[22] Air free from carbonic anhydride indicates after explosion the presence of a small quantity of carbonic anhydride, as De Saussure remarked, and air free from moisture, after being passed over red-hot copper oxide, appears invariably to contain a small quantity of water, as Boussingault has observed. These observations lead to the assumption that air always contains a certain quantity of gaseous hydrocarbons, like marsh gas, which, as we shall afterwards learn, is evolved from the earth, marshes, &c. Its amount, however, does not exceed a few hundredths per cent.
[23] The analyses of air are accompanied by errors, and there are variations of composition attaining hundredths per cent.; the average normal composition of air is therefore only correct to the first decimal place.
[24] These figures express the mean composition of air from an average of the most accurate determinations; they are accurate within ±0·05 p.c.
[25] In Chapter III., Note [4], an approximate calculation is made for the determination of the balance of oxygen in the entire atmosphere; it may therefore he supposed that the composition of air will vary from time to time, the relation between vegetation and the oxygen absorbing processes changes; but as the atmosphere of the earth can hardly have a definite limit and we have already seen (Chapter IV., Note [33]) that there are observations confirming this, it follows that our atmosphere should vary in its component parts with the entire heavenly space, and therefore it must he supposed that any variation in the composition by weight of the air can only take place exceedingly slowly, and in a manner imperceptible by experiment.
[26] The amount of moisture contained in the air is considered in greater detail in the study of physics and meteorology and the subject has been mentioned above, in Chapter I., Note [1], where the methods of absorbing moisture from gases were pointed out.
[27] Soda-lime is prepared in the following manner:—Unslaked lime is finely powdered and mixed with a slightly warmed and very strong solution of caustic soda. The mixing should be done in an iron dish, and the materials should be well stirred together until the lime begins to slake. When the mass becomes hot, it boils, swells up, and solidifies, forming a porous mass very rich in alkali and capable of rapidly absorbing carbonic anhydride. A lump of caustic soda or potash presents a much smaller surface for absorption and therefore acts much less rapidly. It is necessary to place an apparatus for absorbing water after the apparatus for absorbing the carbonic anhydride, because the alkali in absorbing the latter gives off water.
[28] It is evident that the calcium chloride employed for absorbing the water should be free from lime or other alkalis in order that it may not retain carbonic anhydride. Such calcium chloride may be prepared in the following manner: A perfectly neutral solution of calcium chloride is prepared from lime and hydrochloric acid; it is then carefully evaporated first on a water-bath and then on a sand-bath. When the solution attains a certain strength a scum is formed, which solidifies at the surface. This scum is collected, and will be found to be free from caustic alkalis. It is necessary in any case to test it before use, as otherwise a large error may be introduced into the results, owing to the presence of free alkali (lime). It is best to pass carbonic anhydride through the tube containing the calcium chloride for some time before the experiment, in order to saturate any free alkali that may remain from the decomposition of a portion of the calcium chloride by water, CaCl2 + 2H2O = CaOH2O + 2HCl.
[29] Recourse is had to special methods when the determination only takes note of the carbonic anhydride of the air. For instance, it is absorbed by an alkali which does not contain carbonates (by a solution of baryta or caustic soda mixed with baryta), and then the carbonic anhydride is expelled by an excess of an acid, and its amount determined by the volume given off. A rapid method of determining CO2 (for hygienic purposes) is given by the fall of tension produced by the introduction of an alkali (the air having been either brought to dryness or saturated with moisture). Dr. Schidloffsky's apparatus is based upon this principle. The question as to the amount of carbonic anhydride present in the air has been submitted to many voluminous and exact researches, especially those of Reiset, Schloesing, Müntz, and Aubin, who showed that the amount is not subject to such variations as at first announced on the basis of incomplete and insufficiently accurate determinations.
[30] It is a different case in enclosed spaces, in dwellings, cellars, wells, caves, and mines, where the renewal of air is impeded. Under these circumstances large quantities of carbonic anhydride may accumulate. In cities, where there are many conditions for the evolution of carbonic anhydride (respiration, decomposition, combustion), its amount is greater than in free air, yet even in still weather the difference does not often exceed one ten-thousandth (that is, rarely attains 4 instead of 2·9 vols. in 10000 vols. of air).
[31] In the sea as well as in fresh water, carbonic acid occurs in two forms, directly dissolved in the water, and combined with lime as calcium bicarbonate (hard waters sometimes contain very much carbonic acid in this form). The tension of the carbonic anhydride in the first form varies with the temperature, and its amount with the partial pressure, and that in the form of acid salts is under the same conditions, for direct experiments have shown a similar dependence in this case, although the quantitative relations are different in the two cases.