From the time of Van Marum (1785) at least it was known that the passage of electric sparks through air is accompanied by the production of a peculiar smell; but the cause of this remained unknown until 1840, when Schönbein observed that a similar smell is exhibited by electrolytic oxygen (as obtained in the electrolysis of acidulated water), and also develops in the atmosphere of a vessel in which phosphorus suffers spontaneous oxidation at ordinary temperatures in the presence of water. The three kinds of odoriferous gas, he found, had the power of decomposing iodide of potassium with liberation of iodine, and they agreed also in their behaviour to other reagents, whence he concluded that in all the three cases the smell was owing to the same peculiar substance which he called ozone (from oζεw, to emit an odour). Numerous experiments confirmed his first impression that ozone is chemically similar to, though distinctly different from, chlorine, but he got no further towards establishing its nature. Having found, however, that dry phosphorus produces no ozone, and that ready-made ozone is destroyed by being passed through a heated glass tube, he surmised that ozone was a peroxide of hydrogen. This surmise was seemingly raised to a certainty by an investigation of Baumert’s, who found that electrolytic (ozonized) oxygen, when deozonized by heat, yields water, and ascertained that the weight of water thus produced amounted to H 2 O = 18 parts for every 41 = 4 × 127 parts of iodine which the same quantity of gas would have liberated if it had been deozonized by iodide of potassium. This, if true, would prove that ozone is H 2 O 3 ,—a conclusion which passed current as an established fact, in reference to electrolytic ozone at least, until Andrews showed that Baumert’s result was founded upon incorrect observations. The merit of having discovered the true elementary composition of ozone belongs to Marignac and De la Rive, who proved that it can be produced, as easily and abundantly as in any other way, by the electrification of absolutely pure oxygen gas, whence it at once followed that—unless oxygen be a compound of two or more unknown elements—ozone cannot be anything else than an allotropic modification of oxygen.

With regard to the relations of the two kinds of oxygen to one another, our present knowledge is derived mainly from the work of Andrews and Prof. Tait. The first important result which they arrived at was that the ozonization of pure oxygen gas involves a contraction, and that consequently ozone is denser than oxygen gas. Presuming (with all their contemporaries) that in the de-ozonization of oxygen by iodide of potassium all the substance of the ozone is taken up by the reagent with elimination of its equivalent of iodine, they sought to determine the density of ozone by comparing the weight of oxygen-matter which goes into the iodide of potassium with the contraction involved in the process. But they obtained variable results. As their methods became more and more perfect, the weight of unit volume of ozone grew greater and greater, and at last stood at ∞. In other words, what they found and established finally was that the removal of ozone from oxygen by means of -iodide of potassium involves no change of volume whatever, although de-ozonization by heat always leads to a (permanent) increase of volume. This result, to them and everybody else, appeared very singular; but Andrews, after a while, found the correct explanation. Supposing at a certain temperature and pressure one volume of ordinary oxygen contains a grains of matter, then one volume of ozone, being denser, contains a greater quantity of matter, say a+x grains; when the gas acts on iodide of potassium, the a grains come out as one volume of oxygen, while the x grains of surplus oxygen vanish in the iodide. In the decomposition by heat the x grains of surplus oxygen of course assume the form of a∙∕α volumes of additional oxygen gas. It is no addition to Andrews’s explanation, but merely a close translation of it into the language of Avogadros law, to say that, if oxygen (proper) consists of molecules O 2 , ozone must consist of molecules O 2 ⅛ (perhaps O 2+ 1), and tħat in the iodide reaction this molecule breaks up into one molecule of oxygen gas and x atoms of oxygen which go to the reagent. What did constitute a new discovery w¾s Berthelot’s important observation that the conversion of ozone into ordinary oxygen involves an evolution of heat which amounts to 29,600 units for every 16 parts of oxygen matter available for the liberation of iodine from iodide of potassium. What the real density of ozone is was made out with a high degree of probability by Soret. He took two equal volumes of the same supply of ozonized oxygen, and in one determined the contraction produced by shaking with oil of turpentine (which he assumes to take away the ozone as a whole), while the other served for the (direct or indirect) determination of the expansion involved in the destruction of the ozone by heat. He found this increase to amount to half a volume for every one volume of ozone present; hence one volume of ozone contains the matter of one and a half volumes of ordinary oxygen, i.e., its density is 1'5 (if that of ordinary oxygen is taken as unity), and its molecular weight is ⅜× O 2 =O 3 ∙ To check this result Soret determined the rate at which ozone diffuses into air, and compared it with the rate, similarly determined, for carbonic acid. From the two rates, on the basis of Graham’s law, he calculated the ratio of the density of ozone to that of carbonic acid, and found it in satisfactory accordance with O 3 : CO 2 = 48: 44.

From the facts that ozone is destroyed (z.e., converted into O 2 ) at 270° (Andrews and Tait), and that this reaction is not reversible, it at once follows that it is impossible to convert oxygen completely into ozone by electric sparks. Supposing the ozonization to have gone a certain way, each additional spark, besides producing ozone, will destroy some of that previously produced.

From Clerk Maxwell’s notion concerning the distribution of temperatures amongst the molecules of a gas, it would follow that ozonized oxygen, even at ordinary temperatures, will gradually relapse into the condition of plain oxygen, because, although the temperature as indicated by the thermometer may be only 20° C. (say), there are plenty of molecules at temperatures above the temperature of incipient dissociation (which of course lies below 270 c ), and any ozone once destroyed will never come back. But, be this as it may, the lower the temperature of the oxygen treated with sparks the greater the chance of the ozone formed to remain aliVe. This idea forms the basis of an important research by Hautefeuille and Chappuis, who, by operating upon oxygen at very low temperatures, produced unprecedentedly large percentages of ozone. By operating at 0° C. they produced a gas containing 14 - 9 per cent, by weight of ozone (presumably reckoned as O 3 ), λvhile at - 23° the percentage rose to 21 ’ 4. They subsequently (1882; Compt. Rend., xciv. p. 1249) succeeded in producing even liquid ozone, by applying a pressure of 125 atmospheres to richly ozonized oxygen at -100° C. (the boiling point of liquefied ethylene). Liquid ozone is of a dark indigo-blue colour, which, as they tell us, is distinctly visible even in ordinary ozonized oxygen if it is viewed in tubes about one metre long.

According to Carius the coefficient of absorption of ozone by water of +1° C. is about 0 - 8; that is to say, one volume of water of 1°, if shaken with excess of pure ozone at 1° and a pressure of 760 mm., would absorb 0’8 volume of ozone measured dry at 0° and 760 mm. pressure. But it is not certain that Carius’s determinations are correct.

Antozone.— According to a now obsolete notion of Schönbein’s, ordinary oxygen gas is a compound of two kinds of oxygen of which one is positively and the other negatively electrical. Ordinary ozone would be a mixture of the two in equal parts; but certain peroxides, according to Schönbein, contain the one kind, others the other. He supported his view by many ingenious experimental arguments. Meissner and others, while adopting Schönbein’s idea, somehow drifted into the notion that Schönbein’s two kinds of oxygen correspond to two different substances, of which ordinary ozone is one. They naturally searched for the other, and of course did not fail to discover it; but their “antozone,” when critically looked into, turned out to be peroxide of hydrogen. ’