hydraulic machine for raising water by means of the pressure of the atmosphere. See Hydrostatics, no. 23, &c.

Naval Pump, a well-known machine, used to discharge the water from the ship's bottom into the sea. The common pump is a long wooden tube, whose lower end rests upon the ship's bottom, between the timbers, in an apartment called the well, inclosed for this purpose near the middle of the ship's length.

This pump is managed by means of the brake, and the two boxes or pistons. Near the middle of the tube, in the chamber of the pump, is fixed the lower box, which is furnished with a staple, by which it may at any time be hooked and drawn up in order to examine it. To the upper box is fixed a long bar of iron, called the spear, whose upper end is fastened to the end of the brake, by means of an iron bolt passing through both. At a small distance from this bolt the brake is confined by another bolt between two cheeks, or ears, fixed perpendicularly on the top of the pump. Thus the brake acts upon the spear as a lever, whose fulcrum is the bolt between the two cheeks, and discharges the water by means of the valves, or clappers, fixed on the upper and lower boxes.

These sorts of pumps, however, are very rarely used in ships of war, unless of the smallest size. The most useful machine of this kind, in large ships, is the chain-pump, which is universally used in the navy. This is no other than a long chain, equipped with a sufficient number of valves, at proper distances, which passes downward through a wooden tube, and returns upward in the same manner on the other side. It is managed by a roller or winch, whereon several men may be employed at once; and thus it discharges, in a limited time, a much greater quantity of water than the common pump, and that with less fatigue and inconvenience to the labourers.

This machine is nevertheless exposed to several disagreeable accidents by the nature of its construction. The chain is of too complicated a fabric, and the procket-wheels employed to wind it up the ship's bottom, are deficient in a very material circumstance, viz. some contrivance to prevent the chain from sliding or jerking back upon the surface of the wheel, which frequently happens when the valves are charged with a considerable weight of water, or when the pump is violently worked. The links are evidently too short; and the immechanical manner in which they are connected, exposes them to a great friction in passing round the wheels. Hence they are sometimes apt to break or burst asunder in very dangerous situations, when it is extremely difficult or impracticable to repair the chain.

The consideration of the known inconveniences of the above machine has given rise to the invention of several others which should better answer the purpose. They have been offered to the public one after another with pompous recommendations by their respective projectors, who have never failed to report their effects as considerably superior to that of the chain-pump with which they have been tried. It is, however, much to be lamented, that in these sort of trials there... there is not always a scrupulous attention to what may be called mechanical justice. The artificer who wishes to introduce a new piece of mechanism, has generally sufficient address to compare its effects with one of the former machines which is crazy or out of repair. A report of this kind indeed favours strongly of the evidence of a false witness; but this fineness is not always discovered. The persons appointed to superintend the comparative effects of the different pumps, have not always a competent knowledge of hydraulics to detect these artifices, or to remark with precision the defects and advantages of those machines as opposed to each other. Thus the several inventions proposed to supplant the chain-pump have hitherto proved ineffectual, and are now no longer remembered.

Of late, however, some considerable improvements have been made on the naval chain-pump, by Mr Cole, under the direction of Captain Bentinck. The chain of this machine is more simple and mechanical, and much less exposed to damage. It is exactly similar to that of the fire engine; and appears to have been first applied to the pump by Mr Mylne, to exhaust the water from the caissons at Blackfriars bridge. It has thence been transferred to the marine by Captain Bentinck, after having received some material additions to answer that service. The principal superiority of this pump to the former is, 1. That the chain is more simple and more easily worked, and of course less exposed to injuries by friction. 2. That the chain is secured upon the wheel, and thereby prevented from jerking back when charged with a column of water. 3. That it may be easily taken up and repaired when broken or choked with ballast, &c. 4. That it discharges a much greater quantity of water with an inferior number of men.

In Plate CCXLVII. a section of this machine at large is exhibited, as fixed in a frigate of war, fig. 5. wherein A is the keel, V the floor-timbers, X the kelson, a a a the several links of the chain, b b the valves, C the upper wheels, D the lower wheels, c c the cavities upon the surface of the wheels to receive the valves as they pass round thereon, and d d the bolts fixed across the surface of the wheels to fall in the interval between every two links to prevent the chain from sliding back.

The links of the chain, which are no other than two long plates of iron with a hole at each end, and fixed together by two bolts serving as axles are represented on a larger scale, as a a. The valves are two circular plates of iron with a piece of leather between them; these are also exhibited at large by b b.

Upon a trial of this machine with the old chain-pump aboard the Seaforth frigate, it appears, in a report figured by rear-admiral Sir John Moore, 12 captains, and 11 lieutenants of his majesty's navy, that its effects, when compared with the latter, were as follow:

| New Pump | Old Pump | |----------|---------| | Number of Men | Tons of Water | Seconds of Time | Number of Men | Tons of Water | Seconds of Time | | 4 | 1 | 43½ | 7 | 1 | 76 | | 2 | 1 | 55 | 4 | 1 | 81 |

The subscribers further certify, that the chain of the new pump was dropped into the well, and afterwards taken up and repaired and set at work again in two minutes and a half; and that they have seen the lower wheel of the said pump taken up to show how readily it might be cleared and refitted for action after being choked with sand or gravel, which they are of opinion may be performed in four or five minutes.

Air-Pump. See Pneumatics, sect. ii.

One of the principal causes of imperfection in the common air-pump, arises from the difficulty of opening the valves at the bottom of the barrels: to avoid which inconvenience, Mr Smeaton has made use of seven holes instead of one; by which means, the valve is supported at proper distances, by a kind of grating, made by the solid parts between these holes: and to render the points of contact, between the bladder and grating, as few as possible, the holes are made hexagonal, and the partitions filed almost to an edge. He has also made the breadth of each hexagon 1/10 of an inch, so that its surface is more than nine times greater than common; upon which account, as well as by reason of the greater number of holes, the valve may be raised with a sixth part of the force commonly necessary.

Another imperfection is owing to the piston's not fitting exactly, when put down close to the bottom; which leaves a lodgment for air that is not got out of the barrel, and proves of bad effect by hindering the rarefaction from being carried on beyond a certain degree; for as the piston rises, the air will expand itself; but still pressing upon the valve, according to its density, it hinders the air within the receiver from coming out. Hence, were this vacancy to equal the 1/50th part of the capacity of the whole barrel, no air could ever come out of the receiver when once expanded 150 times; though the piston were constantly drawn to the top. This inconvenience Mr Smeaton has endeavoured to overcome, by shutting up the top of the barrel with a plate, having in the middle a collar of leathers, thro' which the cylindrical rod works, that carries the piston. By this means the external air is prevented from pressing upon the piston; but that the air which passes through the valve of the piston from below may be discharged out of the barrel, there is also a valve applied to the plate at the top, that opens upwards. The consequence of this construction is, that when the piston is put down to the bottom of the cylinder, the air in the lodgment under the piston will evacuate itself so much the more as the valve of the piston opens more easily, when pressed by the rarified air above it, than when pressed by the whole weight of the atmosphere. Hence, as the piston may be made to fit as nearly to the top of the cylinder as it can to the bottom, the air may be rarified as much above the piston as it could before have been in the receiver. It follows, therefore, that the air may now be rarified in the receiver, in the duplicate proportion of what it could be upon the common principle; every thing else being supposed perfect.

Mr Smeaton has also improved upon the gages commonly used for measuring the expansion of the air; which his gage will do with much certainty, to much less than the 1000th part of the whole. It consists of a bulb of glass, something in the shape of a pear, and sufficient... sufficient to hold about half a pound of quicksilver. It is open at one end, and at the other end is a tube hermetically closed at top. By the help of a nice pair of scales, he found what proportion of weight a column of mercury of a certain length, contained in the tube, bore to that which filled the whole vessel; and, by these means, was enabled to mark divisions upon the tube, answering to the 1/25th part of the whole capacity; which being about 1/4th of an inch each, may, by estimation, be easily subdivided into lesser parts. This gage, during the exhausting of the receiver, is suspended therein by a slip-wire; and when the pump is worked as much as shall be thought necessary, the gage is pushed down, till the open end is immersed in a cistern of quicksilver placed underneath: the air being then let in, the quicksilver will be driven into the gage, till the air remaining in it becomes of the same density with the external air; and as the air always takes the highest place, the tube being uppermost, the expansion will be determined by the number of divisions occupied by the air at top.

He has also endeavoured to render the pneumatic apparatus more simple and commodious, by making the air-pump act as a condensing engine at pleasure, by only turning a cock: this renders the pump an universal engine for showing any effect that arises from an alteration in the density or spring of the air; and with a little addition of apparatus, it shows the experiments of the air-fountain, wind-gun, &c. This is done in the following manner: The air above the piston being forcibly driven out of the barrel at each stroke, and having no where to escape but by the valve at top; if this valve be connected with the receiver by means of a pipe, and at the same time the valve at the bottom, instead of communicating with the receiver, be made to communicate with the external air, the pump will then perform as a condenser. The mechanism is thus ordered: There is a cock with three pipes placed round it, at equal distances. The key is so pierced, that any two may be made to communicate, while the other is left open to the external air. One of these pipes goes to the valve at the bottom of the barrel; another goes to the valve at the top; and a third goes to the receiver. Thus, when the pipe from the receiver, and that from the bottom of the barrel, are united, the pump exhausts; but turn the cock round till the pipe from the receiver and that from the top of the barrel communicate, and it then condenses. The third pipe in one case, discharges the air taken from the receiver into the barrel; and, in the other, lets it into the barrel, that it may be forced into the receiver.

But the following figures will serve to render the structure and use of this excellent machine still more plain. Fig. 1. is a perspective view of the several parts of the pump together. A is the barrel; B the cistern, in which are included the cock, with several joints: these are covered with water, to keep them air-tight. A little cock to let the water out of the cistern, is marked C. Cc is the triangular handle of the key of the cock; which, by the marks on its arms, shows how it must be turned, that the pump may produce the effect desired. DH is the pipe of communication between the cock and the receiver.; E is the pipe that communicates between the cock and the valve, on the upper plate of the barrel. F is the upper plate of the pump, which contains the collar of leathers d; and V, the valve, which is covered by the piece f. GI is the siphon-gage, which screws on and off, and is adapted to common purposes. It consists of a glass tube hermetically sealed at c, and furnished with quicksilver in each leg; which, before the pump begins to work, lies level in the line ab; the space bc being filled with air of the common density. When the pump exhausts, the air in bc expands, and the quicksilver in the opposite leg rises till it become a counterbalance to it. Its rise is shown upon the scale Ie, by which the expansion of the air in the receiver may be nearly judged of. When the pump condenses, the quicksilver rises in the other leg, and the degree may be nearly judged of by the contraction of the air in bc; marks being placed at 1/2 and 3/4 of the length of bc from c, which shows when the receiver condenses double or treble its common quantity. KL is a screw-frame to hold down the receiver in condensing experiments, which takes off at pleasure; and is sufficient to hold down a receiver, the diameter of whose base is 7 inches, when charged with a treble atmosphere; in which case it acts with a force of about 1200 pounds against the screw-frame. M is a screw that fastens a bolt, which slides up and down in that leg, by means whereof the machine is made to stand fast on uneven ground.

Plate CCXLVII. Fig. 1. represents a perpendicular section of the barrel and cock, &c. of the pump; where AB is the barrel, CD the rod of the piston, which passes through MN, the plate that closes the top of the barrel. K is the collar of leathers through which the piston-rod passes. When the piston is at the bottom of the cylinder, the upper part of K is covered by the cap at D, to keep out dust, &c. L is the valve on the upper plate, which is covered by the piece OP, which is connected with the pipe QR, which makes the communication between the valve and cock. CE is the piston, and EFF the piston-valves. II are two little holes to let the air pass from the piston-valves into the upper part of the barrel. GG is the principal valve at the bottom of the cylinder. HH is a piece of metal, into which the valve GG is screwed, and closes the bottom of the cylinder; out of which is also composed SS the cock, and KTT the duct from the cock to the bottom of the barrel. WW is the key of the cock, X the stem, and VV the handle.

Fig. 2. is an horizontal section of the cock, through the middle of the duct TT. AB represents the bigness of the circular plate that closes the bottom of the barrel, and CD the bigness of the inside of the barrel. EFG is the body of the cock; the outward shell being pierced with three holes at equal distances, and corresponding to the three ducts HH, II, KK, whereof HH is the duct that goes to the bottom of the barrel; II, the duct that communicates with the top of the barrel; and KK, the duct that passes from the cock to the receiver. LMN is the key, or solid part of the cock, moveable round in the shell EFG. When the canal LM answers to the ducts HH and KK, the pump exhausts, and the air is discharged by the perforation N. But the key LMN being turned till the canal LM answers to the ducts II and KK, the perforation N will then answer to the duct HH, and in this this case the pump condenses. Lastly, when N answers to KK, the air is then left in or discharged from the receiver, as the circumstance requires.

Fig. 3. is the plan of the principal valve; where ABCD represents the bladder fastened in four places, and stretched over the seven holes IK, formed into an hexagonal grating, which Mr Smeaton chooses to call the honeycomb. EFGH, shows where the metal is a little protuberant, to hinder the piston from striking against the bladder.

Fig. 4. represents the new gage, called from its shape the pear-gage, which is open at A. BC is the graduated tube, which is hermetically closed at C, and suspended by the piece of brass DE; which is hollowed into a cylinder, and clasps the tube.

In the 67th volume of the philosophical transactions we have an account of a number of experiments made by Mr Nairne with an air-pump constructed after the method recommended by Mr Smeaton; in which several unexpected and for some time unaccountable anomalies were observed. These consisted in certain differences between Mr Smeaton's pear-gage, and the common barometer gage. By the former, a degree of exhaustion would be indicated equal to 4000, 10,000, or perhaps 100,000; while the barometer gage indicated only an exhaustion of 200 or 300, or perhaps much less. The reason of this phenomenon was at last explained by Mr Cavendish in the following manner. "Water, whenever the pressure of the atmosphere on it is diminished to a certain degree, is immediately turned into vapour; and is as immediately turned back again into water, on restoring the pressure. This degree of pressure is different according to the heat of the water: when the heat is 72° of Fahrenheit's scale, it turns into vapour as soon as the pressure is no greater than three quarters of an inch of quicksilver, or one-fortieth part of the usual pressure of the atmosphere; but when the heat is only 41°, the pressure of the atmosphere must be reduced to that of a quarter of an inch before it turns into vapour. Hence it follows, that when the receiver is exhausted to the abovementioned degree, the moisture adhering to the various parts of the machine will turn into vapour, and supply the place of the air which is continually drawn away by the working of the pump; so that the fluid in the pear-gage, as well as that in the receiver, will consist in a good measure of vapour. Now, letting the air into the receiver, all the vapour in the pear-gage will be reduced to water, and only the real air will remain uncondensed. Consequently the pear-gage shows only how much real air is left in the receiver, and not how much the pressure or spring of the included fluid is diminished; whereas the common gages show how much the pressure of the included fluid is diminished, and that equally whether it consist of air or vapour."

To put the truth of this theory to the test, Mr Nairne having wiped the receiver and every part of the machine as clean as possible from moisture, excluded the air by a cement put round the outside of the receiver. In these circumstances the pump being worked for 10 minutes, both the barometer and pear-gage indicated very nearly the same degree of exhaustion, viz. 600. He then began to inquire how far different substances, which might occasionally be put into the receiver, would produce this vapour; and the results of his experiments were as follow:

1. A piece of white sheep-skin, of about four inches diameter, soaked in oil and tallow about a year before, being put into the receiver, and the pump worked for 10 minutes, the barometer-gage indicated an exhaustion of about 300, and the pear-gage of 4000. 2. The piece of leather being taken out, and the pump worked as before, both gages stood at 600. 3. A cylindrical piece of boxwood, an inch in diameter, and three inches broad, being put into the receiver, the barometer gage indicated 300, and the pear-gage 16,000. 4. With two ounces of tallow, the barometer gage was 431, the pear-gage 600. 5. With two ounces of oil the numbers were 377 and 480. 6. With two ounces of alum they were 370 and 580. 7. But with a piece of leather, weighing 100 grains, in the same state in which it came from the leather-sellers, the numbers were 152 and 100,000. 8. With the same piece of leather soaked in the tallow and oil which had been already tried, the exhaustions were 432 and 800.

"From these experiments (says our author) it appears, that the elastic vapour which caused so great a difference in the testimony of the gages, arose principally from the leather, and but little from the tallow, oil, or alum: it even appears by the seventh experiment, that it came from the leather, and supplied the place of the exhausted air so fast, that I could not (at least in the 10 minutes) make the barometer gage indicate a degree of exhaustion of more than 159."

"To determine whether it was the moisture in the leather from which the vapour arose, I made the following experiments.

| Substances put into the receiver | Weight when put into the receiver | Degrees of exhaustion according to Barom. gage | Pear-gage | Variation in weight during the experiments | |---------------------------------|----------------------------------|-----------------------------------------------|----------|----------------------------------------| | Exp. 9. A piece of white leather, fresh from the leather-sellers | 100 grains | 134 | 100,000 | lost 2 grains | | Exp. 10. The same piece of leather, dried by the fire till it would lose no more of its weight; | 80 grains | 268 | 280 | gained 2 grs. | | Exp. 11. The same piece of leather held in the steam of hot water till it had regained the 20 grains it had been deprived of; | 100 grains | 147 | 100,000 | lost 2 grains |

"Ya." In this last experiment, it was full three quarters of an hour before the leather regained the 20 grains of weight, although it was held very near the surface of the hot water.

The same piece of leather used in the 8th experiment was put into a damp cellar, where it was left till the next day; it was then put again into the receiver, and the degree of exhaustion according to the barometer-gage was 300, and according to the pear-

gage 3500.

Being now perfectly satisfied that the variation in the testimony of the pear and barometer gages was occasioned by the moisture contained in the substances I had put into the receiver assuming the form of vapour, I determined next to try what would be the effect of the vapour which might arise from small quantities of different fluids, and from some other substances containing moisture of various kinds.

| Substances put into the receiver | Weight when put in | Degree of exhausting according to Barom. gage | Pear gage | Change in weight during the experiment | |---------------------------------|-------------------|-----------------------------------------------|----------|--------------------------------------| | Exp. 12. Water in a watch-glass, | 3 grains | 148 | 24,000 | lost 1½ grain | | Exp. 13. Water in a glass cup, diameter two inches | 100 grains | 89 | 8000 | lost 2 grains | | Exp. 14. Spirit of wine in the same cup, | 100 grains | 54 | 6000 | lost 9 grains | | Exp. 15. Vitriolic acid, | 100 grains | 340 | 220 | gained 1 gr. | | Exp. 16. A piece of the inside of a China orange with some of the rind, | 100 grains | 160 | 100,000 | lost 2½ grs. | | Exp. 17. A piece of the inside of an onion, | 100 grains | 160 | 100,000 | lost 1½ grs. | | Exp. 18. A piece of tainted beef, | 100 grains | 152 | 100,000 | lost 2½ grs. | | Exp. 19. A piece of fresh beef, | 100 grains | 136 | 100,000 | lost 2½ grs. | | Exp. 20. Spirit of turpentine, | 100 grains | 301 | 1800 | lost 2 grains | | Exp. 21. Pearl-ash, | 2 ounces | 118 | 5000 | | | Exp. 22. The same pearl-ash made very hot, | 198 | | 420 | | | Exp. 23. A lighted candle held in the receiver till it went out, | 297 | | 1800 | | | Exp. 24. A piece of charcoal, | 129 | | 1800 | | | Exp. 25. The receiver heated by holding several pieces of lighted charcoal in it, and then the above piece being thoroughly lighted was put into the receiver, and the pump worked, | 650 | | 600 | | | Exp. 26. Camphire, | 100 grains | 304 | 520 | lost barely ½ a grain | | Exp. 27. Sulphur made to burn on a piece of brads | 247 | | 320 | |

Observing by these experiments, that the small quantity of moisture which exhaled from the substances under the receiver prevented the pump from exhausting it to any very considerable degree, I began to suspect, that whenever wet leather had been used to connect the receiver with the plate, there must have risen to so great a quantity of vapour as to have prevented the degree of exhaustion from being near so great as in some of the foregoing instances. These suspicions induced me to make the following experiments.

Exp. 28. The receiver was taken off, and after the cement was wiped clean from it, and every part made perfectly dry, it was put again on the pump plate, and a little oil only was poured round the outside edge.

Exp. 29. The receiver was taken off again, and instead of the oil it was set on a piece of leather, which had been soaked two days in water.

Exp. 30. The last experiment repeated with the same piece of leather.

Exp. 31. The last experiment repeated again with the same piece of leather.

Exp. 32. The receiver was taken off, and instead of the leather soaked in water, there was put on a piece of the same sort of leather soaked in a mixture of water and spirit of wine, such as Mr Smeaton used.

Exp. 33. The last experiment repeated with the same leather.

Exp. 34. The last experiment repeated again with the same leather.

The great difference in the testimony of the pear-

gage gage in these six last experiments appeared to me exceedingly astonishing, for the leathers seemed each of them to be as moist at last as at first.

By these experiments I was convinced how effectually the use of leather soaked in water, or in water and spirit of wine, prevents the pump from exhausting to any considerable degree. I have made a number of experiments of the same kind as these; but have never been able to exhaust, under such circumstances, to a greater degree than between 50 and 60, when the heat of the room was about 57° by a thermometer of Fahrenheit's scale: but the following experiments will show how much some different degrees of heat affect the degree of exhaustion.

| Height of the Therm. | Degrees of exhaustion according to Barom. gage. | Pear gage. | |---------------------|--------------------------------------------------|------------| | Exp. 35. Receiver set on leather which had lain all night in water, | 46 | 84 | 20,000 | | Exp. 36. Receiver set on a leather soaked all night in two parts water and one of spirit of wine, | 46 | 76 | 8000 |

"The pump having been put in a room of the heat of 57° of Fahrenheit's scale for seven hours together, with the leathers put in the same water and the same spirit of wine and water which they had been soaked in all night, and which had been used in the two last experiments, the following experiments were made."

| Height of the Therm. | Degree of exhaustion according to Barom. gage. | Pear gage. | |---------------------|--------------------------------------------------|------------| | Exp. 73. The receiver set on the leather soaked in water, | 57 | 56 | 16,000 | | Exp. 38. Receiver placed on a leather soaked in water and spirit of wine, | 57 | 49 | 1200 |

"The following table will show the comparative excellency between the pump on Mr Smeaton's principle with which the chief of these experiments have been tried, and one of my common double-barreled table air-pumps under the same circumstances. The leather on the pistons of both was soaked in oil and tallow, and the receiver cemented down to each plate; the pumps were both of them fresh oiled."

Pump on Mr Smeaton's principle.

| Degrees of exhaustion according to Barom. gage. | Pear gage. | |--------------------------------------------------|------------| | Exp. 39. A piece of leather, weighing 100 grains, as it came from the leather-fellers, was put into the receiver of each pump, both pieces being cut from the same skin clothe by each other, | 152 | 100,000 | | Exp. 40. The same pieces of leather dried by the fire till they would lose no more of their weight, | 506 | 520 |

"The following experiments will show the effect of water used in the barrels of pumps to make the pistons move air-tight in them.

I took the same common air-pump used in the last experiment, and having taken off the leathers soaked in oil and tallow from the pistons of this pump, and wiped the barrels as clean as possible, I then put new leathers which had been soaked in water, and new bladder valves; the receiver was then cemented to the pump-plate as before.

Degrees of exhaustion according to Barom. gage. | Pear gage. | |--------------------------------------------------|------------| | Exp. 41. The pump was then worked as usual, | 37 | 38 | | Exp. 42. The last experiment repeated with another common pump, the leathers of the pistons of which were also soaked in water, | 34 | 37 |

"From these experiments it evidently appears, that the air-pump of Otto Guericke, and those contrived by Mr Gratorix and Dr Hooke, and the improved one by Mr Pappin, both used by Mr Boyle, also Hauksbee's, S'Gravesande's, Mutenchenbrook's, and those of all who have used water in the barrels of their pumps, could never have exhausted to more than between 40 and 50, if the heat of the place was about 57°; and although Mr Smeaton, with his pump, where no water was in the barrel, but where leather soaked in a mixture of water and spirit of wine was used to set the receiver on the pump-plate, may have exhausted all..." all but a thousandth or even a ten-thousandth part of the common air, according to the testimony of his pear-gage; yet so much vapour must have arisen from the wet leather, that the contents of the receiver could never be less than a 70th or 80th part of the density of the atmosphere. Nevertheless, it does not seem that any deficiency in the construction of Mr Smeaton's pump was the cause of his not being able to exhaust beyond the low degrees of 70 or 80. Had he been aware of the bad effects of setting the receiver upon leather soaked in water and spirit of wine, and had he made use of the precaution to free all parts of his pump as much as possible from moisture, I make not the least doubt but the air-pump which he executed himself would have exhausted to as great a degree, as that pump has been seen to have done with which the chief of these experiments were made.

"Having read the principal part of this paper to Mr Smeaton, and shown him some of the experiments; one, in particular, where the pear-gage, as he observed himself, was filled to no less than 100,000th part of the whole content; he remarked from memory, that he had in several trials exceeded 1000 times, and once, as he remembered, near or about 10,000 times; but as he never could account how this happened, which appeared to him perfectly accidental, and therefore could not depend upon doing it at pleasure, he consented himself with putting down 1000 times, as being what (under the circumstances mentioned in his papers) he had a tolerable certainty of.

"I must here again observe, that if we only wish to know the quantity of permanent air remaining in the receiver after it is as much exhausted as possible, it seems that it is by Mr Smeaton's gage only that we can know it. Again, when, by the assistance of his gage and the barometer-gage together, we have discovered that there is a vapour which arises and occupies the place of the permanent air which is exhausted, it seems that it is by the means of his gage only that we can discover what part of the remaining contents of the receiver consists of this vapour, and what part of permanent air."

In some other experiments the case was surprisingly reversed; for the pear-gage indicated a less degree of exhaustion than the barometer. This happened particularly when the vitriolic acid was put into the receiver. When two grains of this acid were put into the receiver in a glass cup of two inches diameter, the acid gained one grain in weight, the barometer gage indicated an exhaustion of 602, and the pear-gage only of 380. The same experiment being repeated in the same cup, and with the acid which had already gained one grain, the barometer indicated an exhaustion of 502, the pear-gage only of 350, and the acid gained half a grain more. On a third trial with the same materials, the acid gained a quarter of a grain, the barometer indicated an exhaustion of 502, and the pear-gage of 340. Neither was this circumstance entirely removed by taking away the vitriolic acid; for even when this was done, and the receiver and plate of the pump wiped as clean as possible, the barometer gage indicated an exhaustion of 502, and the pear-gage only of 370.

On this experiment Mr Nairne has the following remarks. "I know of no circumstance attending this experiment that differed from those in which my former experiments were made when the gages agreed so nearly, unless it was that of the weather: I recollect that it was then very damp, and now it had been very dry for some time. How this circumstance could make so great an alteration in the result of these experiments, I cannot pretend to say." The true reason, however, seems to be this. Air, though it will expand itself to a great degree, yet has a certain limit to its expansion; that is, if we suppose any quantity of air to be included in a vessel, and the capacity of that vessel to be increased indefinitely, we must at last arrive at a certain bulk, when the gravity of the aerial particles would overcome their repulsive power, and the air would expand no farther though the vessel should be enlarged ever so much. If this is the case, it must follow, that when we come near to this limit, the resistance of the air will be less in proportion than at a considerable distance from it. Thus, let us suppose the resistance of air in its natural state to be 100, and the utmost limit of expansion also to be 100: we cannot imagine, that, at the expansion of 100, the resistance would be 1; for air can only resist by the difference between the gravity and repulsive power of its particles. As, therefore, at the expansion of 100 the gravitating and repulsive powers of the aerial particles exactly balance each other, the resistance could be nothing. In like manner, at the expansion of 98, the difference between the powers above-mentioned being only \( \frac{1}{50} \), the resistance could be no more than 1.02, instead of 2 which it ought to be if the air's expansion was unlimited. By the same method we shall find, that when the expansion is only 10, the resistance is 90; but this is greater than the other proportion, for according to it the resistance should only have been 88. Hence we may see, that in great degrees of exhaustion, when the spring of the air is much weakened, the instrument which measures the resistance in the expanded state must always indicate a greater degree of exhaustion, or a smaller degree of resistance, than that which measures the resistance of the small quantity of air which remains in the receiver after it has been again condensed.

Hence we see, that, in all cases where the air is very pure, the pear-gage will indicate a smaller degree of exhaustion than the barometer; but where there is a quantity of water mixed with it, it may show an equal, or much greater degree of exhaustion than the other. This seems to have been the reason of the difference between Mr Nairne's experiments when the air was moist, and when it was dry. For air always contains some quantity of water, part of which is deposited on the receiver and plate of the pump when the elasticity of the air begins to be weakened. When this water is again changed into vapour in consequence of a greater degree of exhaustion, it affects the barometer but not the pear-gage, and thus the degrees of exhaustion indicated by both may be equal, as was the case with Mr Nairne's experiment in damp weather, when the quantity of water in the air was considerable; but in dry weather, when the quantity of water was less, or rather when the air was less readily disposed to part with it, the pear-gage indicated a smaller degree of exhaustion than the barometer. The same thing necessarily happened when vitriolic triotic acid was put into the receiver; for thus the air was deprived of that quantity of water which it was most readily disposed to part with, and thus the difference became very remarkable.

To the same cause, viz., the extrication of some quantity of elastic vapour, are we to ascribe that other phenomenon likewise taken notice of by Mr Nairne, viz., that when he had worked the pump for some minutes, it would indicate a pretty perfect degree of exhaustion, which would afterwards become considerably less by working it farther. But for a full account of the generation of vapours in the vacuum of an air-pump, see the articles Evaporation, Vacuum, and Vapour.