The remaining case constitutes the greatest Hubstances obstacle of all in the way of the generalization of the law of definite proportions: and if it cannot be surmounted, the law sinks to the rank of an empirical rule, fit for nothing more than facilitating a certain order of chemical analyses. I refer to the class, anomalous in this view, of substances called organic. And this is the effect, in fact, of the declaration of the chemists of our time, that organic substances do not come under THK PRIXCIPLK OF DUALISM. 359 the principle of definite proportions. It amounts to saying that the law rules all the elements, except oxygen, hydrogen, carbon, and azote. The division between inorganic and organic chemistry is merely scholastic; for ail chemistry is, by its nature, homogeneous, — -that is, inorganic. And thus, if we admit of the enormous exception of the numerical composition of so-called organic substances, the doctrine of definite proportions is overthrown as a rational theory. As it evidently cannot be founded on any a priori considerations, it is only by a strict generality that it can become a rational theory.
If we could not hold at once the grand Application of principle of the dualism which pervades the principle chemistry, and constitutes its homogeneous °^ (uiahsm. character, and the doctrine of definite proportions, I should not hesitate to sacrifice the latter: for it is more important for chemical jirogress to gi-asp the great principle of systematic dualism, than to advance our investigations by the use of the numerical rule. But there is not, in fact, any incompatibility between these two means of progress: and such a brief sketch of my conception on this subject as my limits allow may show how the doctrine of definite pro})ortions can be duly generalized only by discarding organic chemistry as a separate body of doctrine, and extending the principle of dualism to all organic compounds.
If we are to include all organic compounds under one uniform system of chemistry, properly so called, we must refer to physiology, vegetable and animal, the study of the numerous secondary substances which owe their transient and variable existence to the development of vital phenomena, and which have no scientific interest except under the head of biology. We shall see, under that head, hereafter, what the precise classification is; and all that we have to do with it now as to show that it proceeds from the fundamental distinction between the state of death and that of life. The second, and most extended class of oi'ganic substances is chiefly composed of mixtures which, as such, admit of all imaginable proportions, within the limits of vital conditions. As for the substances which, exhibit real combinations, we must conceive of them as subject to the law of definite proportions; but the complexity, and yet more the instability of sucli compounds will probably for ever forbid their being successfully studied under the numerical jjoint of view, which is indeed of very inferior interest in biology. — Even after this clearing of the field, we could not accomplish the desired generalization if we had not taken a new stand with regard to the ternary and quaternary substances contemplated by ordinaiy chemistry. The rigorous dualism which I have before, and in a higher view, shown to be necessary, seems to supjily, naturally and finally, the needs of the doctrine of definite proportions.
As long as chemists persist in regarding organic combinations as ternary and quaternary, — that is, in confounding their elementary with an immediate analysis; — while oxygen, hydrogen, carbon and azote are regarded as immediately united, the compounds from them which must be recognized as distinct, after the severest sifting, *will be enough to constitute an invincible objection to the princij^le of a numerical chemistry. But if they become binary compounds of the second, or at most the third order, whose principles are formed by the direct and binary combination of those three or four elements, we find ourselves able to represent all the actual numerical varieties established by an elementary analysis, conceiving, for each degree of combination, a very small number of distinct and entirely definite proportions. — In the ternary case, — appropriate to compounds of a vegetable origin, — their three elements may be united in three kinds of binary combinations. Combining these again, still employing at once oxygen, hydrogen, and carbon, we have three principal classes of comj^ounds of the second order. But then again, each term of the new compounds really corresponds to two distinct substances: and thus, while admitting only one proportion for the binary composition of tliese bodies, we have already provided for the numerical composition of twelve substances at present called ternary. But, further, we are compelled to suppose at least three different proportions for each l)inary combination: one producing perfect neutralization, and the other two the extreme limits of the reciprocal saturation: and chemical analogies indicate a much larger number of comi^ounds. Putting those APPLICATION OF THE PRINCIPLE. 361 aside, "we have thirty-six compounds, without goiug beyond the second order, by the known combination of three elements on the principle of dualism. We are also entitled now to conceive of a third possible combination between oxygen and carbon, or between carbon and hydrogen, etc., which already furnish two, after being long supposed to admit of only one. Hence, and in view of all these considerations, we may be assured that by dualism we might completely and naturally subject to the law of definite proportions eighty-one compounds of the second order, formed from oxygen, hydrogen, and carbon; and this would unquestiouably more than suffice to represent the elementary analysis of all distinct substances in the range of vegetable chemistry.
Passing on to the quaternary case, — characterizing what is called Animal Chemistry, — it seems as if the principal class of compounds of the second order must be more numerous than in the tei-nary case: but the indispensable condition of employing all the four elements at once restricts the classes to three. But when we examine the terms of the secondary compounds, we find that while two represent only one compound each, a third represents five.
Thus, the three pairs of compounds yield fourteen of one proportion, and forty-two of the three proportions indicated in the last case. But applying, at each degree, the rational rule of a triple binary combination, without stopping at the inevitable gaps of our existing chemistry, we find ourselves in possession of ninety-nine compounds of the second order, now regarded as quaternary. This is probably a larger number than a rational analysis of animal substances will be found to require. Moreover, as animal substances have undergone a greater degree of vital elaboration than vegetable matters, it would be philosophical to admit, with respect to them, the possibility of a higher order of composition, such as physiological combinations must eminently tend to realize. On such an hypothesis, without going beyond the third order, we might obtain ten thousand perfectly distinct compounds from these four elements, all formed by an invariable dualism, and strictly subject to the law of definite proportions. It is true, nature would not jjermit the realization 362 - POSITIVE PHILOSOPHY.
of more than a small part of these speculative combinations; but I have pursued the consequences of my conception to this extreme ideal limit, to show how abundant are the rational resources supjjlied by this new theory for the generalization of the laws of numerical chemistry. If this view is not followed up, or some equivalent one proposed, it is evident that we must give up the doctrine of definite proportions as a law of natural philosophy, and return to Berthollet's theory, merely enlarging the cases of fixed proportions which he admitted. In the present state of the question there is no other choice. But my theory having been, not instituted for this destination, but naturally arrived at by another way, and with higher views, and proceeding from established principles, to meet the needs of chemical philosophy, seems to me to be presumptively entitled to a future, and pei'haps speedy realization. This account of the present aspects of the doctrine of definite proportions will enable any one to judge of its real progress from its institution to this day; of the conditions which must be fulfilled before its principle can be converted into a great law of nature; and of the rational course which alone can lead to such a final constitution of numerical chemistry.
THE ELECTRO-CHEMICAL THEORY.
FROM the beginning of modern cliemistry, Relation of the chemical influence of electricity Electricity to manifested itself uneqiiivocally in many im- Chemistry, portant phenomena, and above all, in the grand experiment of the recomposition of water by the direct combination of oxygen with hydrogen, effected by the aid of the electric spark. But the special attention of chemists was not strongly drawn towards this agency till Volta's immortal discovery disclosed its princij^al energy, in rendering the electric action at once more complete, more profound, and more continuous. Since that time, various series of general phenomena have taught us that electi-icity is a chemical agent more universal and irresistible than heat itself, both for decomposition and combination. The danger now is of exaggerating the relation it bears to the general system of chemical science. Though chemistry is united to j^hysics by this agency more than by any other, it must yet be remembered that the two sciences are distinct, and that there should be no confounding of chemical with electrical properties. In order to ascertain with precision what are the relations of chemistry with electrology, we must briefly review the gradation of ideas which have led up to the present electro-chemical theory, as systematized by Berzelius.
The first important chemical effect of the History of voltaic influence was the decomposition of tlie case. water, established by Nicholson in 1801. It Nicholson's was a necessary result of examination into discovery, the action of the pile, without any chemical intention. It confirmed a truth before well known: but it had a high chemical value, as revealing the chemical energy of the instrument; and it thus constituted the starting-point of eleetro-clieinical research. We may even refer to this origin the first attempts at founding a general theory of electro-chemical phenomena; for the conception offered hy Grothuss, to explain Nicholson's observation by the electric polarity of molecules, contains the germ of all the essential ideas which have expanded, according to the requisitions of the phenomena, into the present electro-chemical theory.
The analytical power of the voltaic pile having been once discovered, it was natural for the chemists to apply the new agent to the decomposition of substances which had hitherto resisted all known means. The first series of, attempts produced, after a few years, the coverv ^ ^'^ brilliant discovery by the illustrious Davy, — of the analysis of the alkalies, properly so called, and the earths. Lavoisier's theory, which showed that every salifiable base must be a result of the combination of oxygen with some metal, had foreshown this analysis; but no chemical means had sufficed to effect it. It was believed in, in spite of Berthollet's discovery of the true composition of ammonia; and the brilliant result obtained by Davy was an easy consequence of a discovery completely jirepared for. M. G-ay-Lussac soon followed, with a more difficult but less striking achievement, — the confirmation, by a purely chemical process, of the electrical analysis of potash.
Nicholson's observation having originated electrochemistry, and Davy's given it a great impulse, the next step was to investigate the chemical influence of electricity, in a scientific view. This was effectually, though indirectly, determined by Davy's great feat; for by it chemistry was proved to have achieved the most important, and hitherto inaccessible analyses: and in fact, the science has not since made any essential acquisition. Electro-chemical action was presently and permanently subjected to direct and regular study; and it was irrevocably constituted a fundamental part of chemical science, when Berzelius accomplished his series of extension investigations on the voltaic decomposition of all the salts, and then of the principal oxides and acids. It was in consequence of these researches that the habitual consideration of electric i:)roperties has SYXTIIESIS. 365 assumed a growing importance in the chemical study of all substances, which are now scientifically divided into the classes of electro-negatives and electro-positives. It was thus the privilege of Berzelius to be the first to conceive of the electro-chemical theory under an entirely systematic form.
One condition remained to be fulfilled, to s - tl f • 1 give its due scientific character to this new process, branch of chemistry. The voltaic action had thus far been regarded only analytically: — it must be regarded synthetically also. This was done by the labours of Becquerel, who fully established the synthetical influence of electricity, fitly applied; and who, moreover, employed it in effecting new and valuable combinations, hitherto impracticable. From this procedure arose the necessity of modifying the method of experimentation. The apparatus which was powerful enough to decompose was much too powerful to combine, because it would probably decompose the immediate principles which were intended to be combined; and thence arose the method of employing the protracted action of very feeble electric powers, — every advantage being given to the disposition of the substances to be acted upon. M. Becquerel did this very successfully by operating with a single voltaic element, and seizing each substance in the T.^„„„Q,.ar ■ ' '^ -.,., l>ecquerel s.
state called nascent, which is agreed upon as most favourable to combination. This change of procedure is the distinctive honour of M. Becquerel. He not only determined direct combinations, not before obtainable, but exhibited in others, which were obtainable, the remarkable property of clearly manifesting their geometrical structure, through the slowness and regularity of their gradual formation; — a character especially marked in the case of certain metallic sulphurets, some oxides, and several salts. It does not lie within our province to point out the results of this method in regard to the natural history of the globe, in explaining a great number of mineral origins, when the time for such concrete questions shall have arrived. It is more in our way to observe the importance of these labours in bringing up chemical synthesis to something like harmony with the progress of analysis, — favoured as the latter pursuit had been by the ease with which we destroy, in comparison with the difficulty with which we recreate. Once more, these researches of M. Becquerel have comj^leted the general constitution of electro-chemistry, which, being henceforth at once synthetical and analytical, can only expand in one of these two directions, however great the improvements remaining to be attained.
Such has been the filiation of electro -chemical discoveries, since the beginning of our century. A little attention to the great phenomenon which was the original subject of the electro-chemical theory will show how this study has gradually led to a new fundamental concejition for the whole of Chemistry.
It has been said, through all periods of combustion chemical research, that the study of Combustion must be the central point of the science. So it was thought in the ancient theological period of the science; and also in the more recent metaphysical stage, when combustibility was called phlogiston, and regarded as an intangible materialized entity: and the advent of the positive period of chemistry was marked by the establishment of Lavoisier's new theory of combustion. In our day it is the recognized necessity of modifying this theoi'y that has especially led to the electric conception of chemical phenomena..., The pneumatic theory of combustion of theorv Lavoisier had two entirely different objects in view; objects which are too commonly confounded, but which we must be careful to keep apart. First, the analysis of the general phenomenon of combustion: and, secondly, the explanation of the effects of heat and light, which is, in the eyes of the vulgar, the more important of the two. Both were treated in the most admirable manner possible in the existing state of knowledge; and in the characteristics of postivity and rationality Lavoisier's theory has not since been surpassed. All combustion, abrupt or gradual, was regarded as consisting in the coml)ination of the combustible body with oxygen, whence, when the body was simjole, must result an oxide, generally susceptible of becoming the basis of a salt, and, if the oxygen THEORY OF COMBUSTION. 367 was preponderant, a true acid, the principle of a certain kind of salts. As for the disengagement of heat and light, it was attributed, in a general way, to the condensation of the oxygen, and in an accessory way, to that of the combustible body, in this combination.
The first part of this anti-phlogistic theory -p. i. i-.• • has a much more philosophical character than the second. It was eminently rational to analyse the phenomenon of combustion, so as to seize whatever was common to all cases. The conclusions drawn might be too general; and were afterwards proved to be so; but Avhatever would stand the test of time miist form a body of indestructible truth, constituting an essential part of chemical science through all future revolutions. The case is different with the explanation of heat and light. The question is not, like the first, of a chemical, but a physical nature; and, whatever may be its final solution, it cannot affect chemical conceptions. It would have been wiser to abstain from offering any general explanation of the effects of heat and light through such a supposition as that of a condensation, which does not necessarily take place, and which is, in fact, found to be often absent. Lavoisier hoped to attach the therniological effect to the great law discovered by Black, of the disengagement of heat proper to the passage of any body fi'om one state to another more dense; but such a connection could not be established on the ground of phenomena not invariably present, or indisputably manifested. However, it would be too much to expect a perfect scientific reserve in discoverers who bring out scientific truths from a region of metaphysical fantasies. It is from their followers that we have a right to demand it; and we are compelled to charge upon the chemists who have been eager to substitute the electro-chemical theory for the antiphlogistic theory, properly so called, a want of care in constructing explanations analogous to those which are dismissed as insufficient. To justify this charge, we must review the proofs of the imperfection of Lavoisier's theory, — still regarding it under the two divisions just exhibited.
Berthollet saw presently that Lavoisier's „.i ii., method of analysis of combustion must be limitation " modified. One of the chief consequences of this analysis was that every acid and every salifiable base must be a result of combustion; that is, of the combination of any element with oxygen; whereas, Berthollet discovered that one of the most marked of the alkalies, ammonia, is formed of hydrogen and azote alone, without any oxygen; and soon after, he proved that sulphuretted hydrogen gas, in which also there is no oxygen, nevertheless presents all the essential properties of a real acid. These facts have since been confirmed in every possible way, and especially by the electric method; and the exceptions, both as to alkalies and acids, have become so multij^lied, that the investigation and comparison of them have given that high character of generality to the study of alkalies and acids which belongs to it in our time.
Moreover, the primitive theory of combustion has been gradually modified by the discovery that a rapid disengagement of heat and light is not always an indication of a combination with oxygen. Chlorine, sulphur, and several other bodies, even non-elementary, have been found to occasion true combustion. And again, the phenomenon of fire is no longer attributed exclusively to any special combination, but, in general, to all chemical action at once very intense and vivid.
It does not follow that because Lavoisier's discoveries have parted with some of their character of generality, they have lost any of their direct value: and such alteration of views as there is relates chiefly to artificial phenomena, while the natural facts remain securely established. Thus, though there are acids and alkalies without oxygen, it is unquestionable that the greater number of them, and especially the most jjowerful, are oxygenated: and again, if oxygen be not indispensable to combustion, it remains the chief agent, and especially in natural combustions. In natural history, the theory is applicable, almost without reserve, though it is insufficient for the severe conditions of abstract science. If the universal sovereignty of oxygen has been overthrown, it will yet be for ever the chief element of the whole chemical system.
the explanation or fare, — it was destroyed by the first direct examination of it. No new facts were re- THEORY OF COMBUSTION. 369 quired for its overthrow, but merely a more scientific appreciation of common phenomena. It will not even serve naturalists for their concrete purposes in any degree, having never really explained the most ordinary effects. The required condensation is found to be only occasionally present, and often absent in the most important cases; so that if it were not for the connection of this aspect of the theory with a sounder one, it would be inconceivable how it could have held its ground up to a recent period, — busy as the chemists were with other theoretical speculations. After finding that in cases where condensation was supposed, expansion exists instead; and that where we find vivid combustion, there should, by the theory, be a great cooling, we are brought to the reflection that if the fire on our hearths was not a matter of daily fact to us, its existence must become doubtful, or be disbelieved, through those very explanations by which the phenomenon has been proposed to be established. To my mind, this is a clear indication that the chemical production of fire does not admit, in a general way, of any rational explanation.
Otherwise it appears incomprehensible that men of such genius and such science, at a time so near our own, should have been so deluded. The electric fire, now proposed for an explanation, must have been sufficiently known to Lavoisier, Cavendish, Berthollet, and others, to have served as a basis to their theory, if its preponderance over the merits of their hypothesis had been so great as is now commonly supposed. This consideration however, striking as it may be, is no dispensation from the duty of examining the electro-chemical conception, for which we have been prepared by this short account of its antecedents.
According to this theory, the fire produced in the greater number of strong chemical reactions must be attributed to a real electric discharge which takes place at the moment of combination (by the mutual neutralization, more or less complete, of the oi:»posite electric conditions) of the two substances under consideration, — one of which must be electro-positive and the other electro-negative. There is every reason to fear, however, that when this theory has been effectually examined, it will be found as defective in rationality as its jjredecessor. If electric effects are I. B B 370 POSITIVK PHILOSOPHY.
concerned in all chemical phenomena, as seems to be now agreed upon by most chemists and jihysicists, they must oftener be supposed than found: and the electric symptoms are most impossible to detect in precisely those chemical phenomena which have been most relied on for overthrowing the old theory. And in the cases in which electrization is evident, its chemical influence is so equivocal that some regard it as the cause, and others as the effect, of the combination. The explanation is not yet positively established for any phenomenon whatever that has been duly analysed: while its vague nature leaves room for fear that it will not be so radically or so speedily destroyed as its predecessor. It could be plainly shown whether the requisite condensation did or did not exist: but there is always a resource, in the more recent case, in the faint or fugitive character of the electric condition, which defies our means of positive exploration, — qualities which are far from being a ground of recommendation of a theory which is to account for very striking and intense effects. I do not desire to say that the disengagement of light and heat can never have an electric origin, any more than that it can never proceed from the condensation proposed: but 1 think an impartial observation woi;ld decide that in most cases of combustion there is neither condensation nor electrization. My own view is that these vain attempts to explain the chemical production of fire jDroceed from the lingering metajihysical tendency to penetrate into the nature and mode of being of phenomena: and that chemical action is one of the various primitive sources of heat and light, which cannot, from their nature, usually admit of any positive explanation; that is, of being referred, in this relation, to any other fundamental influence.
If our chemical science were more advanced than it is, we should not have to point out that the consideration of fire, which, however important, is only a physical accessory of chemical phenomena, cannot afford a rational ground for a radical change in our conceptions of chemical action. When our predecessors regai'ded heat as the tlie theorv^ chief physical agent in composition and decomposition, they did not pervert such a consideration to the point of assimilating chemical to DIFFICULTIES OF THE THEORY. 371