ABCD

ABCD (fig. 2.) is the section of a small digester made of copper. Its lid, which is fastened to the body with screws, is pierced with three holes, each of which its elasticity had a small pipe soldered into it. The first hole was furnished with a brass safety-valve V, nicely fitted to it by grinding. The area of this valve was exactly ¼th of an inch. There rested on the stalk at top of this valve the arm of a steelyard carrying a sliding weight. This arm had a scale of equal parts, so adjusted to the weight that the number on the scale corresponded to the inches of mercury, whose pressure on the under surface of the valve is equal to that of the steelyard on its top; so that when the weight was at the division 10, the pressure of the steelyard on the valve was just equal to that of a column of mercury 10 inches high, and ¼th of an inch base. The middle hole contained a thermometer T firmly fixed into it, so that no vapour could escape by its sides. The ball of this thermometer was but a little way below the lid. The third hole received occasionally the end of a glass-pipe SGF, whose descending leg was about 36 inches long. When this syphon was not used, the hole was properly shut with a plug.

The vessel was half filled with distilled water which had been purged of air by boiling. The lid was then fixed on, having the third hole S plugged up. A lamp being placed under the vessel, the water boiled, and the steam issued copiously by the safety-valve. The thermometer stood at 213, and a barometer in the room at 29.9 inches. The weight was then put on the fifth division. The thermometer immediately began to rise; and when it was at 220, the steam issued by the sides

of the valve. The weight was removed to the 10th division; but before the thermometer could be distinctly observed, the steam was issuing at the valve. The lamp was removed farther from the bottom of the vessel, that the progress of heating might be more moderate; and when the steam ceased to issue from the valve, the thermometer was at 227. The weight was now shifted to 15; and by gradually approaching the lamp, the steam again issued, and the thermometer was at 132$\backslash frac\{1\}\{2\}$. This mode of trial was continued all the way to the 75th division of the scale. The experiments were then repeated in the contrary order; that is, the weight being suspended at the 75th division, and the steam issuing strongly at the valve, the lamp was withdrawn, and the moment the steam ceased to come out, the thermometer was observed. The same was done at the 70th, 65th, division, &c. These experiments were several times repeated both ways; and the means of all the results for each division are expressed in the following table, where column 1st expresses the elasticity of the steam, being the sum of 29.9, and the division of the steelyard; column 2d expresses the temperature of the steam corresponding to this elasticity.

A very different process was necessary for ascertaining the elasticity of the steam in lower temperatures, and consequently under smaller pressures than that of the atmosphere. The glass syphon SGF was now fixed into its hole in the lid of the digester. The water was made to boil smartly for some time, and the steam issued copiously both at the valve and at the syphon. The lower end of the syphon was now immersed into a broad saucer of mercury, and the lamp instantly removed, and every thing was allowed to grow cold. By this the steam was gradually condensed, and the mercury rose in the syphon, without sensibly sinking in the saucer. The valve and all the joints were smeared with a thick clammy cement, composed of oil, tallow, and rosin, which effectually prevented all ingress of air. The weather was clear and frosty, and the barometer standing at 29.84, and the thermometer in the vessel at 42$^\{\backslash circ\}$. The mercury in the syphon stood at 29.7, or somewhat higher, thus showing a very complete condensation. The whole vessel was surrounded with pounded ice, of the temperature 32$^\{\backslash circ\}$. This made no sensible change in the height of the mercury. A mark was now made at the surface of the mercury. One observer was stationed at the thermometer, with instructions to call out as the thermometer reached the divisions 42, 47, 52,

57, and so on by every five degrees till it should attain the boiling heat. Another observer noted the corresponding descents of the mercury by a scale of inches, which had its beginning placed at 29.84 from the surface of the mercury in the saucer.

The pounded ice was now removed, and the lamp placed at a considerable distance below the vessel, so as to warm its contents very slowly. These observations being very easily made, were several times repeated, and their mean results are set down in the following table: Only observe, that it was found difficult to note down the descents for every fifth degree, because they succeeded each other so fast. Every 10th was judged sufficient for establishing the law of variation. The first column of the table contains the temperature, and the second the descent (in inches) of the mercury from the mark 29.84.

Four or five numbers at the top of the column of elasticities are not so accurate as the others, because the mercury passed pretty quickly through these points. But the progress was extremely regular through the remaining points; so that the elasticities corresponding to temperatures above 70$^\{\backslash circ\}$ may be considered as very accurately ascertained.

Not being altogether satisfied with the method employed for measuring the elasticity in temperatures above that of boiling water, a better form of experiment was adopted. (Indeed it was the want of other apparatus which made it necessary to employ the former). A glass tube was procured of the form represented in fig. 3. having a little cistern L, from the top and bottom of which proceeded the syphons K and MN. The cistern contained mercury, and the tube MN was of a slender bore, and was about six feet two inches long. The end K was firmly fixed in the third hole of the lid, and the long leg of the syphon was furnished with a scale of inches, and firmly fastened to an upright post.

The lamp was now applied at such a distance from the vessel as to warm it slowly, and make the water boil, the steam escaping for some time through the safety-valve. A heavy weight was then suspended on the steelyard; such as it was known that the vessel would support, and at the same time, such as would not allow the steam to force the mercury out of the long tube. The thermometer began immediately to rise, as also the mercury

Steam. mercury in the tube MN. Their correspondent stations are marked in the following table :

This form of the experiment is much more susceptible of accuracy than the other, and the measures of elasticity are more to be depended on. In repeating the experiment, they were found much more constant; whereas, in the former method, differences occurred of two inches and upwards.

We may now connect the two sets of experiments into one table, by adding to the numbers in this last table the constant height 29.9, which was the height of the mercury in the barometer during the last set of observations.

²³ Achard. In the memoirs of the Royal Academy of Berlin for 1782, there is an account of some experiments made by Mr Achard on the elastic force of steam, from the temperature 32° to 212°. They agree extremely well with those mentioned here, rarely differing more than two or three tenths of an inch. He also examined the elasticity of the vapour produced from alcohol, and found, that when the elasticity was equal to that of the vapour of water, the temperature was about 35° lower. Thus, when the elasticity of both was measured by 28.1 inches of mercury, the temperature of the watery vapour was 209°, and that of the spirituous vapour was 173°. When the elasticity was 18.5, the temperature of the water was 189.5, and that of the alcohol 154.6. When the

elasticity was 11.05, the water was 168°, and the alcohol 134°.4. Observing the difference between the temperatures of equally elastic vapours of water and alcohol not to be constant, but gradually to diminish, in Mr Achard's experiments, along with the elasticity, it became interesting to discover whether and at what temperature this difference would vanish altogether. Experiments were accordingly made by the writer of this article, similar to those made with water. They were not made with the same scrupulous care, nor repeated as they deserved, but they furnished rather an unexpected result. The following table will give the reader a distinct notion of them :

We say that the result was unexpected; for as the natural boiling point seemed by former experiments to be in all fluids about 120° or more below their boiling point in the ordinary pressure of the atmosphere, it was reasonable to expect that the temperature at which they ceased to emit sensibly elastic steam would have some relation to their temperatures when emitting steam of any determinate elasticity. Now as the vapour of alcohol of elasticity 30 has its temperature about 36° lower than the temperature of water equally elastic, it was to be expected that the temperature at which it ceased to be sensibly affected would be several degrees lower than 32°. It is evident, however, that this is not the case. But this is a point that deserves more attention, because it is closely connected with the chemical relation between the element (if such there be) of fire, and the bodies into whose composition it seems to enter as a constituent part. What is the temperature 32°, to make it peculiarly connected with elasticity? It is a temperature assumed by us for our own convenience, on account of the familiarity of water in our experiments. Ether, we know, boils in a temperature far below this, as appears from Dr Cullen's experiments narrated in the Essays Physical and Literary of Edinburgh. On the faith of former experiments, we may be pretty certain that it will boil in vacuo at the temperature -14°, because in the air it boils at +106°. Therefore we may be certain, that the steam or vapour of ether, when of the temperature 32°, will be very sensibly elastic. Indeed Mr Lavoisier says, that if it be exposed in an exhausted receiver in winter, its vapour will support mercury at the height of 10 inches. A series of experiments on this vapour similar to the above would be very instructive. We even wish that those on alcohol were more carefully repeated. If we draw a curve line, of which the abscissa is the line of temperatures, and the ordinates are the corresponding heights of the mercury in these experiments on water and alcohol,

Steam. we shall observe, that although they both sensibly coincide at $32^\{\backslash circ\}$, and have the abscissa for their common tangent, a very small error of observation may be the cause of this, and the curve which expresses the elasticity of spirituous vapour may really intersect the other, and go backwards considerably beyond $32^\{\backslash circ\}$.

25 These experiments give rise to important reflections. This range of experiments gives rise to some curious and important reflections. We now see that no particular temperature is necessary for water assuming the form of permanently elastic vapour; and that it is highly probable that it assumes this form even at the temperature $32^\{\backslash circ\}$; only its elasticity is too small to afford us any sensible measure. It is well known that even ice evaporates (see experiments to this purpose by Mr Wilson in the Philosophical Transactions, when a piece of polished metal covered with hoar-frost became perfectly clear by exposing it to a dry frosty wind).

Even mercury evaporates, or is converted into elastic vapour, when all external pressure is removed. The dim film which may frequently be observed in the upper part of a barometer which stands near a stream of air, is found to be small globules of mercury flicking to the inside of the tube. They may be seen by the help of a magnifying glass, and are the best test of a well made barometer. They will be entirely removed by causing the mercury to rise along the tube. It will lick them all up. They consist of mercury which had evaporated in the void space, and was afterwards condensed by the cold glass. But the elasticity is too small to occasion a sensible depression of the column, even when considerably warmed by a candle.

Many philosophers accordingly imagine, that spontaneous evaporation in low temperatures is produced in this way. But we cannot be of this opinion, and must still think that this kind of evaporation is produced by the dissolving power of the air. When moist air is suddenly rarefied, there is always a precipitation of water. This is most distinctly seen when we work an air-pump briskly. A mist is produced, which we see plainly fall to the bottom of the receiver. But by this new doctrine the very contrary should happen, because the tendency of water to appear in the elastic form is promoted by removing the external pressure; and we really imagine that more of it now actually becomes simple elastic watery vapour. But the mist or precipitation flows incontrovertibly, that there had been a previous solution. Solution is performed by forces which act in the way of attraction; or, to express it more safely, solutions are accompanied by the mutual approaches of the particles of the menstruum and solved: all such tendencies are observed to increase by a diminution of distance. Hence it must follow, that air of double density will dissolve more than twice as much water. Therefore when we suddenly rarefy saturated air (even though its heat should not diminish) some water must be let go. What may be its quantity we know not; but it may be more than what would now become elastic by this diminution of surrounding pressure; and it is not unlikely but this may have some effect in producing the vesicles which we found it so difficult to explain. These may be filled with pure watery vapour, and be floating in a fluid composed of water dissolved in air. An experiment of Fontana's seems to put this matter out of doubt. A distilling apparatus AB (fig. 4.)

was so contrived, that the heat was applied above the surface of the water in the alembic A. This was done by inclosing it in another vessel CC, filled with hot water. In the receiver B there was a sort of barometer D, with an open cistern, in order to see what pressure there was on the surface of the fluid. While the receiver and alembic contained air, the heat applied at A produced no sensible distillation during several hours: But on opening a cock E in the receiver at its bottom, and making the water in the alembic to boil, steam was produced which soon expelled all the air, and followed it through the cock. The cock was now shut, and the whole allowed to grow cold by removing the fire, and applying cold water to the alembic. The barometer fell to a level nearly. Then warm water was allowed to get into the outer vessel CC. The barometer rose a little, and the distillation went on briskly without the smallest ebullition in the alembic. The conclusion is obvious: while there was air in the receiver and communicating pipe, the distillation proceeded entirely by the dissolving power of this air. Above the water in the alembic it was quickly saturated; and this saturation proceeded slowly along the still air in the communicating pipe, and at last might take place through the whole of the receiver. The sides of the receiver being kept cold, should condense part of the water dissolved in the air in contact with them, and this should trickle down the sides and be collected. But any person who has observed how long a crystal of blue vitriol will lie at the bottom of a glass of still water before the tinge will reach the surface, will see that it must be next to impossible for distillation to go on in these circumstances; and accordingly none was observed. But when the upper part of the apparatus was filled with pure watery vapour, it was supplied from the alembic as fast as it was condensed in the receiver, just as in the pulse glass.

Another inference which may be drawn from these experiments is, that Nature seems to affect a certain law in the dilatation of aeriform fluids by heat. They seem to be dilatable nearly in proportion of their present dilatation. For if we suppose that the vapours resemble air, in having their elasticity in any given temperature proportional to their density, we must suppose that if steam of the elasticity 60, that is, supporting 60 inches of mercury, were subjected to a pressure of 30 inches, it would expand into twice its present bulk. The augmentation of elasticity therefore is the measure of the bulk into which it would expand in order to acquire its former elasticity. Taking the increase of elasticity therefore as a measure of the bulk into which it would expand under one constant pressure, we see that equal increments of temperature produce nearly equal multiplications of bulk. Thus if a certain diminution of temperature diminishes its bulk $\backslash frac\{1\}\{4\}$th, another equal diminution of temperature will diminish this new bulk $\backslash frac\{1\}\{4\}$th very nearly. Thus, in our experiments, the temperatures $110^\{\backslash circ\}$, $140^\{\backslash circ\}$, $170^\{\backslash circ\}$, $200^\{\backslash circ\}$, $230^\{\backslash circ\}$, are in arithmetical progression, having equal differences; and we see that the corresponding elasticities 2.25, 5.15, 11.05, 22.62, 44.7, are very nearly in the continued proportion of 1 to 2. The elasticity corresponding to the temperature $260$ deviates considerably from this law, which would give 88 or 89 instead of 80; and the deviation

Steam. deviation increases in the higher temperatures. But still we see that there is a considerable approximation to this law; and it will frequently assist us to recollect, that whatever be the present temperature, an increase of 30 degrees doubles the elasticity and the bulk of water vapour.

This is sufficiently exact for most practical purposes. Thus an engineer finds that the injection cools the cylinder of a steam-engine to 192°. It therefore leaves a steam whose elasticity is three-fifths of its full elasticity, = 18 inches $\backslash frac\{1\}\{2\}$. But it is better at all times to have recourse to the table. Observe, too, that in the lower temperatures, i. e. below 110°, this increment of temperature does more than double the elasticity.

This law obtains more remarkably in the incoercible vapours; such as vital air, atmospheric air, fixed air, &c. all of which have also their elasticity proportional to their bulk inversely: and perhaps the deviation from the law in steams is connected with their chemical difference of constitution. If the bulk were always augmented in the same proportion by equal augmentations of temperature, the elasticities would be accurately represented by the ordinates of a logarithmic curve, of which the temperatures are the corresponding abscissæ; and we might contrive such a scale for our thermometer, that the temperatures would be the common logarithms of the elasticities, or of the bulks having equal elasticity; or, with our present scale, we may find such a multiplier $m$ for the number $x$ of degrees of our thermometer (above that temperature where the elasticity is equal to unity), that this multiple shall be the common logarithm of the elasticity $y$; so that $mx\; =\; \backslash log.\; y$.

But our experiments are not sufficiently accurate for determining the temperature where the elasticity is measured by 1 inch; because in these temperatures the elasticities vary by exceedingly small quantities. But if we take 11.04 for the unit of elasticity, and number our temperature from 170°, and make $m\; =\; 0.010035$, we shall find the product $mx$ to be very nearly the logarithm of the elasticity. The deviations, however, from this law, are too great to make this equation of any use. But it is very practicable to frame an equation which shall correspond with the experiments to any degree of accuracy; and it has been done for air in a translation of General Roy's Measurement of the Base at Hounslow Heath into French by Mr Prony. It is as follows: Let $x$ be the degrees of Reaumur's thermometer; let $y$ be the expansion of 10,000 parts of air; let $e$ be = 10, $m\; =\; 2.7979$, $n\; =\; 0.01768$; then $y\; =\; e^\{m+nx\}\; -\; 627.5$. Now $e$ being = 10, it is plain that $e^\{m+nx\}$ is the number, of which $m+nx$ is the common logarithm. This formula is very exact as far as the temperature 60°: but beyond this it needs a cor-

rection; because air, like the vapour of water, does not expand in the exact proportion of its bulk. Steam.

We observe this law considerably approximated to in the augmentation of the bulk or elasticity of elastic vapours; that is, it is a fact that a given increment of temperature makes very nearly the same proportional augmentation of bulk and elasticity. This gives us some notion of the manner in which the supposed expanding cause produces the effect. When vapour of the bulk 4 is expanded into a bulk 5 by an addition of 10 degrees of sensible heat, a certain quantity of fire goes into it, and is accumulated round each particle, in such a manner that the temperature of each, which formerly was $m$, is now $m+10$. Let it now receive another equal augmentation of temperature. This is now $m+20$, and

the bulk is $\backslash frac\{5\; \backslash times\; 5\}\{4\}$ or $6\backslash frac\{1\}\{4\}$, and the arithmetical increase of

bulk is $1\backslash frac\{1\}\{4\}$. The absolute quantity of fire which has entered it is greater than the former, both on account of the greater augmentation of space and the greater temperature. Consequently if this vapour be compressed into the bulk 5, there must be heat or fire in it which is not necessary for the temperature $m+20$, far less for the temperature $m+10$. It must therefore emerge, and be disposed to enter a thermometer which has already the temperature $m+20$: that is, the vapour must grow hotter by compression; not by squeezing out the heat, like water out of a sponge, but because the law of attraction for heat is deranged. It would be a very valuable acquisition to our knowledge to learn with precision the quantity of sensible heat produced in this way; but no satisfactory experiments have yet been made. M. Lavoisier, with his chemical friends and colleagues, were busily employed in this inquiry: but the wickedness of their countrymen deprived the world of this and many other important additions which we might have expected from this celebrated and unfortunate philosopher. He had made, in conjunction with M. de la Place, a numerous train of accurate and expensive experiments for measuring the quantity of latent or combined heat in elastic vapours. This is evidently a very important point to the distiller and practical chemist. This heat must all come from the fuel; and it is greatly worth while to know whether any saving may be made of this article. Thus we know that distillation will go on either under the pressure of the air, or in an alembic and receiver from which the air has been expelled by steam; and we know that this last may be conducted in a very low temperature, even not exceeding that of the human body. But it is uncertain whether this may not employ even a greater quantity of fuel, as well as occasion a great expense of time. We are disposed to think, that when there is no air in the apparatus, and when the condensation can be speedily performed, the proportion of fuel expended to the fluid which comes over will diminish continually as the heat, and consequently the density of the steam, is augmented; because in this case the quantity of combined heat must be less. In the mean time, we earnestly recommend the trial of this mode of distillation in vessels cleared of air. It is undoubtedly of great advantage to be able to work with smaller fires; and it would secure us against all accidents of blowing off the

Steam. the head of the still, often attended with terrible consequences (B).

We must not conclude this article without taking notice of some natural phenomena which seem to owe their origin to the action of elastic steam.

We have already taken notice of the resemblance of the tremor and succussions observed in the shocks of many earthquakes to those which may be felt in a vessel where water is made to boil internally, while the breaking out of the ebullition is stifled by the cold of the upper parts; and we have likewise stated the objections which are usually made to this theory of earthquakes. We may perhaps resume the subject under the article VOLCANO; but in the mean time we do not hesitate to say, that the wonderful appearances of the Geyzer spring in Iceland (see HUER; and ICELAND, N^o 3—5.) are undoubtedly produced by the expansion of steam in ignited caverns. Of these appearances we suppose the whole train to be produced as follows.

A cavern may be supposed of a shape analogous to CBDEF (fig. 5.), having a perpendicular funnel AB issuing from a depressed part of the roof. The part F may be lower than the rest, remote, and red hot. Such places we know to be frequent in Iceland. Water may be continually trickling into the part CD. It will fill it up to B, and even up to E e, and then trickle slowly along into F. As soon as any gets into contact with an ignited part, it expands into elastic steam, and is partly condensed by the cold sides of the cavern, which it gradually warms, till it condenses no more. This production of steam hinders not in the smallest degree the trickling of more water into F, and the continual production of more steam. This now presses on the surface of the water in CD, and causes it to rise gradually in the funnel BA; but slowly, because its cold surface is condensing an immense quantity of steam. We may easily suppose that the water trickles faster into F than it is expended in the production of steam; so that it reaches farther into the ignited part, and may even fall in a stream into some deeper pit highly ignited. It will now produce steam in vast abundance, and of prodigious elasticity; and at once push up the water through the funnel in a solid jet, and to a great height. This must continue till the surface of the water sinks to BD. If the lower end of the funnel have any inequalities or notches, as is most likely, the steam will get admission

along with the water, which in this particular place is boiling hot, being superficial, and will get to the mouth of the funnel, while water is still pressed below. At last the steam gets in at B on all sides; and as it is converging to B, along the surface of the water, with prodigious velocity it sweeps along with it much water, and blows it up through the funnel with great force. When this is over, the remaining steam blows out unmixed with water, growing weaker as it is expended, till the bottom of the funnel is again stopped by the water increasing in the cavern CBD. All the phenomena above ground are perfectly conformable to the necessary consequences of this very probable construction of the cavern. The feeling of being lifted up, immediately before the jet, in all probability is owing to a real heaving up of the whole roof of the cavern by the first expansion of the great body of steam. We had an accurate description of the phenomena from persons well qualified to judge of these matters who visited these celebrated springs in 1789.