of 1967 as “Ancienne mesure française de longuer, représantant la douzième partie du pouce, soit 2,25 mm environ”.

5. Aqua regia (“royal water”) is a mixture of nitric acid and hydrochloric acid (in the proportion of 1:3) that can dissolve gold.

6. For a contemporary discussion of the mathematics involved here see Dominique François Rivard (1739) Élémens de Geometrie avec un Abbregé d’Arithmetique et d’Algebrie. Paris: Chez Jean Desaint. Pages cxxxv-cxxxvj.

7. Samuel Morland (1685) Élévation des eaux par toute sorte de machines, réduite à la mesure, au poids, à la balance... Paris: Impr. De G. Martin.

8. Robert Hooke (1665) Micrographia: or some Physiological Descriptions of Minute Bodies made by Magnifying Glasses. With Observations and Inquiries thereupon. London: John Martyn and James Allestry. Figure 1.

9. Guillaume Amontons (1695) Remarques et experiences phisiques sur la construction d'une nouvelle clepsidre sur les baromètres, termomètres, and higromètres. Paris: J. Jombert.

10. Robert Boyle (1660) New experiments physico-mechanicall, touching the spring of the air, and its effects: (made, for the most part, in a new pneumatical engine) : written by way of letter to the Right Honorable Charles, Lord Vicount of Dungarvan, eldest son to the Earl of Corke . Oxford: H. Hall.

11. Otto van Guericke (1672) Experimenta nova (ut vocantur) magdeburgica de vacuo spatio primùm à r. p. Gaspare Schotto ... nunc verò ab ipso auctore perfectiùs edita, variisque aliis exmerimentis aucta. Quibus accesserunt simul certa quaedam de aëris pondere circa terram; de virtutibus mundanis, and systemate mundi planetario; sicut and de stellis fixis, ac spatio illo immenso, quod tàm intra quam extra eas funditur . Amsterdam: J. Jansson.

12. The toise was a pre-metric French measure of length, equal to six French feet [1.949 meters]. It is essentially the French equivalent of the fathom, or six English feet, but was slightly longer, the French foot being equivalent to 1.066 English feet.

13. Snowdon, in northwestern Wales, is currently calculated by Britain’s Ordnance Survey to have a height of 1,085 meters (3,560 feet). https://www.ordnancesurvey.co.uk/blog/2014/10/re-surveying-snowdon/. The height given by d’Alembert, 1,240 toises, would have been equivalent to about 7,930 English feet. The highest mountain in the British Isles is Scotland’s Ben Nevis, measured by the Ordnance Survey at 1,345 meters (4,411 feet).

14. Edmond Halley (1698) “A Letter from Mr. Halley of June the 7th. 97. Concerning the Torricellian Experiment Tryed on the Top of Snowdon-Hill and the Success of It.” Philosophical Transactions of the Royal Society 19, 582-584.

15. William Derham (1698) “Part of a Letter of Mr. William Derham, Rector of Upminster, dated Dec. 6 1697. Giving an account of some Experiments about the Height of the Mercury in the Barometer, at Top and Bottom of the Monument: and about portable Barometers.” Philosophical Transactions of the Royal Society 20, 2-4, published 1 January 1698.

16. Edmé Mariotte (approx. 1684 [1923]) Discours de la nature de l'air. De la végétation des plantes. Nouvelle découverte touchant la vue. Paris: Gauthier-Villars.

17. Philippe de La Hire (1689) Trouver la correction des observations correspondantes devant and aprés midi, pour determiner le vray midi / par monsieur de La Hire . Paris: Impr. d'Antoine Lambin.

18. Edmé Mariotte (1717) Oeuvres de Mr. Mariotte, de l’Académie royale des sciences; divisées en deux tomes, comprenant tous les traitez de cet auteur; tant ceux qui avoient déja paru séparément que ceux qui n’avoient pas encore été publiez; imprimées sur les exemplaires les plus exacts and les plus complets . Leiden, P. Vander Aa.

19. Philosophical Transactions of the Royal Society, Volume 26 (1709): 497-499.

20. Jean-Baptiste le Rond d’Alembert (1747) Reflexions sur la cause generale des vents: piéce qui a remporté le prix proposé par l’Academie royale des sciences et belles lettres de Prusse pour l’année MDCCXLVI / par M. D’Alembert ...; a laquelle on a joint les pieçes qui ont concouru . A Berlin: Chez A. Haude and J.C. Spener, 1747.

21. Jean-Jacques Dortous de Mairan (1715) Dissertation sur les variations du baromètre, qui a remporté le prix à l’académie royale des belles-lettres, sciences et arts de Bordeaux, le 1. de May 1715, 2e éd. revue et corrigée par l’auteur. Béziers: E. Barbut.