ence of this impurity was deemed so probable, that in our first experiments, a small vessel containing nitrate of silver was interposed in the current of issuing gas. Finding no milkiness to arise in the time required for the experiment, we prolonged the trial to a period of several hours, during which a stream of gas from the gasometer was passed in bubbles through the nitrate. But no precipitate was formed. This result, many times repeated, sufficiently attests the absence of hydrochloric acid from the gas. It is important, however, to remark, that using a much larger proportion of hydrochloric acid with the water of the reservoir, distinct traces of this substance may be detected in the issuing gas, and to remove all chance of error, therefore, after each new charge of acid, the gas was carefully tested by transmission through the solution of nitrate of silver.

Of the Hygrometric State of the Gas.-Excepting in the experiments on sulphuric and some other acids, the gas was transmitted through the long submerged tube into the flask, without desiccation. This course was rendered necessary by the fact, that in repeated trials, with gas previously dried and saline solutions, there occurred an irregular expansion of the gaseous volume in the first stage of the action, which did not show itself when the undried gas was used. Such an enlargement, due evidently to the rise of aqueous vapor into the dry space in the first moments of the agitation, being of variable amount according to the solution used, would form a serious obstacle to the exact measurement of the absorption.

In the case of the sulphuric acid and other bodies referred to as exceptions, this would not take place. On the contrary, the presence of aqueous vapor in the gas would here involve other errors, due to the absorption of the vapor by the liquid, or to the heat disengaged by their reaction.

The drying of the gas being thus prohibited in a great majority of the experiments, it became important to ascertain whether the carbonic acid, coming from the gasometer, was saturated with vapor, as in this case, from the observed absorption of the moist gas, it would be easy to compute the amount of dry gas which had actually disappeared.

For this purpose, the acid solution in the gasometer was allowed to continue its action until it became entirely neutral. A measured volume of the gas was then passed very slowly through a long drying tube of chlorid of calcium, previously counterpoised. By preliminary trials with an additional smaller tube, similarly charged, it was ascertained that scarcely a trace of moisture escaped absorption in the long tube. Before the second weighing of the latter, it was freed from carbonic acid by aspiration, the smaller tube being attached to prevent the entrance of atmospheric moisture. In repeated experiments thus performed, the

weight of vapor contained in the gas was found to correspond closely with that proper to the temperature and a state of saturation. In other words, the vapor mingled with the gas was at its maximum tension. As in its neutral state, the liquid of the gasometer contained most dissolved matter, it was to be inferred that the effect upon the tension of the vapor rising from it would then be most perceptible, and that therefore in the working condition of the apparatus, the saturation of the vapor could not be less, although it might be a small fraction more. Similar experiments with the gas under these conditions gave us, however, the same results. We concluded, therefore, that the dissolved matter in the gasometer, is not in sufficient quantity to produce a sensible modification of the tension of the aqueous vapor evolved, and that in all our experiments, we may assume the gas to be saturated with vapor proper to the temperature at which we operate.

Of the Correction for Moisture. This being deduced from the pressure of the atmosphere and the vaporous tension jointly, requires a record of the barometer for each experiment. By the equal adjustment of the columns in the measuring tube, the entire tension of vapor and gas together, is the same at the close as at the beginning of the experiment, and is measured by the height of the barometer. The tension of the vapor remains unchanged, because it is vapor of saturation, and is condensed into water in proportion as the gaseous space contracts in the progress of the absorption. If therefore V represent the apparent absorption, or the volume which has disappeared, and v the volume of dry gas in V, estimated under the full atmospheric pressure; and if p denote that pressure, in other words, the height of the barometer, and f the tension of the vapor proper to the temperature, we have v = y.P-f

Р

It is important to remark, that the tension of the gas under which this absorption takes place, is p-f, and not p, and that in tabulating the results, the corrected absorption should refer to the actual pressure of the gaseous atmosphere in the flask, and not to the entire atmospheric pressure.

From experiments upon the absorption of carbonic acid gas at various pressures by water, Dr. Henry, as is well known, was led to infer that equal volumes are absorbed at all pressures, or what is the same thing, that the quantities of gas absorbed are exactly proportioned to the pressures. This very simple law, if true, would render the correction for moisture superfluous. For in that case, the volume absorbed at the pressure p-f, would equal the volume absorbed at p. Thus V, the apparent absorption, that is, the volume disappearing at the pressure p, consisting partly of gas and partly of vapor, would be precisely the same as the volume of dry gas alone which would disappear at the same SECOND SERIES, Vol. VI, No. 16.-July, 1848.

14

pressure. But further experiments are, we think, needed, to determine with precision the law of absorption as dependent on pressure, and in the mean time, the law of Henry can only be looked upon as approximately true. From observations on this subject in which we have lately been engaged, and which we hope to continue, we have been led to infer that, in comparing widely variant pressures, there is a marked departure from this law.

Although therefore, from the small difference of gaseous pressure (that between p and p-f) in our experiments, we believe that no sensible error could be introduced by applying the law of Dr. Henry to the results, we have thought it proper in reporting them, to state the volume of dry gas absorbed and the reduced pressure, as well as the apparent absorption and entire barometric pressure.

Having now presented all the details of our mode of operating, and of the precautions and corrections we have used, we proceed to give an account of the results, treating of them in the following order:

I. Of the absorption by water.

II. Of that by sulphuric and other acids, and by other unmixed liquids.

III. Of that by various saline aqueous solutions.

I. Absorption of Carbonic Acid by Water.-The water used in these experiments, as well as in making the solutions employed in others to be described hereafter, was prepared by careful distillation in a copper vessel. Its purity was such, that several cubic inches evaporated in a platinum capsule, gave no indication of alkaline matter to the most delicate test paper, and when entirely volatilized, left scarcely a trace of residuum. Before being used, it was briskly boiled for half an hour, quickly transferred to a well stopped bottle, and when sufficiently cooled, exposed to the exhausting action of a good air-pump. The bottle was then suspended in the large reservoir, to bring it to the proper temperature, before the charge was introduced into the flask.

The absorption was seen to begin as soon as the first drop descended from the flask, and with brisk agitation, the process was completed in about five minutes after the liquid was brought in contact with the gas. To satisfy ourselves that no further absorption would occur, we repeatedly prolonged the agitation to fifteen or twenty minutes, allowed the apparatus to rest, and again resumed the shaking, but without producing any appreciable change in the column of the measuring tube.

Although, from the purity of the gas used, the closeness of the apparatus, and the care with which it was charged with gas, we had no reason to apprehend any dilution of the CO,, yet as such a change would cause the absorption to terminate short of the saturation of the liquid proper to an unmixed atmosphere of the

gas, experiments were made to determine if any further absorption was caused by a renewal of the charge. This was done by removing the flask, driving a stream of CO, into the bottle, closing the orifice by an air-tight stopper, readjusting the levels, and submitting the liquid to further agitation. Repeated trials at 60°, gave no indications of additional absorption. We would therefore regard our results as furnishing a nearly accurate measure of the absorption of carbonic acid by pure water.

These results, together with the conditions under which the observations were made, are comprised in the following table.

Table of the Absorption of Carbonic Acid by Water, from 32° to 212°.

No. Temp. External of of CO2 thermomEx. & HO.

Mean abs. Tension of dry CO2

=v.

1

eter.

of gas

Abs. reduced

100....HO.

=V.

32

54

29.48

166.5

40

40

50

50

60 64.5 29.42 100-5

12

60

On comparing the second and third columns in the above table, it will be seen that, in the observations from 50° to 80° inclusive, the temperature of the contiguous air in no case differed from that of the apparatus by more than 50-5. Hence during the few minutes occupied in each experiment, the temperature even of the smaller reservoir experienced only a very slight and quite unimportant change, amounting in none of the experiments to as much as one-tenth of a degree. In the experiments at 40°, it was found easy to maintain a uniform temperature by a few fragments of floating ice, and in those at 32°, the use of a large amount of ice in both vessels, preserved the temperature entirely unchanged. The observations at 90° and 100°, were attended with a slight cooling in the small reservoir, which however, in no case exceeded one degree, an amount too small to produce any measurable change in the mercurial column.

The above table presents we believe the first systematic series of observations on the comparative absorption of CO2, by water at

different temperatures, yet made known. The experiments of Dalton, Henry, Manchester and Saussure, were made almost exclusively at 60°. The only results, referring to other temperatures, which we have seen numerically noted, are one by Cavendish at 55° and one by Henry at 85°. According to Cavendish the absorption by one hundred volumes of water at 55° is one hundred and sixteen. In Henry's experiment the same volume of water at 85°, is said to have absorbed eighty-four volumes of the gas. The latter result departs very widely from the mean of our experiments at 80° and 90°, which is about sixty-six volumes instead of eighty-four. The experiment seems to have been made with little care and merely to test the effect of a higher temperature upon the amount of absorption. The number obtained by Cavendish in his observation at 55°, corresponds more nearly with our results, which, taking the mean of the experiments at 60° and 50°, would be about one hundred and ten, instead of one hundred and sixteen, the number which he has given.

In the more numerous and important experiments at 600, the observed absorption as given by Saussure, is one hundred and six, by Henry, one hundred and eight, and by Dalton, one hundred. The two former present a marked excess over our result, the latter agrees with it very closely. The larger absorption obtained by Saussure and Henry, is we think explained by their mode of conducting the experiment. We have found that when a column of mercury is shaken briskly in a tube containing water and carbonic acid, the water is made to absorb a larger volume of the gas than is proper to the normal pressure. The concussive movement, violently compresses the gas at each vibration, and the additional quantity which in these circumstances is promptly taken up by the water, is very slow in separating after the quiescent pressure has been restored.

Referring to the arrangement of the preceding table, it will be seen that the numbers in the 7th column express the absorption, reduced to volumes of dry gas and to the density corresponding top in the 4th column. The obvious formula for this has already been explained. The numbers in the 8th column, represent the actual tension of the gas under which the absorption took place. These two columns give the direct experimental relation of the absorption of dry gas with the tension of the same. But assum

ing Henry's law to be correct, and in the present case it can involve no sensible error, this relation would be equally expressed by the corresponding numbers in columns 4 and 6. Thus while it is clearly proved, from the observations at 50°, that under the pressure 29.1-p-f, 118.4 volumes of dry gas are absorbed, it would also be true that under the pressure 29.46 p, 120 volumes of dry gas would be absorbed, for p: V-p-f: v.