109

110

111

be examined. It is found to be a colourless, odourless, gas, a very little lighter, bulk for bulk, than ordinary air; it does not support combustion, nor is it combustible; it reacts chemically with but few elements and compounds. Every attempt to separate a specified mass of this gas into unlike parts has failed. But very many compounds are known of each of which this gas is a constituent. The gas is an element; it is called nitrogen.

We know that oxygen is also an element. Hence we have obtained from air two elements nitrogen and oxygen.

Is air a compound or a mixture of these gases? If it is a compound, the properties of air must differ considerably from the properties of either oxygen or nitrogen; and these elements must be united in air in a ratio expressed by the formula NO, where x is 1, 2, 3, 4...n times the combining weight of nitrogen (14), and y is 1, 2, 3, 4...n times the combining weight of oxygen (16). If air is a mixture of nitrogen and oxygen, it must be possible to recognise both of these elements in air by making use of the properties which each possesses when unmixed with other kinds of matter.

Whichever hypothesis is adopted as a guide in experimental inquiry, we must begin by determining the properties of air, the properties of oxygen, and the properties of nitrogen. We already know some of the properties of oxygen and nitrogen. Both are colourless, odourless, gases; nitrogen is 14 times, and oxygen is 16 times, heavier than hydrogen. Combustible bodies burn rapidly and brilliantly in oxygen, but they cease to burn in nitrogen. The ratios of diffusion of both are nearly equal; but oxygen passes through a thin sheet of india-rubber about 2 times more rapidly than nitrogen. Oxygen is slightly soluble, nitrogen is less soluble, in water; 1 vol. of water at 16° dissolves 0295 vols. of oxygen, and 0145 vols. of nitrogen. The combining weight of oxygen is 16, and the combining weight of nitrogen is 14.

The prominent physical properties of air are known to all. Accurate analyses of air have shewn that 100 parts by weight of dry air freed from carbon dioxide (v. infra, par. 113) are composed of 23 parts of oxygen and 77 parts of nitrogen, by weight. The simplest formula which will fairly accurately represent this composition, assuming air to be a compound, is NO13; this compound, if it existed, would be composed of 22.5 parts of oxygen by weight, and 77.5 parts of nitrogen, per 100 parts. Five definite compounds of nitrógen and oxygen are known;

51

51 13°

their compositions are represented by the formulae NO, NO, NO, NO, NO,. It is improbable that a sixth compound of these elements should exist having as complex a composition as N12O, But this is the simplest formula which can be given to air if air is a compound of nitrogen and oxygen. Hence the argument based on analogy of composition leads to the conclusion that air is probably a mixture and not a compound.

Assuming air to be a mixture of nitrogen and oxygen, we next inquire, what volume of air ought to be dissolved by 1 vol. of water, say at 16°? The solution of a mixture of gases by a liquid between which and the gases there is no chemical interaction follows the same course as if each gas were dissolved separately in the liquid. The solution of a gaseous compound, on the other hand, in a liquid which does not interact chemically with the compound follows a course of its own; the vol. dissolved is independent of the vols. of the gaseous constituents of the compound dissolved under the same conditions.

1 vol. of water at 16° dissolves '0295 vols. of oxygen, and 0145 vols. of nitrogen; now 1 vol. of dry air freed from carbon dioxide (v. infra, par. 113) is composed of 2096 vols. of oxygen and 7904 vols. of nitrogen; therefore, if air is a mixture, 1 vol. of water will dissolve (0295 x 2096) + (0145 x 7904) = 01765 vols. of air, at 16o.

=

Experiment proves that 1 vol. of water dissolves '0177 vols. of air at 16o.

1 vol. of water at 16° dissolves 7535 vols. of nitrous oxide (NO); but if this gas were a mixture of nitrogen and oxygen in the ratio in which these gases unite to form 1 vol. of nitrous oxide, water would dissolve 02925 vols. of the nitrous oxide.

These calculations and experiments shew that air is dissolved by water exactly as if the air were a mixture of oxygen and nitrogen and not a compound of these elements. In other words: one of the properties of oxygen is to dissolve in water to a certain definite extent, and one of the properties of nitrogen is to dissolve in water to a certain definite extent; but both oxygen and nitrogen retain this property when they are present in air; therefore air is a mixture, and not a compound, of oxygen and nitrogen.

We may carry the inquiry further on the same lines. If air is a mixture of oxygen and nitrogen, and if oxygen passes through a thin sheet of india-rubber about 2 times quicke

112

than nitrogen it ought to be possible to effect a partial separation of air into its constituent gases by passing it through a sheet of india-rubber. Experiment proves that when all, or almost all, the air is pumped out of an india-rubber bag, and the bag is closed and left in the atmosphere, air passes into the bag through the walls, and that the composition of the air found in the bag is approximately 40 p. ct. of oxygen and 60 p. ct. of nitrogen, by volume. But the composition of ordinary air is approximately 21 p. ct. oxygen and 79 p. ct. nitrogen, by volume. Therefore the air has been partially separated into its constituents by passing it through a sheet of india-rubber; therefore air is a mixture, not a compound, of oxygen and nitrogen.

The argument may be extended to chemical events. If air is a mixture, it ought to interact chemically with other substances both as oxygen interacts and also as nitrogen interacts. If air is a compound, its interactions with other substances ought to be different from those of either oxygen or nitrogen. We have learned that nitrogen is a very inert substance; it does not support combustion, it is not combustible, it combines directly with only a few elements, and it does not react chemically with many compounds. On the assumption that air is a mixture, we should, therefore, expect its chemical properties to resemble those of oxygen, but to be less strongly marked because of the presence of the inert nitrogen.

For instance, we should expect substances which burn rapidly and brilliantly in oxygen to burn in air but to burn more slowly and less brilliantly. If we can find an element which combines directly with nitrogen when heated in that gas, we should expect that element to form a compound with nitrogen when strongly heated for some time in air.

We need not go into details regarding individual experiments, but suffice it to say that these expectations are realised; that the chemical behaviour of air is exactly what the hypothesis of its being a mixture asserts ought to be its behaviour. Air then is a mixture, not a compound, of oxygen and nitrogen. 113 But besides these gases, air contains small quantities of the compound gases, carbon dioxide, ammonia, and water

vapour.

The composition of air varies within narrow limits. Thus air has not that fixity of composition which as we have seen characterises chemical compounds.

The experiment described in par. 108 shewed that the composition of air is, roughly, 4 vols. of nitrogen to 1 vol. of oxygen. In order accurately to determine the volume-composition of air, a quantity of air is passed into a graduated glass tube fitted with two platinum wires passing through the glass near the closed end; the tube is filled with mercury, and is then inverted in a trough containing mercury. The air to be analysed is freed from carbon dioxide and ammonia, and is then passed into the tube, and the volume is measured by reading off the level of the mercury. A quantity of hydrogen equal in volume to nearly of the volume of air is passed into the tube, and the level of the mercury is again read off; the tube is pressed down on a pad of india-rubber and securely clamped; an electric spark is then sent from one platinum wire to the other; the effect of this is that the whole of the oxygen in the air combines with a portion of the hydrogen to produce water which condenses. After a little time the tube is slowly raised from the india-rubber pad; mercury rushes in; the level of the mercury is read off. As we know that 2 vols. of hydrogen combine with 1 vol. of oxygen, we conclude that of the diminution of volume which occurs when the spark is passed represents the volume of oxygen in the volume of air employed. The volume occupied by the small quantity of water produced is so small that it may be neglected. Many precautions are necessary in carrying out such an analysis as this; corrections must be made for temperature and pressure ; the volumes of wet air and wet gas after the explosion must be reduced to the corresponding volumes of dry gases, &c.

The carbon dioxide in air may be determined, (1) by slowly passing a large measured volume of air, freed from ammonia and water-vapour, through a series of weighed U tubes filled with caustic potash, and determining the increase in the weight of these tubes; or (2) by adding, to a measured volume of air, a known quantity of barium oxide dissolved in water, and determining the quantity of this oxide which remains when the carbon dioxide in the air has all been absorbed by a portion of the barium oxide. The chemical reactions on which these methods are based may be represented in equations thus ;

2

2

3

(1) KOH (moist) + CO2 = K ̧CO2+ H ̧O + (x − 2) KOH. (2) xBaOAq + CO2 = BaCO ̧ + (x − 1) BaOAq.

The potassium carbonate (K,CO,) and water (H,O) formed

in (1) remain in the weighed U tubes along with the potash (KOH) which has not been changed by the carbon dioxide (CO): the barium carbonate (BaCO,) formed in (2) is a solid, it settles down in the liquid, and the unchanged barium oxide (BaO) remains in solution and is determined by a method which need not be described here.

114 The quantities of oxygen and nitrogen in average country air freed from water-vapour, ammonia, and carbon dioxide,

115

The quantity of carbon dioxide averages about '03 volumes per 100 vols. of air. The quantity of ammonia varies very much; it may perhaps be taken as about 1 part in 10,000,000 parts of air, by weight. The quantity of aqueous vapour also varies with variations in the season, the district, &c. &c.

A

The fact that the quantities of oxygen and nitrogen in country air vary, although within very narrow limits, has been definitely established. The oxygen sometimes amounts to 20.999 vols. per 100 e.g. in air from the seashore or from inland moors; in towns the oxygen sometimes falls to 20.82 vols. ; in inhabited rooms and crowded halls it may be as little as 20.28; in mines it averages about 20-26. decrease in the volume of oxygen is usually accompanied by an increase in that of carbon dioxide; in crowded rooms the volume of this gas may be as large as 3 to 5 vols. per 100. Air which contains as much as 1 vol. carbon dioxide per 100 is unpleasant, and harmful to health. The air of towns contains many gases, liquids, and solids, produced by the changes which go on among the living beings, and also by the manufactures conducted in the towns.

Our examination of air has afforded an application of the statements made in Chap. III. regarding the differences between mixtures and compounds; it has shewn us how we may determine to which of these classes a given substance belongs; it has also made us acquainted with some of the prominent characters of air; and it has a little familiarised us with the methods pursued in chemical inquiries.