entirely into common oxygen; and finally, it is destroyed by black oxide of manganese and some other substances, which are not themselves oxidized by it.

All these, and a great number of similar facts, we owe to Schönbein, who has pursued his discovery with a perseverance and zeal which cannot be overrated. It is true that his researches have been almost entirely qualitative in their character, and that he has therefore arrived at erroneous conclusions as to the causes of the phenomena which he has studied; but this circumstance must not induce us to forget the gratitude we owe him for his very good and useful work. He stands to ozone something in the relation that Priestly stood in to oxygen, while the exact labours of Lavoisier have their parallel in those of Fremy and Becquerel, Andrews and Tait, Soret, and several other investigators.

Passing by the crude theories of the earlier workers on ozone, we find that the first step towards the true theory was made by Marignac and De la Rive, who proved that ozone contained no other element than oxygen, and consequently that it could only be some altered, or "allotropic" form of that element. In 1852 another important stage in the progress was reached by Becquerel and Fremy, who not only confirmed the conclusions of Marignac and De la Rive, but demonstrated that pure oxygen might, by the prolonged action of electricity, be converted entirely into ozone. It is true that this can only be effected when the ozone is absorbed as fast as it is produced-when, for instance, the electric sparks are passed through a tube of oxygen inverted over mercury or iodide of potassium-and that it has hitherto been found impossible to prepare ozone free from common oxygen, but the total conversion is none the less significant of the true nature of the substance. In 1856 Dr. Andrews showed conclusively that ozone was the same by whatever process it was prepared, and finally demolished the arguments by which Williamson and Baumert had sought to prove that it was a teroxide of hydrogen.

So far the investigation of ozone had been of a satisfactory character. Erroneous experiments had indeed been made, and incorrect theories had, as a matter of course, been founded upon them. But still the theories followed logically from the supposed facts, and the experiments themselves were conducted with care and honesty. But Schöbein, although he scarcely touched the quantitative methods, on which alone a theory can be safely based, ventured, in a letter to Faraday, dated June 25, 1858, to propound a new hypothesis to which he has ever since adhered, and which has never

wanted advocates, although it has been distinctly and formally disproved. He assumed the existence of two different and opposite kinds of oxygen, a negative and a positive kind. The former-the variety obtained by electricity, the oxidation of phosphorus, etc.he continued to call ozone, the latter he distinguished as antozone, and he asserted that common, or neutral oxygen was formed by the union of both kinds. These hypothetical constituents of oxygen were furthermore assumed to exist in a great variety of oxides. Those which contained ozone were called ozonides, and among them were numbered the higher oxides of manganese, chromium, and iron, and the oxides of the noble metals. The opposite class of oxides, the antozonides, comprised the peroxide of the metals of the alkalies and alkaline earths, the peroxide of hydrogen, and a few other substances. This elaborate hypothesis was based almost entirely upon the circumstance that when one of the so-called ozonides was mixed, under suitable conditions, with an antozonide, common oxygen was evolved, the ozone of one, according to Schönbein, combining with the antozone of the other. But without denying that Schönbein's hypothesis is capable of accounting for facts like these, we are compelled by the researches of Sir B. C. Brodie to believe that they can be explained as satisfactorily and more simply by a reference to the ordinary laws of chemical change. The recently ascertained density of ozone is, moreover, quite inconsistent with the truth of Schönbein's hypothesis, which, indeed, I should not have referred to here, but that it is still accepted by a considerable section of scientific men.

We come now to a much more important and genuine set of discoveries. In 1860, Andrews and Tait published in the" Philosophical Transactions" a paper "On the Volumetric Relations of Ozone," which must be regarded as the most important memoir upon the subject which has been produced since the original discovery by Schönbein. The authors found that, during the formation of ozone by the passage of the electric discharge through oxygen, a contraction of volume took place, and, consequently, that ozone must be heavier than oxygen. The amount of contraction was in direct proportion to the amount of ozone formed. It was greatest when the silent electrical discharge was employed, which likewise developed the greatest quantity of ozone; but in no case did it exceed onetwelfth of the original volume of the oxygen. On heating the gas, so as to destroy the ozone, the original volume was exactly restored. They then set to work to determine what further contraction of volume could be produced by the removal by mercury, or some

other absorbent, of the ozone which had previously been produced. This second contraction, they anticipated, would give the volume of the ozone which had been absorbed by the mercury; and as its weight could be easily ascertained by finding the amount that the mercury had gained, it would be easy to find the actual density of ozone. The result is a powerful example of the way in which experiment often contradicts hypothesis. The removal of the ozone did not alter in the slightest degree the volume of the gas. An imaginary example will make this more evident. We take :

100 cubic inches of oxygen.

By the action of the electric discharge this is reduced to
92 cubic inches of ozonized oxygen,

which is really a mixture of ozone and oxygen. After the absorption of the ozone by mercury there remains

92 cubic inches of oxygen.

So that the ozone appears to occupy no volume at all, and its density to be absolutely infinite. This curious experiment was repeated by the authors in several forms, but the result was always the same. They honestly expressed themselves surprised and puzzled at the phenomenon, and were very guarded in their attempts to account for it. New light, however, soon poured in. The experiments were too striking and too carefully performed to remain long unfruitful; and the very absurdity which they seemed to involve suggested to the acute mind of Dr. Odling a simple solution of the problem. To appreciate the force of this solution, it is necessary to bear in mind the theoretical conception of the nature of gases which modern investigation has enabled chemists to frame. To prevent unnecessary digression, I will state it dogmatically, giving none of the reasons which have led to its adoption, and omitting the real or apparent exceptions, which are, indeed, very few in number.

Every gas, whether elementary or compound, consists of minute particles called molecules. The molecules of all gases, whether elementary or compound, have an equal size, and, at the same temperature and pressure, a given volume contains always the same number of them. Hence all gases are equally affected by purely physical operations, such as increase or diminution of temperature or pressure. The differences between gases depends entirely upon the nature, or, so to speak, upon the structure of the molecules. The molecules are in reality aggregations or clusters of ultimate indivisible atoms. The nature, the number, and the arrangement of the atoms in each molecule determines its weight and its properties.

Elementary molecules contain atoms of only one kind, the number differing in different elements. Thus the molecules of mercury and some other elements contain but one atom; the molecules of hydrogen, oxygen, potassium, etc., two atoms; and the molecules of phosphorus and arsenic four atoms. The molecules of compound gases contain two or more different kinds of atoms, the total number of which may be only two, but may amount to sixty or eighty, or even more. The formulæ used by chemists are now invariably constructed so as to denote one molecule of each element or compound, each symbol denoting one atom. Thus, Hg. H2. 03. P‘., represent single molecules of mercury vapour, hydrogen, oxygen, and phophorus vapour; and HCl. H2O. H3N., single molecules of hydrochloric acid, steam, and ammonia. The hypothesis is, of course, only a convenient explanation of well-known and certain facts; but even if the atomic theory were abandoned, the formulæ might still be used to express the facts.

Dr. Odling's theory of ozone may now be given in very few words. The molecule of oxygen contains two atoms, the molecule of ozone contains three; so that the formation of the latter body simply means the condensation of oxygen into two-thirds of its former volume. As the formula for oxygen is O', so that of ozone is 03, and its oxidizing power is due to the ease with which each molecule loses its third atom of oxygen. On this view Andrews' and Tait's results become mere matters of course, as may easily be seen by returning to our former example.

100 cubic inches of oxygen yield 92 cubic inches of ozonized oxygen, because 8 cubic inches coalesce with 16 to form 16 cubic inches of ozone.

If the gas is heated, the original volume is restored, because the 16 cubic inches of ozone, O', yield 24 cubic inches of oxygen, O'.

When the ozone is absorbed by mercury, it is really only the third atom which combines with the mercury, the 16 cubic inches of ozone therefore become 16 cubic inches of oxygen, and the volume remains unaltered

This beautiful hypothesis, although accounting perfectly for all known facts, was yet, nevertheless, but a probability. One link was lacking in the chain of evidence, and it is just that link which M. Soret has supplied by a happily-devised experiment. He has discovered that whereas most substances only remove the third atom of oxygen from ozone, oil of turpentine is capable of absorbing the whole molecule. If the 92 cubic inches of ozonized oxygen in our imaginary experiment were treated with oil of turpentine in

stead of mercury, a white cloud would be produced, and the residual oxygen would be found to occupy a volume of only 76 cubic inches. The only possible explanation here is that the 92 cubic inches consisted of 16 of ozone, O3, and 76 of unaltered oxygen, O', and that the former was seized upon entire and removed in the solid form by the oil of turpentine. There can hardly be a doubt that this confirmatory experiment will settle the question, and that the nature of ozone, and the cause of its peculiar powers, will henceforth be regarded as established.

With this knowledge in their possession, and a very considerable mass of valuable information about the properties of the gas, chemists can now approach with some confidence the very difficult question of the presence and functions of ozone in the atmosphere. It will hardly be believed by some, but it is nevertheless true, that it is only within the last few months that ozone has been distinctly proved to exist in the air at all. Schönbein, in 1840, found that his test-paper became blue when exposed to the air, and hence concluded that ozone was present in it. Here was a new and easy field for scientific discovery! Test-papers were exposed in all directions, and countless observations on the amount of ozone in the air were recorded. But unfortunately a few radical errors and doubts have enshrouded these well-meant efforts, and the greater part of them are therefore worthless. Ozone is not by any means the only gas which affects the papers. Nitrous and nitric acid and chlorine, the two former of which, at any rate, are known to exist in the atmosphere, have an equal effect upon them, and sunlight alone is capable of effecting the decomposition, even when the paper is enclosed in a sealed tube. Hence, in the great majority of researches, it was quite impossible to say whether any-and, if any, how much of the coloration was really due to ozone. Even in those experiments where most pains were taken to avoid doubt, as in the recent researches of Dr. Daubeny, some ambiguity still remained, so that the more cautious chemists hesitated to pronounce absolutely that ozone was a constituent of the atmosphere. Within the last few months, however, Dr. Andrews, to whom we were already so largely indebted, has communicated to the Royal Society some careful experiments, which seem to prove that the observed effects can only be due to ozone. The most decisive proof consists in passing the air through a moderately-heated tube, which is found to destroy all traces of its power on the test-paper.

This point also, then, seems to be settled, and it only remains to ascertain with equal exactness how the ozone is formed the