an instrument for trying the salubrity of air, founded on a fact discovered by Dr Priestley; namely, that nitrous air diminishes the bulk of common atmospheric air in proportion to the salubrity of the latter.*—The Abbé Fontana and Chevalier Landriani, were the first, as it seems, who availed themselves of this discovery. Both proposed to the public an instrument for measuring the salubrity of the air we breathe. They gave to these instruments, called eudiometers, different forms, as appears by the printed descriptions that each of them has separately published: and the chevalier Landriani transmitted to England, as a present to Dr Priestley, the very instrument he had made use of to estimate the respective salubrity of the air in different parts of Italy. This eudiometer consists of a glass tube, ground to a cylindrical vessel, with two glass cocks, and a small basin, all fitted in a wooden frame. Quicksilver is there used instead of water; and that part of it which replaces the bulk lost by the diminution of the two mixed airs, is conducted either through a kind of glass siphon, or through the capillary holes of a glass funnel: so that, by its fall, the whole mixture of the two kinds of air is more readily made. Dr Falconer of Bath sent, some time ago, to Dr Falconer of London, a glass tube, neatly divided; by means of which one may be enabled to know the quantity of diminution produced in a certain bulk of the mixture of nitrous air with another air, in order to judge of its salubrity, which Dr Priestley has shown to be in proportion to the diminution suffered in the sum of their original bulk, after they are mixed together. This method is the readiest of all, when no great nicety is required in observations of this kind; but, in order to determine this matter with the greatest exactness, others have been contrived by J. H. De Magellan, F. R. S. of which he gives the following description in his letter to Dr Priestley. "Of the three eudiometers I have contrived, which are represented fig. 8, 15, and 16.* I think the latter is the easiest in its application, and the most exact in its results. It is represented also (fig. 12, t. q. and 17.) in different positions, for the better understanding of its application; and it consists of the following parts, viz. a glass tube m n e d, fig. 16, about 12 or 15 inches long, and of an equal diameter; with a ground glass stopple m; a vessel c, the neck of which is ground air-tight to the lower end d of the tube; and two equal phials a and b, whose necks are also ground air-tight to the respective mouths of the vessel c. Both these phials contain nearly as much as the whole tube m n e d. There is, moreover, a sliding brass-ring, marked z, which slides in the tube n d, and may be made tight at pleasure by a finger-screw; and, lastly, a ruler, either of brass or of wood, represented fig. 11, which is divided into equal parts, and indicates the contents of both the phials a and b, when thrown into the tube, by the number of parts which is engraved or stamped about the middle of it. The two bent pieces of brass z t serve to hold it easily by the side of the tube n d. Experiments with these endiometers, which are easily constructed, may be made either with water or with quicksilver; with this difference, that when the last is made use of, the endiometers (particularly the third, represented fig. 8, which seems the fittest to be used with quicksilver) will be more convenient if made of a still smaller size. Mercury, however, is a fluid that, I think, never ought to be used preferably to water, in the inside of endiometers; because it suffers a sensible action from the contact with nitrous air, as yourself have observed: and this must have an influence on the result of the experiments. Water, on the contrary, seems less liable to mistakes, although it imbibes some part of the nitrous air. In fact, this effect only takes place in a long time, or with much agitation; and after duly weighing the question on both sides, I should think water might be generally used, without the fear of any sensible error. The weight and the clearness of quicksilver, are likewise two other considerations to give the preference to water in these experiments.

"The Process." In the first place there must be either a trough, as represented fig. 17.; or at least a common tub, nearly filled up with water, unless the tall glass receiver, of which I shall afterwards speak, be at hand. I take out the stopple m (fig. 16.) and fill the endiometer entirely with water, keeping it in the position represented fig. 16. and 17. I then shut it with the stopple m, without leaving any bubble of air in the inside; and put the lower part c under the surface of the water in the tub (fig. 17.) in an erect position as it is therein seen. I take the phial a, filled with water; and keeping its mouth downwards under the surface of the water, I fill it with that air, the salubrity of which I want to ascertain (a). This is done either by putting the phial a on the shelf n o of the tub (fig. 17.) and throwing the air into the glass-funnel t, which is there cemented to the shelf; or by holding in the left hand the same phial a, together with the glass funnel B (which is represented fig. 18., and has no pipe at all) applied to the mouth of the phial, whilst I pour the air with my right hand into it. But lest the heat of my hand should produce any considerable expansion in this air, I generally use in hot weather the wooden tongs represented fig. 21. with two bent wires xx, in order to hold the glass funnel z close to the mouth of the phials; unless they are made with a solid lump at their bottoms, as represented in the plate.—There are some niceties to be observed in order to fill up exactly any phial intended to serve as a measure of air. The easiest method to succeed is the following: Let a glass funnel t (fig. 17.) be cemented under the hole n of the shelf n o in the trough. In this case I hold the phial a, filled with water, with its mouth downwards over the hole n of the funnel t; I throw the air into the funnel; and, when

(a) "The case I am speaking of, is when I have a bottle of air, which has been taken at any distant place, and sent for trial. If a glass-bottle, with a ground glass-stopper, is filled with water or with mercury, and emptied in the place whose atmospheric air is intended for being examined, it will, of course, be filled with that air; and, being closely shut with the glass-stopper, may be carried to any distant place for a trial. By this means the atmospheric air of any part of a country may be sent to any distant place, in order to ascertain its comparative salubrity; and many useful inquiries and discoveries may be made hereafter on this subject, with great ease, and at very small expense. But if I only want to try the air of the room, where I have the endiometer, I then only pour out of the phial a the water it contains. I find that, however, after some trials with nitrous air, the atmosphere about me is loaded with phlogistic miasma; and for that reason I always empty the phial a out of the window of the room, in order to have nearly the same kind of air in all the experiments." EUD

Eudiometer mouth of which is ground air-tight; the crooked tube \( n z \) in the shape of an S. I fill the half of this phial with thin brafs wire, the thickness of which is equal to \( \frac{3}{16} \) of an English inch, nicely cut by a pin-maker to this length. I fill the three quarters of the phial with common water; and the remainder with strong nitrous acid. I put the crooked tube \( n z \) into the phial; and, as soon as the effervescence causes the liquor to rise to the end \( z \) of the tube, I pass it under water into the mouth of the bottle \( E \) (fig. 20.) which is filled with water, and inverted with its mouth downwards upon the hole of the shelf \( n o \), which appears covered with water within the trough or pan, (fig. 17.) This figure represents the most commodious shape a trough must have for any experiments on different kinds of air. It is made with straight boards of elm-wood one inch thick. The inside dimensions are 25 inches long, \( \frac{13}{4} \) wide, and 11 deep, English measure. The two end boards, \( c d \) and \( e f \), are fitted into a groove cut in the other three boards; this is daubed with thick white painting, as a cement, to keep well the water in; and the whole is fastened with nails from the outside. The shelf \( w a n o \) is eight inches wide, and two inches thick. It has three holes of three tenths of an inch diameter, with as many separate cavities underneath, so as to serve as so many funnels. The figure, however, represents a glass funnel \( t \), cemented to the middle hole \( n \); which is equally convenient. This shelf is supported by four metallic hooks \( V w z z \), which may be raised or lowered at pleasure by the wooden wedges there represented. When the bottle \( F \) is entirely filled by the nitrous air, I shut it up with its stopple \( x \) (fig. 20.) which I pass under the surface of the water, to avoid any communication with the external air; and I push this bottle under the shelf, where I let it remain for a quarter of an hour, to acquire the same temperature of the surrounding water; and the same I always observe with the bottle, containing that atmospheric air which I desire to try, before I put it into the phial \( b \). I must acknowledge, however, that, notwithstanding these precautions, I cannot say that all the results of my experiments, even when made upon the same atmospheric air, have as yet agreed to exactly as I flattered myself they would. Perhaps there was some difference in the strength of the nitrous air, the density of which I thought might easily be brought to a settled standard, to be determined by means of a glass hydrometer. Perhaps there was some other little variety in the circumstances of the experiments, the influence of which I was not aware of. But let it be as it may, I very willingly leave this problem to be resolved by abler chemists than I can pretend to be; and I heartily wish they may succeed better than I have done; for, without being assured of getting everywhere a certain standard nitrous air, by which the same atmospheric air may be equally affected, we cannot draw with certainty any general decisive conclusions from eudiometrical experiments made in distant times or places.

I take afterwards the eudiometer with my left hand, holding it near the lower part \( d \), over the surface of the water in the trough, to avoid breaking any of the phials, if it chances to fall; and, with my right hand, I turn the vessel upwards, so that the two phials may be downwards, as represented fig. 14. By this operation the two kinds of air come up to \( x \) from the phials \( a b \); and there they mix together in the best possible manner; the particles of each having a large room to come into contact with each other; since the foremost ones do not detain those which are behind, as it happens when this mixture is made in a narrow vessel. This being done, I immediately dip the eudiometer in the water of the trough, (fig. 17.) leaving the mouth of the instrument above its surface; so that no more water may enter into it than what it had at first. I then observe with attention the moment when the mixture \( x \) (fig. 14.) of the two kinds of air comes to its greatest diminution, after which its bulk will begin to increase again. In order to catch this moment with certainty, I slide down the brass ring \( z \) of the instrument, as the surface of the water in the tube falls. This point of the greatest diminution will be easily perceived, by observing when that inside surface is stationary; which will happen in a few minutes, if the nitrous air has a proper strength. The bulk of the mixed air will decrease to a certain degree, within a few minutes, according to the strength of the nitrous air. Afterwards it will begin to expand again; but this will do to a very short limit, much below its former bulk. This is a phenomenon which, I think, I have observed the first on these experiments; having made a very great number of them with nice eudiometers, of the kind I am now describing. It certainly deserves the attention of philosophers; and, although I have communicated it to some of my acquaintance, none have as yet, in my humble opinion, given a satisfactory solution of this phenomenon.

As soon as the diminution of the two kinds of air appears to be stationary, I fill up the whole tube of the eudiometer with water; I shut it up with the stopple \( w \); and incline the top of the instrument forwards, till the air comes from \( x \) (fig. 14.) up to the top \( n \) of the tube. I then keep the lower part of the instrument dipped in the water; take off the glass vessel \( c \) with the two phials \( a b \), and raise or lower the tube of the eudiometer, so as to see the surface of the water, in the inside, even with that in the outside; which I mark by sliding to it the brass ring \( z \). Otherwise I apply the rulers, fig. 11. (without making any use now of the brass ring) to the side of the eudiometer, whilst it is immersed in the water of the trough; and there I see the true

---

(8) Two striking circumstances relating to nitrous air deserve to be remarked. The first is the great quantity produced by the action of nitrous acid on metals; which may still be carried to a greater extent, if helped by bringing the flame of a candle to the phial, which contains the solution, when it seems to be nearly done with emitting air. The second is the antiseptic power of nitrous air to preserve animal-matters from corruption. A beef-stake, almost entirely putrid, and with an intolerable stench, being put into a jar of nitrous air, in less than two days was perfectly restored, and very eatable when dressed. A pigeon was very well preserved above six weeks by the same treatment; and, when roasted, was found so good as to be eaten without any dislike. Two other pigeons were kept in it full six months without corruption; they were still very firm and of a good colour; but the flesh had lost all its flavour, and was far from being eatable when dressed. But the nitrous air for these economical purposes, which may be of a great advantage at sea as well as at home, must be made out of nitrous acid with iron, or other metal less exceptionable than brass or copper, the effluvia of which are pernicious to animals. true dimension of the remaining bulk of the two kinds of air already diminished. Perhaps the best method for this observation would be to allow time enough that the mixed air may take its settled bulk; but this requires sometimes 24 hours. I leave, however, the choice of these two methods to the observer, who may use both if he pleases, provided he keeps distinctly the result of each method in his account of the experiment.

"The number marked about the middle of this ruler (fig. 11.), as for instance, * * = 96, means that the contents of both phials a and b are equal to 96 divisions of the ruler, when put into the tube of that eudiometer: that is to say, they are equal to a solid cylinder as thick as the inside of the glass tube, and whose length is 96 divisions of the ruler, which has been divided into tenths of an English inch.

"Now if, for instance, this remaining bulk of mixed air corresponds to the 56th division of the ruler, it shows that, out of 96 parts, only 40 (=96−56) have been lost or contracted: and, in this case, the wholesomeness of that air, which I call A, will be \( \frac{40}{96} \).

If another equal quantity of different air, which I shall call B, had also been tried by the same eudiometer, and its residuum was equal to 60 parts of the same ruler, the respective salubrity of the air B will then be to that of the air A, as 36 (=96−60) to 40.

"But if the air B had been tried by another eudiometer, whose proportional dimensions, marked about the middle of its ruler, were * * = 108, then the respective salubrity of these two kinds of air A and B, would be in the compound ratio of \( \frac{36}{96} \times \frac{108}{96} = \frac{3456}{4320} = \frac{54}{67.5} \), that is to say, the wholesomeness of the air B would be to that of the air A, as 54 to 67.5 (c).

"Nearly the same results would be found, if the ruler (fig. 11.) was applied to the side of the eudiometer, as soon as the inclosed mixture of air came to its utmost diminution, as above-mentioned; because as much water must fall in the tube n d, as corresponds to the diminution suffered by the two mixed airs in x. But there are some varieties, which arise from the different pressure of the column of water, which presses more or less upon the air at x (fig. 14.) as it is longer or shorter: and these varieties ought not to be overlooked in nice experiments.

"Whenever I have at hand a tall glass receiver, like that represented fig. 14., the whole process is then more easily performed: for in this case I dip the eudiometer, inverted as it appears fig. 12., into the water contained in the vessel V S q l: I then put the two kinds of air into the phials a and b, as above said: I turn the instrument upright, as represented fig. 14., and finish the process, as I have already described.

"I must, however, warn the operator, that, unless every trial, and even almost every part of the process, be made in the same temperature, or at least unless the varieties arising from this cause be accounted for, no reliance can be had on the result of such experiments; it being well known, that air is apt to increase or diminish very considerably in its bulk, by the influence of heat and cold. It is for this reason that I constantly keep a good thermometer K, which hangs by the wire r, and is immersed in the water of the glass vessel fig. 14., or in the trough fig. 17., whenever I make any of these experiments. For the same reason, I take care to leave the eudiometer and the vessels of air, immersed in water time enough, as above-mentioned, to get the same temperature: and I make use of the wooden tongs mentioned p. 2051, par. 2, whenever I handle the phials a b filled with air, chiefly if they have not the solid lump at their bottoms, as represented in the plate; unless I feel the heat of my hands to be the same as that of the water, in the trough, I make use of.

The eudiometer, represented fig. 15., consists of a second glass tube c e, two or three feet long, and of an uniform diameter: the end c is bent forwards; and the other end t is wide open, as a funnel, unless a separate one is made use of: this tube is fastened, by two loops, to the brass scale c w t v. There is a glass phial n, the neck k of which is ground air-tight to the end t of the tube; and contains only half of the whole inside capacity of the divided tube c t. It has, at the other end c, a large round phial a b c, containing three or four times the bulk of the phial n: its neck is also ground air-tight to the mouth c of the tube. The brass scale c w t v is divided into 128 equal parts: this being a number that can be divided to unity in a subduplicate ratio without fraction, by continual biflections; on which account it is one of the numbers the late famous Mr Bird had adopted for his dividing mathematical instruments with the utmost accuracy. These numbers are set out in the scale from t towards c. The contents or capacity of the tube till the number 128 is the double of the capacity of the phial n. Besides this there is a tin vessel x s d r o (fig. 15.) which may serve as a packing-case for the whole instrument, and its necessary appendages; and also as a trough, when experiments are made, it being then filled with water. Both the glass tube represented fig. 22. and the glass stopple m (fig. 15.), belong to this eudiometer; and both are fitted in, air-tight, to its mouth v.

"Let the instrument be immersed under the water Method of z z of the tin vessel fig. 15.; and let the phial n, filled with water, be put in the inside socket c e d of the tin vessel. Let it be filled with nitrous air, as above-di-

Vol. IV.

(c) "It is supposed that the inside of the tube is of an uniform diameter; but it often happens, that there are some varieties in different parts of its whole length. When they are not very considerable, we may neglect their influence in the result of these eudiometrical experiments; but, when the contrary happens, it will be very easy to make a proper allowance for them in the calculation. It is for this reason, that I have always ordered that the contents of one single phial be marked also upon the scale of each eudiometer, as well as the contents of both phials; for instance, as in this manner:

* * = 96

Which means, first, that the contents of both phials a and b are equal to a cylinder, whose diameter is the same as that of the inside bore of the tube n d (fig. 16.), and whose height is equal to 96 equal divisions of the ruler: secondly, that the contents of a single phial are equal to 47 divisions in the upper part of the same tube n d; and, of course, to 49 divisions (=96−47) of its lower part. By this difference it appears, that the tube of such eudiometer is wider in the top than at the bottom, by \( \frac{2}{9} \) of the whole. Eudiometer rected: and let this quantity of air be thrown into the phial \(a b c\), as directed above, which I fix a little tight to the mouth \(c\) of the eudiometer. I afterwards fill the same phial \(n\) with the air I want to try; and raising the end \(c\) of the instrument, I put it into its mouth \(V\); when this is done, I set the instrument upright, as represented fig. 15., hanging it on the hook \(w\); and, as soon as this last air goes up to the phial \(a b c\), I take off the phial \(n\), that the diminution of the two mixed airs may be supplied from the water in the tin vessel; which must be the case, as the mouth \(V\) of the eudiometer is then under the surface of the water. I then put to the lower end \(V\) of the eudiometer, the bent tube fig. 22., to which is fitted the brass ring \(K\), and is filled with water. It is by observing the surface of the water in this small tube (which then forms a true siphon with the tube of the instrument), and by means of the brass ring \(K\), that I can distinguish the stationary state of the diminishing bulk of the two mixed airs, above-mentioned: which being perceived, I take off the small tube \(g h\) from the eudiometer, and lay down, for some minutes, the whole instrument, in an horizontal position, under the water of the tin vessel. I shut up the mouth \(V\) with the glass stopple \(m\); and, reverting the instrument, I hang it up by the end \(V\), on the hook \(w\). By this position the whole diminished air of the vessel \(a b c\) goes up to the top, where its real bulk is shown by the number of the scale, facing the inside surface of water. This number being deducted from 128, gives the comparative wholeness of the air already tried, without any further calculation.

"But this process will be still easier, when the last diminution of the two mixed kinds of air, is only required in the observation: because no use will be then made of the siphon (fig. 22.). In such a case, the instrument is left hanging on the hook \(w\) for 48 hours: after which it is laid down under the water of the trough (fig. 15.) in an horizontal position, for 8 or 12 minutes, in order to acquire the same temperature of the water: the mouth \(V\) is then shut up with the stopple \(m\); the instrument is hung by the end \(V\) in a contrary position, and the last real bulk of the good mixed air will be then shown by the number of the brass scale answering to the inside surface of the water. This number being subtracted from 128, will give the comparative salubrity of the air employed in the trial, without any further calculation. I need not say that all the circumstances already mentioned for the better obtaining exact results in these experiments, are to be carefully observed, when this second or the third eudiometers are used: but chiefly that circumstance ought never to be omitted. The thermometer is to be kept dipped in the water of the tin vessel; and the eudiometer must be kept there immersed some minutes, as I have said just now, before it is raised for the last time, to read off the quantity of the total diminution of the mixed air. The same method must be applied to the third new eudiometer I am going to describe; and even the first eudiometer, already described, may be treated in the same manner: for if it be laid down in an horizontal position under the water in the tub, before it be shut up with the stopple, there will be no variation produced by the expansion of the air in the inside; because the proper quantity of water is then shut up within the glass vessel \(c\) of the instrument: so that raising it up, as it is, together with the vessel \(c\), and its phials \(a b\) (fig. 14.), the weight of the column of water will press totally upon them, without expanding the inclosed air, or causing any variation beyond the trifling one which may proceed from the natural elasticity of the sides of the glass tube and vessels.

"I must, however, acknowledge, that the long way through which the air passes, in going at first to the large phial \(a b c\), in this second eudiometer, must leave some doubt whether it has not then suffered some sensible change in its quality before it is mixed with the nitrous air; since, as you have observed, the air that has been long agitated in water, changes for the better from its bad qualities: and this objection must be still greater in the use of the third eudiometer. It is on this account that I have mentioned the first eudiometer, as the least exceptionable of all that we know till the present; and perhaps the nature of the thing is not capable of a further perfection. Indeed that instrument, I mean my first eudiometer, has not only the advantage of offering a very small way through the water to the two kinds of air, on their going to mix at \(x\) in the vessel \(c\) (fig. 14.), but they are kept separate till that moment, in the two respective phials \(a\) and \(b\), without any other contact with the water, but only in the narrow diameter of the necks of these phials.

"The third eudiometer consists of a straight glass tube \(e n\) (fig. 8.) of uniform diameter, about one or two feet long, with a large ball \(s\), and a glass stopple \(m\), fitted air-tight to the mouth \(n\), which ought to be wide open, as a funnel, unless a separate one is made use of. There is also a small siphon (fig. 23.) with a brass ring \(x\); a small phial \(z\) (fig. 9.) the contents of which may be received in the third part of the ball \(s\); and, when put into the glass tube \(n\), must take there no more than the half of its length. Lastly, this instrument has a ruler (fig. 13.) which is divided and stamped like that other already described above; and a glass funnel, which is ground to the mouth \(n\) of the instrument, when this is not wide open, as already said.

"The use of this instrument is easily understood by what I have already said of the two preceding ones, using it. First, it is filled with water, and set in a vertical position, with the mouth \(n\) under the surface of the water in a tub, or in a trough, (fig. 17.) Secondly, the phial \(z\) (fig. 9.) is filled, as above, with nitrous air; and thrown into the tube by means of the glass funnel \(y\) (fig. 10.) which is ground to the mouth \(n\) of the eudiometer; unless it be wide enough not to be in need of any funnel. Thirdly, the same phial \(z\) is again filled with the air to be tried; and thrown into the same. Fourthly, the siphon (fig. 23.) is added immediately to the mouth \(n\) of the eudiometer, under the surface of the water; some of which is to be poured into it. Fifthly, the stationary moment of the greatest diminution of the mixed air at \(s\), is watched by means of the ring \(x\), as above-mentioned. Sixthly, when that moment arrives, the siphon \(K l\) (fig. 23.) is taken off; the eudiometer is laid for some minutes under the water, in an horizontal position, or nearly so, but in such a manner that no part of the inclosed air may get out; EUD

When I want only to know the last diminution of the mixed air, the process then becomes easier, as no use is made of the siphon (fig. 23). The method of conducting the process in such a case being respectively the same as that already described, it is unnecessary to describe it here again. The same precautions I have spoken of, must be observed when this eudiometer is made use of, in order to form a true judgment concerning those places, where people will be able to live without danger of hurting their constitutions by breathing and being continually surrounded by noxious air; which they have not yet been able to distinguish from the most wholesome, except by a long and too late experience.

The eudiometers already described are the fittest instruments for philosophical experiments on the bulk of air and other fluids, when mixed together; and even when mixed with some solid substances, which can be introduced into the lower vessel c of the first of the three eudiometers. It will be better, however, to have them made purposely for such objects, with a tube two or three times longer than I have indicated above. Whenever deplogisticated air is to be tried by these instruments, proper care is to be taken to observe the precise point of its full saturation, which is that of its greatest diminution by the addition of nitrous air. In order to make this experiment with great accuracy, let a narrow glass tube of an uniform diameter (fig. 24) be provided: let one of the two phials a or b (fig. 16) filled with quicksilver, be thrown into it; and the tube cut exactly to that size, so as to contain neither more nor less. Let its whole length be divided into some number of equal parts, by which number the value marked on the ruler (fig. 11) of this eudiometer, can be divided without any fraction: for instance, the number \( \star \star = 108 \), marked in the ruler, means, that the contents of the two phials a and b, are equal to a cylinder of 108 divisions long, as those of the ruler: and, of course, it shows that a single phial a or b contains but 54 of these parts. In this case, this tube (fig. 24) may be divided either into 27 parts, each containing two of the ruler; or into 54, into 108, &c.

N.B. If the top of the tube is not very flat in the inside, it will be more exact, to divide the weight of the quicksilver in two parts; to put one of them into the tube; to mark the space occupied by it; to divide the part of it which was empty, into half the number intended for this tube; and afterwards to divide the other half into similar equal parts, as the first half, carrying them towards the closed end.

If the deplogisticated air is very pure, it will require almost double the quantity of nitrous air to be completely saturated. In order to do this without exceeding the necessary quantity, I throw into the tube n d (fig. 17) a second measure b or a of nitrous air, after I have brought the process to the moment above-mentioned; in this case the whole volume or bulk of the deplogisticated and nitrous air will be 162 \( = 108 \times 54 \); I observe where the surface of the inside water in the tube stops, and I mark it by the sliding brass ring z. I then fill up the divided tube (fig. 24) with nitrous air: I throw a small quantity into the eudiometer tube n d; and, if it becomes of a reddish colour, the inclosed air will diminish: I then push up the ring z; and by this means, I go on throwing in the nitrous air, by little and little, till I see that the whole diminishes no more; which shows me that it is fully saturated. Let us suppose, for example, that the tube (fig. 24) was divided only into 27 equal parts; and that the saturation of the deplogisticated air was completed at the eighth division of it: this shows that 19 parts \( [27 - 8 = 19] \), equal to 38 of those marked in the ruler, have been thrown into the eudiometer; that is to say, that the whole bulk of both kinds of air is equal to 200 \( = 162 + 38 \) such measures as those marked by the divisions of the ruler (fig. 11.) Now if the remaining quantity of air within the eudiometrical tube is only equal to two measures or numbers of the ruler, it is clear that such deplogisticated air is 99 times of 100 \( \frac{200 - 2}{200} = \frac{198}{200} = \frac{99}{100} \) pure air; since its bulk is reduced, by the combination of nitrous air, to the \( \frac{1}{100} \)th of the whole.