ge i os iit “ oaths " PPA A hee a Ente Hela 5 rae : SRAEE Delia ate 45 af). S70 Peiren tate e! Senet BGR EE Spates te pace sain ies ets TRAE REE IE Petia centr s a Ft a : narzate ty te Okt % 3 rE tT: PEPER AS ry 4 . O53 wiptrarere eins Belch Aye ote ete , a8 xb Ai fo Mal vets int Wlee dp Aedbgegel =. ‘ ‘ : . vee Uy? tee et 8e etl fae oer obs Foy righ Gdte bie wish (Gk a MeCaee Eek SF dvtetce % CHAPTER III. ‘ PROUT’S HYPOTHESIS—LAW OF SPECIFIC HEATS— ISOMORPHISM, Dulong and Petit: Law of Specific Heats—Mitscherlich: _ _Isomorphism—Berzelius’s System of Atomic Weights . ae BA Ss PAGE 33 49 pl CONTENTS. CHAPTER IV. SYSTEM OF CHEMICAL EQUIVALENTS —EQUIVALENT NOTATION, Objections to the Principle of Berzelius’s Notation—Discus- sion of the Objections brought forward by Gmelin—Incon- sistencies of the Equivalent Notation ° CHAPTER V. PRESENT SYSTEM OF ATOMIC WEIGHTS: GERHARDT AND LAURENT—CANNIZZARO. Notation of Gerhardt—Ideas of Laurent—Reform of Canniz- zaro—Table of Atomic Weights—Law of Volumes—Present System of Atomic Weights deduced from the Law of Avo- gadro—Apparent Exceptions to the Law of Avogadro —Atomic Constitution of the Elements—New Atomic Weights in Harmony with the Law of Dulong and Petit — Molecular Heats—New Atomic Weights in Harmony with the Law of Isomorphism CHAPTER VI. THE NEW SYSTEM OF ATOMIC WEIGHTS RESPECTS AND RENDERS EVIDENT THE ANALOGIES WHICH EXIST . BETWEEN BODIES. It agrees with Chemical Analogies—Ideas of Dumas—Mende- lejeff’s Principle of Classification—New Atomic Weights in Harmony with Physical Properties, with Chemical Proper- ties e . © e ° ° © ° CHAPTER VII. ATOMIC AND MOLECULAR VOLUMES, Researches of Hermann Kopp—Molecular Volumes of Salts , 67 149 187 CONTENTS. Vii BOOK II. ATOMICITY. —_—- CHAPTER I. PAGE DEFINITION AND HISTORIC DEVELOPMENT OF THE IDEA OF ATOMICITY: . ° : . : ° - 196 CHAPTER II. AFFINITY AND ATOMICITY, TWO DISTINCT PROPERTIES OF ATOMS. Atomicity a Relative Property of Atoms—Molecular Com- pounds ° ° . ; ° e ° e 224 CHAPTER III, CONSTITUTION OF BODIES DEDUCED FROM THE THEORY OF ATOMICITY. Atomicity applied to the Interpretation of Isomers—Atomicity . applied to the Interpretation of Molecular Dissymmetry . 259 CHAPTER IV. HYPOTHESIS UPON THE CONSTITUTION OF MATTER. Conclusion, A - = . 16 . 3805 Vili CONTENTS. APPENDIX. NOTE I. WATER OF CRYSTALLISATION . 3 - a ++ 33a NOTE II. THE CONSTITUTION OF DOUBLE SALTS - e - $834 NOTE IIL. THE ISOMERISM OF THE AMYL ALCOHOLS . ° « vaG NOTE IV. THE ACTION OF HEAT UPON GASES, ; ‘ ; >; 888 PLATE. THE RELATIONS BETWEEN THE ATOMIC WEIGHTS OF THE ELEMENTS AND THEIR PHYSICAL PROPERTIES, AFTER LOTHAR MEYER - i - . < * = at end THE ATOMIC THEORY. —_— oO PG Orr. ATOMS. CHAPTER I. HISTORICAL INTRODUCTION—-RICHTER— DALTON. Tue hypothesis of atoms, put forward by the Greek philosophers, and revived in modern times by great thinkers, acquired a definite form at the begin- ning of this century. John Dalton was the first to apply it to the interpretation of the laws which he and Richter recognised as governing chemical combinations. Confirmed by the great discoveries of Gay-Lussac, Mitscherlich, Dulong and Petit, the hypothesis has assumed a definite form, connecting many various facts of a chemical and physical nature. Fundamentally it consists of modern ideas upon the constitution of matter. ~ In common with correct ideas, it has grown with time, and nothing has as yet happened to stop its pro- 2 THE ATOMIC THEORY. gress; but, in common with all fruitful ideas, it has been an instrument of progress even in the hands of its detractors. The latter are now few, and the hypothesis seems to make a firm stand against the regular opposi- tion of some and the subtle attacks of others. In these pages we propose to discuss both its historical evolution and its present form, and we shall thus show the influ- ence it has exercised upon the progress of science since the beginning of the century. Dalton revived the hypothesis of atoms to explain the fact that in chemical combinations elements unite in fixed proportions, and in certain cases in multiple proportions. He admitted that these proportions repre- sent the relative weights of indivisible particles of the bodies, which particles are brought into contact and grouped by the fact of combination. This led to the consideration of atomic weights, and the idea of representing the composition of bodies by symbols which indicate both the nature and the number of these particles and the proportion of the elements entering into combination. We have here two things which must not be confounded— facts and an hypothesis. We shall retain the hypothesis as long as it gives a faithful interpretation of facts, and enables us to group them, to connect them together, and to anticipate fresh ones —as long, in fact, as it proves fertile. An hypothesis thus formed rises to the rank of a theory. We shall endeavour to show, in demonstrating its origin, progress, and results, that this is the case with Dalton’s concep- tion. DEFINITE CHEMICAL PROPORTIONS. 3 Ps Simple bodies combine in definite proportions. This is one of the most firmly established truths of natural philosophy. It includes the two following facts :— Firstly, the relative weight of combining bodies is always fixed in every combination; secondly, the numbers which express these relations are interproportional for all kinds of combinations. We must clearly understand the meaning of these propositions. Two simple bodies unite so as to form a given com- pound. As long as the compound lasts the relative weights of the two elements will remain perfectly con- stant, whether the quantities acting upon each other have been great or small; the smallest particles, as well as the whole mass, will contain strictly proportional weights of these elements, which no physical circum- stances, such as pressure or temperature, can modify. This is true for all kinds of combinations, the most simple as well as the most complicated. This fixity of the proportion in which bodies combine was acknow- ledged and admitted as a truth more than a century ago by some eminent chemists, and by all in the year 1806. Bergman was conscious of the truth, even if not logically convinced of it; in fact, the numerous quan- titative analyses for which we are indebted to him would have been aimless or useless if he had been under the impression that the compounds he was analysing were formed in chance proportions. Lavoisier demon- strated in the clearest manner the fact of the constancy of the relations in which bodies combine. In every 4 THE ATOMIC THEORY. oxide, in every acid, he said, the relation of oxygen to the metal is constant; and this relation should be exactly determined for every oxygen compound. He admits, moreover, that the difference between the acids of sulphur and the oxygen compounds of nitrogen is due to the power possessed by these simple bodies of uniting with oxygen in several proportions, each degree of oxidation corresponding to a fixed and constant rela- tion between the weights of the two elements. The law of fixity was thus distinctly admitted and clearly stated by Lavoisier ; one step more, and he would have discovered the law of multiple proportions. He did not, however, make this decisive step. Even as regards the fixity of several proportions, though he was himself convinced of the fact, he was not successful in making it universally accepted. In the month of July 1799 his pupil Berthollet read at the Egyptian Institute, which was sitting at Cairo, a memoir entitled ‘Researches upon the Laws of Affinity.’ He there for the first time brought forward profound ideas upon the influence exercised by the physical condition, the cohesion, solu- bility, insolubility, and volatility of bodies upon the affinity and progress of chemical decompositions. With- out denying the fixity of the composition of certain compounds, he attributed this fact to the chance influ- ence of these physical conditions, which in some cases were constant, and would not allow that it partook of the character of a general law. It is true, he said, that in sulphate of baryta the relation between the sulphuric acid and the baryta is constant, simply because the acid and base must PROUST—FIXITY OF PROPORTIONS. 5 unite in this precise and fixed proportion to form a salt of absolute insolubility. Thus, here as in many other eases, constancy of composition is dependent upon a physical property, cohesion—in other words, the insolu- bility of the sulphate of baryta. But the rule is as follows: Chemical combinations take place in propor- tions which may vary within certain limits. A salt formed by a soluble acid and a slightly solu- ble or insoluble base may be precipitated in an insoluble form, unvarying in composition, when the proportion of the base is exactly such as to cause the precipitation of the salt of this composition ; but if the proportion of the base is increased the salt will still be precipitated, but its composition will be different, for it now consists of greater quantities of base for the same quantity of acid. A metal, such as mercury, dissolved in nitric acid, will unite, in the process of oxidation, with quantities of oxygen varying between a maximum and a minimum. We canrfot, therefore, maintain with Lavoisier that when a salt is formed by the action of an acid upon a metal, there is a constant relation between the quantity of the metal and the quantity of oxygen which the former takes from the acid in the process of oxidation. These propositions of Berthollet were first opposed and successfully refuted by S. L. Proust. Having remarked, in 1799, that upon dissolving native carbonate of copper in an acid, and then precipitating the solution ' by an alkaline carbonate, he obtained a quantity of car- bonate of copper equal to that of the native carbonate which had been dissolved, Proust drew from this fact 6 THE ATOMIC THEORY. the conclusion that the composition of carbonate of copper is fixed and invariable, whether the salt has been formed in the depths of the earth or artificially by a chemical process. His subsequent researches enabled him to generalise this conclusion; and in speaking of these researches we must specially quote those upon the composition of the two oxides of tin, the sulphides of iron, and sulphide of antimony. In all these com- pounds the relation in weight between the two elements is constant ; and if two simple bodies, by combining in different proportions, are able to form several com- pounds, as is the case with tin and oxygen, iron and sulphur, it is evident that in every degree of combina- tion the relation in question is invariable. Proust brought forward these facts, which he had discovered in opposition to those upon which Berthollet took his stand, and showed that the latter allowed a different interpretation. Metallic solutions, where the metal enters into combination with variable quantities of oxygen ;-salts, which, when precipitated, may contain variable quantities of bases ; or oxides of tin and lead, which have been obtained by the calcination of metals in contact with air, and which have fixed variable quantities of oxygen—in no case consist of, or constitute, definite chemical compounds, but are mixtures, in dif- ferent proportions, of several compounds, all of which possess a fixed composition. The fixity of composition, indeed, seemed to Proust an essential attribute of com- binations, a great law of nature—the pondus nature, justly recognised by Stahl. This discussion, which is one of the most memorable FIXITY OF PROPORTIONS. 7 of which science possesses a record, lasted from 1799 till 1806, and was maintained on both sides with a power of reasoning and a respect for truth and propriety which have never been surpassed. The fullest deve- lopment of Berthollet’s views appeared in his celebrated work entitled ‘ Essai dune Statique chimique,’ which was published in 1803. The great idea developed in this book is that chemical affinity and astronomical attraction are different manifestations of an identical property of matter, which led the author to regard not only the energy of affinities as producing chemical reactions, but also the influence of the masses. In a great number of reactions this influence does undoubtedly govern the progress of decomposition or combination ; it augments or diminishes the proportion of compounds which are formed or destroyed in a reac- tion, but it does not govern the proportions in which the elements unite in these compounds. On this latter point Berthollet held a different opinion; he main- tained that mass does exercise an influence upon the combining proportions of two bodies when no physical condition is present to determine the separation of a compound in fixed proportions. Thus, when an acid acts upon a base in such a manner as to produce a solu- ble salt, the point of neutrality undoubtedly corresponds to fixed proportions-of combined acid and base; but if an excess of one or other of these elements be added, it also will enter into combination, and, moreover, in variable proportions, till a physical property—cohesion, for example—determines the separation of a compound of fixed proportions. In a great number of chemical 8 THE ATOMIC THEORY. combinations, therefore, this fixity in the proportions of elements may be observed; but, in the opinion of Berthollet, they are exceptional cases, to which it would be wrong to ascribe the dignity of a general law. Proust, on the contrary, maintained the generality of this law. If it is impossible, he says, to make an ounce of nitric acid, an oxide, a sulphide, or a drop of water in other proportions than in those which nature, from all eternity, has assigned to these compounds, we must acknowledge that for chemical combinations there is a sort of ‘ balance,’ which is subject to the immutable laws of nature, and which, even in our laboratories, deter- mines the relation of the elements in these compounds. The latter are of several orders. The most simple are generally formed of two elements—at most of three, very rarely of four. But these compounds of a simple order may combine with each other, so as to form more com- plex compounds; in other cases they are merely mixed together. In these mixtures the proportion of the elements is naturally subject to variation; in all che- mical combinations properly so called it is, on the contrary, fixed. The opinion of Proust was well founded; it won the day, in spite of the opposition of his powerful antagonist ; and we cannot too much admire the per- severing energy and discernment displayed by the chemist of Angers in this contest, when he took one by one the arguments of Berthollet, and opposed to the facts collected and arranged by the latter in support of his theory fresh facts and fresh analyses of his own, which, it must be confessed, were not always models of FIXITY OF PROPORTIONS. 9 accuracy. The superior intelligence, however, of an accurate and lofty mind saved him from error in the discussion of results, and made up for the insufficiency of the methods of that time. This great truth of the fixity of chemical propor- tions was, then, definitely established in the year 1806. But the discussions between Berthollet and _ Proust, which agitated the scientific world during the first years of this century, only gave an incomplete idea of it, for they dealt solely with the composition of each compound taken individually. The question as to whether sulphide of antimony was a constant com- pound, and whether this was also the case with the sulphides of iron, the oxides of tin and cobalt, was answered in the affirmative by Proust, in the negative by Berthollet. It is now definitely decided in the affirmative. We must not, however, forget that Proust and Berthollet only attacked the question from one side, for there is another. It is true that this sulphide of antimony, these sulphides of iron, and, in fact, that all sulphides present a fixed composition; and, again, it is equally true that in every metallic oxide the metal and the oxygen unite in invariable proportions. But this is not all. Analysis shows, further, that the relations between quantities of different metals uniting with a fixed weight of sulphur are the same as those between different metals uniting with a fixed weight of oxygen. Independently, therefore, of the fact of fixity, there is the further fact of the proportionality of the combining quantities or weights of bodies; and the case in question is not an exceptional one, but 10 THE ATOMIC THEORY. belongs to a whole order of. similar facts—is, in short, a law. We have, in demonstrating this law of propor- tionality, employed as examples the very compounds which enabled Proust to establish the law of fixity. It may, however, be demonstrated under a more general and striking form. A is a certain weight of a simple body. B is a certain weight of another simple body, which is exactly sufficient to form with a the combination a B. . A . The relation — is constant. B C is a certain weight of a third simple body, exactly sufficient to form with a the combination ac. The re- ay Oe lation — is constant. C D is a certain weight of a fourth simple body, exactly sufficient to form with a the combination ap. The . A J relation — is constant. D This is Proust’s law. Let us now take the second body B, and form com- binations between this body and the third c and the fourth p. Experience shows us that the quantities c and Dp which combine with a will also combine with 3B —in other words, that the weights of the bodies 3, c, p, which formed definite compounds with A, are unchanged when they combine with each other. From the fact of the existence of compounds AB, AC, AD, We may assume the existence of compounds BC, BD, CD, in which the quantities A,B, C, D, are constant. In short, there LAW OF PROPORTIONALITY—RICHTER. ¥y exists between all compound bodies formed by the union of two elements such a definite relation of composition that we have only to determine the proportions in which the most widely differing elements unite with one of their number, and we shall also have determined the proportions in which they combine with each other. This is the law of proportionality, discovered by Richter, who lived at Berlin towards the close of the last century. For many years another German chemist—C. F. Wenzel—was considered the author of this great dis- covery. It was attributed to him by Berzelius.' M. Dumas also claims it for him,? and all chemical treatises 1 The following are the terms in which Berzelius claimed for Wenzel the discovery of the proportionality of quantities of acids and bases which exactly saturate each other:—‘ He published the result of these experiments in a memoir entitled Lehre von den Verwandtschaften, or the Theory of Affinities, at Dresden in 1777, and proved, by singularly accurate analyses, that this phenomenon (the preservation of neutrality after the mutual decomposition of two neutral salts) was due to the fact that the quantities of alkalies and earths which saturate a given quantity of the sameacid are the same for all acids; so that if we decompose, for example, calcium nitrate by potassium sulphate, the potassium nitrate and the calcium sulphate obtained will preserve their neutrality, because the quantity of potash which saturates a given quantity of nitric acid is to the quantity of lime which saturates the same quantity of nitric acid as the potash is to the lime which neutralises a given quantity of sulphuric acid.’— Yraité de Chimie, French edition, 1831, t. iv. p. 524. 2 Chemical Philosophy, p. 200. The error concerning the part attributed to Wenzel in the discovery of the law of proportionality has been corrected by several scientific writers—first by Hess (Journal fiir praktische Chemie, t. xxiv. p. 420); then by Schweigger, in the work‘ entitled Ueber stéchiometrische Rethen im Sinne 12 THE ATOMIC THEORY. fifty years ago quoted him as the precursor of Richter. He was rather the rival of Bergman and Kirwan. The analyses of neutral salts which he published were accu- rate ; but he nowhere mentions the fact of the reserva- | tion of neutrality after the double decomposition of | the two neutral salts; he admits, on the contrary, that in the phenomenon in question the quantity of the two | neutral salts which react upon each other being cal- culated after their known composition, a certain excess _ of one of the elements may remain after the decom- position has taken place. This opinion is contrary to facts, and must necessarily have rendered it impossible for the author to discover the law of proportionality. — This law was demonstrated a few years later by a much — less experienced chemist than Wenzel, who was obscure and diffusive in his productions, but endowed with singular penetration and rare perseverance. ee a iT: J. D. Richter was preoccupied with the idea of applying mathematics to chemistry, and particularly to that of discovering numerical relations between combin- — ing bodies. His efforts in this direction did not meet with success; for, though he was the first to recognise and demonstrate the law of proportionality between _, the quantities of bases uniting with a given weight of Rai pts ' base, and between the quantities of acids uniting with a given weight of base—a most important and well-esta- blished fact—he fell into error in trying to show that these Richter’s, Halle, 1853 ; lastly by R. A. Smith (Memoir of J. Dalton, and History of the Atomic Theory up to His Time, London, 1856). LAW OF PROPORTIONALITY—RICHTER. 13 quantities form numerical series, the terms of which bear to each othera simple ratio.! But'we need not pay much attention to this point. Let us rather gather from the work of Richter the great truths and fundamental dis- coveries which demand the grateful recognition of pos- terity, all the more strongly from the fact that they were neglected and almost ignored by his contemporaries. Richter founded his researches upon the then well- known fact of the permanence of neutrality in the double decomposition of two neutral salts. Richter found and clearly demonstrated the required explana- tion of this fact. In the first volume, published in 1 Richter tried to show that the quantities of bases which saturate a given weight of acid represent the terms of an arithmetical pro- gression, and that the quantities of acids which combine with a given weight of base form the terms of a geometrical progression. Thus, for example, he found that 1,000 parts of hydrochloric acid are saturated by-734 of alumina, 858 of magnesia, 1,107 of lime, and by 3,099 of baryta. These numbers form the terms of a series a, a+b, a+3b, a+19b, in which a@=734 and 0=1245. Having afterwards discovered the saturating capacity of strontia for hydro- chloric acid, he found that this base would occupy the place @+116 in the preceding series, a result which he soon corrected to a+ 9b. A different but very simple relation exists, in his opinion, between the quantities of acids which saturate a given quantity of base. Thus the quantities of fluoric (hydrofluoric) acid 696-4, muriatic acid 1160-0, sulphuric acid 1630:0, and nitric acid 2290-4, which saturate 1,000 of magnesia, form the first, third, fourth, and fifth terms of geometrical progression—e, cd, cd*, ed*®, cd’—the first term of which c is = 696°4, and d=1°1854, Again, the quantities of carbonic, sebacic, oxalic, formic, succinic, acetic, citric, and tartaric acid necessary to neutralise a given base increase according to a geometrical progression a, ab, ab*, ab’. Metallic acids, on the con- trary, are subject to another law: the quantities of tungstic, chromic, arsenic, and molybdic acid which saturate a given weight of base constitute the terms of an arithmetic progression. 14 THE ATOMIC THEORY. 1792, of his ‘ Elements of Stoichiometry’! he expresses himself as follows :—Let a and B represent the weights or masses of two neutral compounds (salts) which exactly decompose each other; the new bodies will remain neutral: let a represent the mass of an element in A, and b that of an element in B; the masses of the two elements in a will be a, A—a, and in B will be b, B—b. Before decomposition the ratio of the masses (weights) of the neutral compounds A and B will a b 22s and sae A—a B—b After decomposition the masses of the elements in the All these propositions are founded upon inaccurate data, a fact which doubtless did not escape the notice of some of Richter’s con- temporaries, and contributed to throw discredit upon his labours. He himself sometimes saw the necessity of correcting some of these errors; but though he gave up a few details, he still held to the numerical laws demonstrated above—the new figures always adapted themselves to it. Thus in 1797 soda changes its place in the series of bases neutralising a given weight of sulphuric acid. Richter now finds that 1,000 parts of this acid are saturated by 672°1 of volatile alkali (instead of 638), by 858°6 of soda (instead of 1,218), and by 1604:6 of potash (instead of 1,606). These numbers increase as the terms a, a+b, a+ 5b, while the original numbers formed the terms of a series a, 4+ 3b, a+ 5d. ; These are great imperfections in the work of Richter; but, though we cannot but regret that his memory should be charged with them, they must not cause us to forget the great truths which he had the honour of discovering. — 1 Anfangsgriinde der Stichiometrie, oder Messkunst chemischer Elemente. Pa 2 We have reversed these fractions, which the author wrote— ~ A—@ B—b ee d . an b iin. 4 « a. 4 2 —— LAW OF PROPORTIONALITY—RICHTER. 15 new products will be a, B—b and b, a—a, and the ratio of these masses will be a tee Atl B—b A—G@ If, then, the ratio of the masses (elements) is recog- nised in the original compounds, the same ratio must be acknowledged in the new compounds. Richter drew up, in 1793, a table which he termed series of masses—the quantities of analogous elements (acids or bases) which combine with a given weight of another element. In another part of the work which we have just quoted he definitely states the following proposition :—The different quantities of bases which form neutral salts with 1,000 parts of anhydrous muriatic acid also form neutral salts with a given weight (1,394 parts) of anhydrous sulphuric acid. It follows, from the formula given above, that if we take a weight a of a muriate (chloride) containing 1,000 parts of acid and a weight a—1000 of base, and a weight B of a sulphate containing 1,394 of sulphuric acid, and B—1394 of a second base, this quantity of the latter base will exactly neutralise 1,000 parts of muriatic acid, while the quantity a—1000 of the first base will exactly neutralise 1,394 parts of sulphuric acid. If, therefore, we mix the two original salts, the neutral muriate and sulphate, we shall obtain from the double decomposition a new sulphate and a new muriate, which again will be neutral. Richter thus explains the fact of the permanence of neutrality when two neutral salts exchange their bases and acids. He at this time 16 THE ATOMIC THEORY. (1793), and in the same work, gave the first ‘series of masses’ for the alkaline bases and for the earths—that is to say, the equivalent quantities of bases which saturate a given weight (1,000 parts) of sulphuric, hydrochloric, and nitric acids. The following is the series :— Sulphuric Acid | Muriatice Acid Nitric Acid Potash. . ] : 1,606 2,239 1,143 Soda . ‘ 4 : 1,218 1,699 867 Volatile alkali . ‘ 638 889 453 Baryta . , ; : 2,224 3,099 1,581 ene ee BPs ine 796 1,107 565 Magnesia . ; ; 616 858 438 Alumina : i : 526 734 374 Although these figures are far from correct, they allow the deduction of the law of proportionality, with which the name of Richter is justly connected. He afterwards completed and corrected them. Having ascertained the quantities of lime and potash which neutralise 1,000 parts of fluoric (hydrofluoric) acid, he proved that these quantities are very nearly propor- tional to those which neutralise 1,000 parts of muriatic acid. On this point he affirms that ‘the masses of alkalies or alkaline earths, when they maintain neutrality -with a given mass of either of the three other volatile acids,' will always bear to each other the same ratio.’ The idea is correct, though the form of expression is not happy. Richter, indeed, generally failed in the latter respect. Thus he endeavours to generalise the 1 Sulphuric, muriatic, and nitric acids. RICHTER—LAW OF PROPORTIONALITY. 17 law he has discovered by terming the substance (the acid, for example) which enters into combination with a series of analogous substances (bases) the determining element, and the latter the elements determined. ° Let P represent the mass of a determining element, the masses of ‘its’ elements determined being a, ), ¢, d, e, &c.; Q the mass of another determining element, a, B; y, 5, ¢, &c., being the masses of ‘its’ elements determined; so that a@ and a, b and £, ¢ and y, d and 6, and e and « shall represent the same elements; and, further, that p+a@ and q+, P+) and Q+y¥, P+e and Q+a, &c., are decomposed by double affinity, so that the new products will remain neutral. We shall observe that the masses a, b, c, d, e, &c., bear to each other the same ratio as the masses a, 8, y, 6, ¢, &c. Such is the discovery of Richter as he himself published it in 1795 in the fourth part of his ‘ Mittheilungen uber die neueren Gegenstiinde der Chemie.’ This is not all. We owe to his penetration another important discovery which is closely connected with the one we have just mentioned. We shall now direct our attention to metallic salts properly so called. When two of these salts are de- composed by double affinity—that is to say, when they exchange their acids and bases—the metal of the one finds in the other exactly the quantity of oxygen neces- sary to keep it dissolved in the acid; in other words, the quantities of different metals necessary for the formation of neutral salts absorb the same quantity of oxygen when they dissolve in a given weight of acid. This proposition, which is, moreover, very accurate, 2 18 THE ATOMIC THEORY. assumed a clearer form when Lavoisier some time afterwards worded it thus :—The different quantities of oxides which combine with a given weight of acid con- tain the same quantity of oxygen. Richter followed up these investigations with great success. He admits that the ratios in which oxygen combines with other bodies, particularly metals, are perfectly fixed, and that the quantity of oxygen fixed by a metal during solution in an acid is not always the same as that which it absorbs when heated in contact with air. He is thus led to distinguish several degrees of oxidation, notably in the case.of iron and mercury. ‘The latter forms two oxides capable of producing salts. Each of these salts presents a perfectly fixed composition, and passes with- out alteration of composition by double exchange from one salt to another. These researches date from the close of the last century, and it seems as if, from the manner in which they were conceived and expressed, the influence of Lavoisier had made itself felt, unknown to the author and in spite of his opposition to the doctrines of the reformer. The very fact of this oppo- sition seems, in a great measure, to have been the cause of the discredit thrown upon the labours of Richter ; his time was not yet come; other topics created more interest; and in Germany, as also in France and England, men’s minds were engrossed by the influx of new ideas. | There is some difficulty in harmonising the signifi- cation and even the publication of Richter’s great dis- coveries with the phlogistic theories which he main- tained, and which apparently influenced his observations. —- RICHTER—LAW OF PROPORTIONALITY. 19 Strictly speaking, we can understand that he could have regarded acids as undecomposable bodies, for he only considered their relative weights, which are independent of their constitution. But when we turn to his opinions upon the nature of oxides, upon the fixity of their com- position, upon the equality of the weights of oxygen absorbed by metals when dissolved in equivalent quan- tities of acid, how can we reconcile these correct and simple notions with the erroneous conception of phlo- giston? It must be confessed that Richter adapted all this to his theory. He held that the metallic calces or oxides were formed by the combination of metals with oxygen, causing a loss of imponderable phlogiston and most curious contortion of the phlogistic theory. Had he but said heat instead of phlogiston, he would have been quite right. We may, therefore, absolve Richter on this head ; but his contemporaries were more severe, and he himself confesses that in 1799 he was declared by the partisans of antiphlogistic doctrines to have taken leave of his senses, The profound but perplexed author of the great discovery in question—the proportionality which exists between the weights of elements in chemical combina- tions—was fortunate in having an intelligent and ingenious commentator. G. E. Fischer published in 1802 a German translation of Berthollet’s ‘ Researches upon Affinity,’ and further endeavoured in this work to explain and simplify the deductions which Richter made from the fact of the permanence of neutrality after the decomposition of two neutral salts. He succeeded, and simplified the demonstration of the law of proportionality 20 THE ATOMIC THEORY. in the following manner :—Richter had given a series of neutralisation for each acid and each base; he had determined the quantities of bases which saturate 1,000 parts of sulphuric acid, 1,000 parts of nitric acid, and 1,000 parts of hydrochloric acid; and then, again, he had indicated the quantities of acids which would saturate 1,000 parts of each base. Though admitting that the quantities of acids and the quantities of bases composing these series are proportional, he uselessly multiplied the number of the latter. Fischer saw that they might be reduced to one by giving the ratio which the quantities of acids and bases contained in the series bear to one number, 1,000 parts of sulphuric acid. In fact, he drew up the first table of chemical equivalents as follows :— Bases. Acids. Alumina. : pips Fluoric acid. o 427 Magnesia ‘ . 415 Carbonic ,,; « pr tact Ammonia : . 572 Sebacic mh? Se - 7; UG Lime ; : Athy © Muariatic’. 5;7°3 e TiS Soda 4 : fate tits, Oxalic wae a FLO8 Strontia . ; . 1,829 Phosphoric,, . uo Long Potash . A . 1,605 Formic Seas oat ists) Baryta . : . 2,222 Sulphuric ,, . . 1,000 Succinic ,, . . 1,209 Nitric ak . 1,405 | Acetic vs hee . 1,480 Citric a ate . 1,583 Tartare? 05,°% . 1,694 The figures in these two columns represent equiva- lent quantities of acids and bases. To neutralise a given base of the first series with a given acid of the second series, we must take of that base and that acid the quantity indicated by the accompanying figures. ane ri RICHTER—LAW OF PROPORTIONALITY. 21 The ratios of neutrality between the bases and acids are expressed by these numbers, and the table demonstrates in a striking and convenient form the coon a of a large number of neutral salts. The foregoing table forms part of a note which was inserted by Fischer in Berthollet’s ‘ Chemical Statics.’ ! It is to his translator that the latter owed his acquaintance with the researches of Richter. He had treated of the same subject in a chapter of the ‘Statics’ entitled * Acidity and Alkalinity,’ and had mentioned the opinion of Guyton de Morveau upon the inference which may be drawn from the permanence of neutrality after the decomposition of certain neutral salts, so as to calculate beforehand or control the composition of the salts produced. Both chemists, however, acknowledged that Richter had anticipated them in this direction. Berthollet expresses himself on this point as follows :— ‘The preceding observations appear to me necessarily to lead to the conclusion that in my researches I have only hinted at the laws of affinity, but that Richter has positively established the fact that the different acids follow proportions corresponding with the different alkaline bases in order to produce neutrality. This fact © may be of the greatest utility in verifying the experi- ments which have been made upon the proportions of . the elements of salts, and even to determine those which have not yet been decided by experiment, and so furnish the surest and easiest method of accomplishing this object, so important to chemistry.’ Thus Berthollet admitted the law of proportionality, 1 Vol. i. p. 134, 1802. 22 THE ATOMIC THEORY. discovered by Richter, though at the same time he questioned the fixity of certain chemical combinations. He considers it possible for neutral salts, precipitated in an insoluble state or separated as crystals from their solutions, to exist in physical conditions compatible, according to him, with a fixed composition. As we remarked above, we owe to Richter another important discovery. He observed that the quantities of different metals which dissolve in a given weight of acids also combine with the same weight of oxygen. This discovery met with no recognition, and was made afresh by Gay-Lussac in 1808. It was the same with the following fact, which Richter established: that cer- tain metals, such as iron and mercury, have the power of combining with oxygen in several proportions, so as to form two degrees of oxidation. Proust rediscovered this fact, and laid great stress upon it in his discussion with Berthollet, but he failed to observe that the quan- tities of oxygen contained in the different oxides of a given metal increase in a very simple ratio.! We find, therefore, that at the close of the last and the commencement of the present century a number of definite facts were discovered concerning the composi- tion of salts and chemical compounds in general, but that these facts were isolated and without connection. Their deep signification escaped the observation of 1 Proust admitted that 1,000 parts of copper combine with 172 to 18 parts of oxygen to form the first or sub-oxide of copper, and with 25 parts of oxygen to form the second or black oxide. The correct numbers are 12°6 and 25:2. Had the analyses of the two oxides been more correct, Proust might have recognised the law of multiple proportions, DALTON—HYPOTHESIS OF ATOMS. 23 chemists, and the theoretical link which unites them was entirely unknown. It was reserved for an English _ chemist to complete them by a discovery of the first order and to arrange them by an hypothesis both simple and fruitful. Ill. In 1802 John Dalton, at that time a professor in Manchester, was investigating the action of air upon bin- oxide of nitrogen in the presence of water. He observed that the oxygen contained in 100 volumes of air united with either 36 or 72 volumes of binoxide of nitrogen, leaving a residue of pure nitrogen gas above the water. He concluded from this fact that oxygen combined with a certain quantity of binoxide of nitro- gen or with double that quantity, but not with any intermediary quantities, nitric acid being formed in the first instance, nitrous acid in the second. In this - observation we have the germ of the law of multiple proportions, although it was not as yet formally stated in the memoir in question.' It was announced at the same time as the atomic theory, by which it is theoreti- cally explained, in a communication by Dalton to Thomson in August 1804. He was then studying the composition of marsh gas, and observed that for the same quantity of carbon this gas contains a quantity of hydrogen exactly double that which is contained 1 Memoirs of the Literary and Philosophical Society of Man- chester, Vol. Vv. p. 535. 24 THE ATOMIC THEORY. in bicarburetted hydrogen. We learn further from Thomson ! that the foundations of Dalton’s theory were derived from his researches into the composition of combinations of oxygen and nitrogen, and that, in fact, the observation mentioned above upon the absorption of oxygen by binoxide of nitrogen first gave him an in- sight into the composition of those combinations. Be- yond this it is difficult to affix any precise date to the discoveries of Dalton, or at least to trace the logical sequence and successive evolution of his ideas, and to separate the origin of the law of multiple proportions from the origin of the conception of the atomic theory. In fact, in a memoir upon the absorption of gases by water, read in October 1803 before the Literary and Philosophical Society of Manchester, Dalton attributed— erroneously, moreover—the unequal solubility of the different gases to the circumstance that their ultimate particles are not equal in weight, and that the ele- mentary atoms of which they are formed are not equal in number. ; In this memoir he remarked that he had for some time been occupied with an endeavour to determine the relative weights of the ultimate particles of bodies—a new and, as he says, most important consideration. Without laying any special stress upon the development of these ideas, he gives in his memoir the first table of atomic weights as follows :— Hydrogen . ‘ : ° : ; 5 aed Oxygen : . : : . . . 2be Nitrogen. : , ° ° : s £2 1 History of Chemistry, vol. li. p. 289. London, 1831. DALTON—HYPOTHESIS OF ATOMS. 25 Phosphorus Sulphur Carbon . ‘Water . Ammonia ., : Protoxide of nitrogen . Binoxide of nitrogen Nitric acid : ’ Phosphoretted hydrogen Sulphuretted hydrogen Sulphurous acid Sulphuric acid Carbonic oxide Carbonic acid Marsh gas Olefiant gas . Ether . Alcohol. 7:2 14-4 4°3 6°5 5:2 13:7 9°3 15°2 8°2 15-4 193. 25°4 98 15°3 6°3 53 9°6 15:1 Let us first remark that not only the law of multiple proportions but also the atomic theory are clearly con- tained in this table. following data :-— 4°3 of carton are com- bined with 1 of hydrogen in 5:3 of olefiant gas. 4-3 ” ” 2 4:3 ” ” 4-3 a 2 Roe 14-4 of sulphur 4 55 14-4 » » 2x55 4:2 of nitrogen __,, 55 4°2 3 wie aa x 5:6 Sx42 55 s 55 ” ” This result is evident from the 6°3 of marsh gas. 5:5 of oxygen in 9°8 of carbonic oxide. 15°3 of carbonic acid. 19°9 of sulphurous acid. 25-4 of sulphuric acid. 9-7 (9:3) of binoxide of nitrogen. 15°2 of nitric acid. 13°9 (13°7) of protoxide of nitrogen, It is true that these figures are very inaccurate, but the inaccuracy of the numerical data cannot conceal or 26 THE ATOMIC THEORY. diminish the grandeur and simplicity of the theoretical conception. Dalton here regards chemical combinations as formed by the addition of elementary atoms, the rela- tive weights of which he endeavours to determine, referring these weights to one of the elements—hydrogen —asunity. Whentwo bodies combine in several propor- tions, the combination can only be effected by the addi- tion of entire atoms: it follows, therefore, that, the proportion of the one body remaining constant, the proportions of the second must be exact multiples of each other. It is, then, clear that as early as the year 1803 or 1804 Dalton had, if not formally stated, at least conceived and implicitly admitted the law of multiple proportions, as well as the atomic hypothesis, which may almost be regarded as the theoretical representa- tion of the fact of fixed and of multiple proportions. A few years afterwards he gave his opinion upon this subject in the following terms: !-— ‘In all chemical researches great importance has with justice been attached to the determination of the relations according to which elements unite to form compound bodies; but, unfortunately, the subject has not been followed up, though the consideration of these relations might have led to important conse- quences concerning the relative weights of the smallest particles, or atoms, of bodies.’ From the year 1804 the atomic theory inspired all Dalton’s labours and influenced all his thoughts; he confesses himself the influence which this idea had upon 1 A New System af Chemical Philosophy, part i. London, 1808. DALTON—HYPOTHESIS OF ATOMS. 27 him in his representation of the constitution of marsh gas, which he was studying in 1804. It occasioned the discovery of multiple proportions, and afterwards fur- nished, by a happy reaction, a solid foundation for the atomic hypothesis. The latter was not, however, new to the epoch of Dalton. Not to speak of the atomists of the seventeenth century, who had revived, though at the same time distorted, the ancient conception of the Greek philosophers, we must not forget that Van Helment, N. Lemery, and Boerhaave had mentioned the indivisible particles of bodies, and had termed them ‘atoms,’ and that Boyle had tried to explain the differ- ences between chemical attractions by the inequality of the ‘ massulz’ or particles. _ This was a correct conception: the ultimate parti- cles of bodies differ in their relative weight, and doubt- less in their size and form. In 1790 Higgins opposed this hypothesis, erroneously attributing the same weight to atoms which combine in very simple proportions to form compound bodies. Thus Higgins admitted that one ponderable part of sulphur in sulphurous acid is combined with one ponderable part of oxygen, and in sulphuric acid with two ponderable parts of oxygen, and that these compounds may be represented as con- sisting, the first of one atom of sulphur with one atom of oxygen, the second of one atom of sulphur with two atoms of oxygen, the atoms of these two elements being, moreover, of the same weight. He consequently re- presented these compounds as formed by the union of particles or atoms, of the same weight, but united in different proportions—one of nitrogen for two of oxygen 28 THE ATOMIC THEORY. in binoxide of nitrogen, and one of nitrogen for five of oxygen in nitric acid. This is all perfectly clear, but the starting point is wrong. The proportions in which bodies combine do not represent equal weights. This was an established fact in the time of Higgins, and he is obliged to ac- knowledge it in the case of water, the two elements of which unite in very unequal proportions ; if therefore, as he admitted, water was composed of one atom of oxygen and one atom of hydrogen the atoms of these two elements could not be equal in weight. Higgins’s conception was, therefore, spoilt by errors and contradic- tions, and it is useless to attempt to represent him as one of the authors of the modern atomic theory. This honour belongs to Dalton alone. This great man began to consolidate and publish his views about the year 1808. Thomson and Wollaston were at that time developing the law of multiple proportions. In a* memoir upon oxalic acid Thomson showed that the acid oxalate of potash contains twice as much acid as the neutral oxalate. Wollaston demonstrated that this law applies to the quantities of bases and acids contained in basic and acid salts, these quantities bearing a simple ratio to each other. He showed that this is the case in the com- pounds of potash and soda with carbonic and sulphuric acids, and especially in the compounds of oxalic acid with potash. He points out that the latter are three in number, and that the quantities of acid which they contain for the same proportion of base increase as the numbers 1, 2, 4. At this time Dalton himself published his theory in DALTON’S NOTATION. 29 the first part of his ‘ New System of Chemical Philoso- phy,’ which appeared in 1808. The new and compre- hensive idea of representing compound bodies as formed of groups of atoms, fixed in number, and possessing different, but at the same time fixed, relative weights, might, it seemed to him, be graphically expressed by the adoption of symbols representing these atoms, and grouped in such a manner as to indicate the composi- tion of bodies. Each atom was represented by a small circle bearing a particular sign. This is the origin of chemical notation, the lan- guage of symbols and numbers, which is clearer and more concise than that of words, and has since been a great instrument of progress in science and a great assistance in instruction. In the work mentioned above Dalton gives a new table of atomic weights, more com- plete and less incorrect than the preceding one. We give a few of these atomic weights. Dalton’s Correct Atomic Weights. Numbers. Hydrogen : : - ot 1 Nitrogen . ° . . : ae 4°66 Carbon . : ° : in a? EF 6 Oxygen . E - : ER i 8 Phosphorus . : ‘ a. o 10°3 Sulphur . : : : were 16 Iron. ° : ; : - 38 28 Zinc. : . ° ° . 56 65°2 Copper . . ; : - 56 64:5 Lead d ; ‘ ‘ - 95 104 Silver. ° “ ° - 100 108 Platinum . ° ‘ ; - 100 98:5 Gold : : ; ‘ . 140 197 Mercury . ~ ‘ . « 167 200 30 THE ATOMIC THEORY. We have omitted in this table the atomic weights of the alkalies and earths which are still placed among the elements, though Dalton must have already been acquainted with the great discovery of H. Davy upon the nature of the alkalies. The above figures give, however, a sufficiently good idea of the accuracy, or rather the inaccuracy, to which Dalton had attained in his own determinations, or in the discussion of those of others. At the same time they show us the exact sense in which we must regard these atomic weights. They are not, properly speaking, atomic weights in the sense which we now ascribe to the term; they are proportional numbers referred to unity, which represents the weight of hydrogen in hydrogen compounds. ‘This may be seen from the following table, in which, for the sake of brevity, we have employed the symbuls in use at the present day :— Atomic Weights. Water contains 1 at. H, which weighs 1, and 1 at. O, which weighs 7 . ° 8 Sulphuretted hydrogen Soran 1 at, TH, hice we} srg ib aa 1 at. 8, which weighs 13 - : 14 Ammonia contains 1 at. H, which weighs 1, se 1 “ N, milion weighs 5 : ; 6 Olefiant gas contains 1 at. H, rainth wens a; ve 1 eh C, Poe weighs 5 * 3 6 Marsb gas contains 2 at. H, ian wah 2, ani 1 fe C, hich weighs5 . =f Carbon protoxide contains 1 at. 0, which meee 5, BN 1 = 0, which weighs 7 . 3 12 Carbonic acid contains 1 at. C, eich Hei? 5, oa 2 Ja 0, which weigh 14 . ‘ 19 Protoxide of nitrogen contains 9 at. N, hich Loigt LO; a 1 at. O, which weighs 7 . ee ‘ 2 . ; - te ES DALTON’S NOTATION. 31 Atomic ; ‘ Weights. Binoxide of nitrogen contains 1 at. N, which weighs 5, and 1 at. O, which weighs 7 ‘ he Nitrous acid contains 2 at. N, hich weld 10, aid 3 af O, weigh 21. : » ol Nitric acid contains 1 at. N, which Soinhe 5, Sie 2 ve O, which weigh 14, - . : : p - : : 19 We see that the atomic weights of oxygen, sulphur, nitrogen, carbon, and phosphorus are deduced from the composition of their combinations with hydrogen, in which the existence is admitted of one atom of hydro- gen combined with one atom of another body; and when there are two combinations with hydrogen, as is the case with carbon, the atomic weight is determined from that containing the least quantity of hydrogen. Thus the atomic weight of carbon is the quantity of carbon combined with 1 of hydrogen in olefiant gas. In marsh gas this quantity of carbon is combined with 2 of: ~ hydrogen. Such are the principles by which Dalton was guided in the determination of atomic weights, as they were conceived by him in 1808, and in the notation which was deduced from them. These principles are clearly demonstrated in the following table, which expresses the atomic constitution of the compounds mentioned above; the formule are analogous to those now in use :— Dalton’s Notation (1808). Atomic (Molecular) . Weight. Formule. 8 of water are represented by . ; a. HO 14 ,, sulphuretted hydrogen ; F . HS 6, ammonia . ‘. r : ; a: SELES 6 ,, olefiant gas . : ‘ . ‘ « HO 32 THE ATOMIC THEORY. Dalton’s Notation (1808)—continued. Atomic (Molecular) = Weight. 7 of marsh gas 12 19 17 12 31 19 », carbon protoxide . 5 carbonic acid », protoxide of nitrogen », binoxide of nitrogen ») nitrous acid . . » Ditricacid . ‘ Formule. 2 co co, N,0 NO N,0, _ NO, CHAPTER II. LAW OF VOLUMES. GAY-LUSSAC—AVOGADRO AND AMPERE—BERZELIUS. 1p TuE atomic weights established by Dalton were really proportional numbers; they represented the proportion in which bodies combine, expressed by the relative weights of their ultimate particles. The atoms of simple bodies are equivalent to each other. We may, therefore, consider the terms atomic weights, propor- tional numbers, and equivalents as at this time syno- nymous. We owe the last term to Wollaston; H. Davy preferred the expression ‘ proportional numbers.’ The atomic constitution of bodies follows very naturally from the ideas of Dalton. In binary com- pounds atoms unite in the ratio of 1 to 1, and in multiple compounds formed by two given elements in the ratio of 1 to 1,1 to 2,1 to 3,2 to3, &. This simple conception, which is clearly demonstrated in the table upon the preceding page, had to be modified in accord- ance with Gay-Lussac’s great discovery. 34 THE ATOMIC THEORY. The relations between the combining volumes of gases are very simple, and the volume of the compound formed bears, moreover, a very simple ratio to the sum of the volumes of the combining gases. This proposition embraces a great number of facts, which present no exceptions and which together consti- tute a great law of nature, the law, namely, of Gay- Lussac. Suitably interpreted, it has become one of the foundations of chemical science. ‘The following are the facts; the interpretation will be developed presently :— 2 vol. of hydrogen unite with 1 vol. of oxygen to form 2 vol. of aqueous vapour.! 2 vol. of nitrogen unite with 1 vol. of oxygen to form 2 vol. of nitrogen protoxide. 1 vol. of nitrogen unites with 1 vol. of oxygen to form 2 vol. of nitrogen dioxide. 1 vol. of nitrogen unites with 2 vol. of oxygen to form 2 vol. of nitrogen peroxide. 1 vol. of chlorine unites with 1 vol. of hydrogen to form 2 vol. of hydrochloric acid gas. 2 vol. of chlorine unite with 1 vol. of oxygen to form 2 vol. of hypochlorous anhydride. 1 vol. of nitrogen unites with 3 vol. of hydrogen to form 2 vol. of ammonia. | ; 2 vol. of carbon protoxide unite with 2 vol. of chlorine to form 2 vol. of phosgene gas. | 2 vol. of ethylene unite with 2 vol. of chlorine to form 2 vol. of vapour of ethylene chloride, Thus it appears that very simple relations exist not only between the volumes of gases entering into combination, but also between these volumes and the volume occupied by the gas or-vapour of the com- 1 The volumetric composition of water WAS discovered in 1805 by Gay-Lussac and Humboldt. This observation formed the starting point of Gay-Lussac’s discoveries, GAY-LUSSAC—LAW OF VOLUMES. 35 pound body. It should be remarked, moreover, that, as far as we know at present, the volumes of the combining gases are always reduced to 2 vol. after combination." Bearing this fact in mind, we may return to our his- torical account. } Gay-Lussac rendered unexpected assistance to the ideas of Dalton. The fixed relations which are ad- mitted between the weights of elements entering into combination, the simple relations which exist between the weights of a given element in the multiple combina- tions of that element, are again encountered when the combining volumes of gases are considered. Connecting these two orders of facts, and following up the interpre- tation which Dalton gave of the former, may we not conclude that the relative weights of the gaseous volumes entering into combination exactly represent the relative weights of the atoms—in other words, that there exists a simple relation between the specific gravities of elementary gases and their atomic weights ? Gay-Lussac perceived this simple relation, and Berzelius defined it a few years afterwards; but Dalton refused to accept it, ignoring and repudiating the solid support which the great French chemist gave to his ideas. In fact, the relation which exists between the densities of gases and their atomic weights is not so simple as we should at first sight be led to expect, and as for a long time it was thought to be. It is a difficulty which will soon be apparent, and 1 This applies particularly to the first seven cases, in which the -volumetric relations are as simple as possible, and cannot be reduced. The two last cases will be discussed presently. 36 THE ATOMIC THEORY. which has only quite recently been overcome, after sixty years of investigation and labour. Nevertheless the theoretical conception which embraces the two orders of phenomena in question, and which establishes a link between fixed and multiple chemical proportions and the law which regulates the combinations of gaseous volumes, was accurately formulated in 1813 by the Italian chemist Amedeo Avogadro. Starting from the discoveries of Gay-Lussac, Avoga- dro arrived at the conclusion that there exists a simple relation between the volumes of gases and the number of elementary or compound molecules which they con- tain. The most simple and at the same time the most probable hypothesis which can be brought forward upon this point is, he says, to admit that all gases contain in a given volume the same number of integral mole- cules.!| These molecules must, therefore, be equi- distant from each other in ditferent gases, and placed at distances which, in relation to the dimensions of the molecules, shall be exactly sufficient to neutralise their mutual attraction. This hypothesis, according to Avogadro, is the only one which gives a satisfactory explanation of the fact of the simplicity of relations between the volumes of gases entering into combina- tion. The following result of this hypothesis is im- portant: if it is true that equal volumes of gases contain the same number of molecules, the relative weights— that is to say, the densities of equal volumes—ought to represent the relative weights of the molecules. Thus the molecular weights of hydrogen, oxygen, and nitro- 1 Journal de Physig e, vol. xxxili. p. 58. HYPOTHESIS OF AVOGADRO. ST gen will be expressed by the ratio of their densities —i.e. 1, 15, 13.1 But in considering the molecular weights of compound bodies we encounter a difficulty, which arises from the difference of contraction experi- enced by gases in the act of combination. Supposing water to be formed by the union of two volumes of hydro- gen and one volume of oxygen contracted to one volume, it is clear that the weight of this single volume, com- pared with that of one volume of hydrogen, would be 17 (15+2);? or again, supposing one volume of . ammonia were formed by the contraction of three volumes of hydrogen and one volume of nitragen, the weight of this volume of ammonia must be 16. Now, experiment proves that the densities of aqueous vapour and of ammonia are half the above numbers —namely, 8°5 and 8—a result which agrees with the fact that two volumes of hydrogen and one volume of oxygen are condensed into two volumes of aqueous vapour, and, on the other hand, that three volumes of hydrogen and one volume of nitrogen are condensed into two volumes of ammonia. Since one volume of aqueous vapour contains only one volume of hydrogen and 5 volume of oxygen, a molecule of water can only be formed of one molecule of hydrogen and 4 molecule of oxygen ; and, for the same reason, one molecule of ammo- nia must be formed of 14 molecule of hydrogen and 4 molecule of nitrogen, and a molecule of hydrochloric acid gas of $ molecule of hydrogen and 4 molecule of chlorine. It follows that the matter contained in the unit 1 The correct numbers are 1, 16, 14. ? These numbers are those of Avogadro, 38 THE ATOMIC THEORY. of volume of the elementary gases does not represent the ultimate particles which exist in certain combinations of these gases, for the matter contained in one volume of oxygen, nitrogen, hydrogen, and chlorine must be halved in order to form the quantity of water, ammonia, and hydrochloric acid gas contained in one volume of — these compounds. This is the difficulty which strikes us, a difficulty which, according to-Avogadro, is easily solved by supposing that the wntegral molecules, an equal number of which are contained in the gases or vapours of elementary bodies, are themselves composed of a certain number of elementary molecules of the same kind, just as the antegral molecules of compound gases and vapours are formed of a certain number of element- ary molecules of different kinds. This is a fundamental idea, and at that time was quite new. Whilst Dalton had only distinguished one kind of ultimate particles—atoms—Avogadro admits the existence of two kinds—an important distinction, which has been established by the progress of science. Avogadro’s elementary molecules are atoms, while the integral molecules, which are equidistant from each other in gaseous bodies, which are set in motion by heat and excited by affinity, are what we at the present time call molecules. Ideas similar to those of the Italian chemist were published in 1814 by Ampére,' who established a distine- 1 A letter from M. Ampére to M. le Comte Berthollet upon the determination of the proportions in which bodies combine, from the respective number and arrangement of the molecules of which their integral molecules are composed. (Annales de Chimie, vol. xe. p. 43.) HYPOTHESIS OF AMPERE. 39 tion between particles and molecules. The particle, he says, ‘is a collection of a definitive number of mole- cules in a definite situation, occupying a space incom- parably greater than that of the volume of the molecules.’ And he adds, ‘ When bodies pass into the gaseous state, their several particles are separated, by the expansive force of heat, to much greater distances from each other than when the forces of cohesion or attraction exercise an appreciable influence, so that these distances depend entirely upon the heat to which the gas is sub- jected, and that, under equal pressure and temperature, the particles of all gases, whether simple or compound, are equidistant from each other. The number of par- ticles is, on this supposition, proportional to the volume of the gases.’ This passage is so remarkable that we have quoted it word for word. But, as a natural consequence of Ampére’s proposition, it follows that, the distances between the gaseous particles being the same, and depending solely upon pressure and temperature, the same variations of pressure and temperature should produce the same change of volume in the gases. This, as we know, is actually the case, and this great physical fact of the sensible equality of the expansion of gaseous volumes follows as the result of the principle propounded by Avogadro and Ampére —namely, the equality of the number of particles contained in equal volumes of gases and the equality of the distances by which they are separated. Neither of them laid any stress upon this result, which clearly supported their hypothesis. Ampére clearly alludes to it in the passage quoted 40 THE ATOMIC THEORY. above, but he only adds that if the hypothesis which he has offered ‘agrees with the established results of experiment, and if such results can be deduced from it as will be confirmed by subsequent experiments, it may acquire a degree of probability approaching to what in physics is termed certainty.’ Thus, equal volumes of gases and vapours contain the same number of particles, and the latter are formed of groups of molecules. This, in other words, is the con- ception of Ampere. From geometrical considerations Ampére was led to conclude that each particle consists of four molecules. ‘In accordance with this idea,’ he says, ‘ each particle should be regarded as a collection of a definite number of molecules in a definite position, occupying a space incomparably greater than that of the volume of the molecules; and, in order that this space may be of three comparable dimensions, one particle must comprise at least four molecules. In order to express the respec- tive position of the molecules in a particle, we must conceive, by means of the centre of gravity of these molecules, to which we may suppose them to be reduced, planes so placed as to leave on the same side all the molecules which are outside each plane. Supposing that no molecule should be contained in the space included between these planes, this space will be a polyhedron, of which: each molecule will occupy an angular point. If the particles of oxygen, nitrogen, and hydrogen are composed of four molecules, it would follow, accord- ing to Ampére, that those of nitrogen dioxide are HYPOTHESIS OF AMPERE. 41 composed of four molecules—namely, two of oxygen and two of nitrogen—and those of nitrogen protoxide of six molecules—namely, four of nitrogen and two of oxygen. Thus when gases combine together the molecules contained in the unit of volume of either of the com- bining gases, and which form its particle, are not always contained integrally in the unit of volume of the com- pound gas. These molecules may be grouped or separated in forming a particle of the compound gases. Let us translate this idea of Ampére’s into the language of formule, confining ourselves to the examples quoted above. The unit of volume of gas contains— Oxygen . : - : : : aamye Hydrogen : ‘ - 3 : welig Nitrogen . " ‘ : : : aly Water. ‘ ‘ ° . , » H,O, Nitrogen protoxide ; : : ae PoE Nitrogen dioxide . : . ‘ - N,O, Ammonia . . : ° ° Ee N*HG The analogy will at once be seen between this conception and that which is generally adopted at the present day, and which is expressed by the following formule :-— Two volumes of gas contain— Oxygen . ; 3 , ‘ ; Ls Od Hydrogen , : é . ° wie EL Nitrogen ’ . ; : fpr Re3 Water - ; 3 ‘ : : , - HO Nitrogen protoxide ; ‘ : « N,O Nitrogen dioxide . ‘ : 2 a WO Ammonia ° ‘ ‘ 3 : - NH; 3 42 THE ATOMIC THEORY. For the first time gases are compared under the . same volume, and the matter contained in equal volumes represents the magnitude of the molecules, which is a most essential point. The integral molecules of Avogadro and the particles of Ampére.are, in fact, the material parts contained in the unit of volume. These integral molecules, or particles, may be subdivided into elementary molecules, or simply into molecules ; and Ampére, from geometrical considerations, ingeniously, though uselessly, multiplies the number of the latter. Both for the first time introduce into science the dis- tinction between two kinds of ultimate particles, and admit that the number of integral molecules or particles is proportional to the gaseous volume. We now give a more simple form to the same ideas by admitting that gases (and all other bodies) are formed of molecules and atoms; and, in order to avoid the subdivision of molecules referred to one volume, we find it more convenient to refer them to two volumes, assigning the term molecule to the matter contained in two volumes. When for an elementary gas this molecule, as is often the case, consists of two atoms, the atom represents the matter contained in one volume; but the general rule is that equal volumes of gases and vapours contain the same number of molecules, and conse- quently that the relatwe weights of these molecules are proportional to the densities. This is the law, or, if we prefer it, the hypothesis, of Avogadro and Ampére, | for we must acknowledge that a hypothesis here becomes mingled with the interpretation of positive facts ; but the hypothesis seems to be legitimate, and will be justified INTERPRETATION OF BERZELIUS. 43 presently. Chemists long ignored its import. Another conception was soon substituted for it, which acquired an important place in science, without, however, gaining general consent, though supported by the authority of a great name—that of Berzelius. ? Hae is In 1813 Berzelius conceived the idea that, in order to represent the composition of bodies, we must take into consideration the relative volumes in which simple gases combine to form compound gases. He developed this idea in a memoir upon the nature of hydrogen. It is well known, he says, that one volume of a gaseous body combines with one, two, or three volumes of another gaseous body; we have only to determine, therefore, the weights of these volumes to know the relative weights according to which the gases combine with one another. It is obvious that the numbers thus obtained are similar to the atomic weights of Dalton, without, however, being identical. Again, though declaring himself a partisan of the atomic hypothesis, Berzelius held that it was better to keep to the theory of volumes, as having the advantage of being founded on well-established facts. In order to express the com- — position of bodies by weight from this point of view, it — is only necessary to find:—firstly, the number of element- ary volumes which unite to form a compound body; secondly, the relative weights of these volumes—that is to say, the densities of the elements in the gaseous state. Hence the importance of the determination of gaseous densities. In 1814 Berzelius worked out a 44 THE ATOMIC THEORY. number of these densities, referring them to that of oxygen, which he took as 100. Thus, the weight of an ‘elementary volume’ of oxygen being 100, that of an elementary volume of hydrogen will be 6:218, that of an elementary volume of nitrogen 88°6, that of an elementary volume of chlorine 221°4.! These weights express the quantities of the bodies which enter into combination. In a great number of cases there are many difficulties in the way of the determina- tion, which can only be made in an indirect manner. In fact, at this period Berzelius was only acquainted with two simple gases of-which he could obtain the density—oxygen, namely, and hydrogen. He at that time regarded nitrogen as a body composed of nitrogen and oxygen, and was not as yet converted to Davy’s opinion concerning the simple nature of chlorine. It would, therefore, be only in a very small number of cases that the weight of ‘ elementary volumes’ could be directly proved by experiment, and to determine the . weight of non-gaseous simple bodies we should be forced to have recourse to hypotheses upon the composition by volumes of non-gaseous elements. Let us take a few: 1 We can judge of the accuracy of these numbers by referring them to 6-218 of hydrogen taken as unity. They then bear the following ratios to each other :— Correct Numbers. Hydrogen : . . oi Bs. 1 Oxygen . A ; : . 16:08 16 Nitrogen : ‘ : . 14°45 14 Chlorine . 5 Ny . 361) 35°5 INTERPRETATION OF BERZELIUS. 45 examples. What is the weight of an ‘elementary volume’ of carbon? We know that carbonic acid gas contains its own volume of oxygen. But do the two volumes of carbonic acid gas, which contain two volumes of oxygen, contain one volume or two volumes of carbon vapour? In the first case the three volumes, two of oxygen and one of carbon vapour, are reduced to two from the effect of combination, a condensation similar to that of water; in the second the condensation is one-half. Thus, on the first hypothesis, it is evident that the weight of the ‘elementary volume’ of carbon is twice that which is attributed to it in the second. Referred to oxygen as 100, the weight of the ‘ element- ary volume’ of carbon is 75:1 in the first case and 37:55 in the second; and the corresponding formulx of carbonic acid gas will be CO, and ©,0,. Berzelius adopted the first hypothesis, allowing himself to be guided by analogy. It seemed to him probable that the condensation of the elements of carbonic acid gas was similar to the condensation of the elements of water. He also at this time admitted that the powerful bases must be composed of two elementary volumes of oxygen and one volume of metal. The composition of the oxides of sodium, potassium, calcium, iron, zinc, and lead was, therefore, represented by the formule Na0O,, KO,, CaO,, FeO,, PbO,, the weights of the element- ary volumes of a great number of metals thus assuming a value double that which Berzelius attributed to them later. The theory of volumes, as it stood at that time, was therefore bristling with hypotheses and full of uncer- 46 THE ATOMIC THEORY. tainties. And yet this conception long held its ground in science, especially in France, where at a certain period it was the fashion to express the composition of bodies in § volumes,’ under the impression that the substitution of volumes for atoms had the advantage of offering a representation more in accordance with facts. But in reality it was not so: the volume occupied by carbon vapour, and the degrees of condensation of the elements of carbonic acid gas, were hypothetical ideas, and these ‘elementary volumes’ represented the atoms themselves, at least in notation. Berzelius recognised this fact in 1818. In his essay upon the theory of chemical proportions he modified considerably the views which he had published in 1813. The prevailing idea is no longer that of establishing the system of atomic weights upon the theory of volumes. Though still giving weight to the indications furnished by this theory, he endeavours to reconcile it with what he terms the ‘corpuscular theory,’ which is founded upon chemical proportions. The indivisible corpuscules, or the ultimate particles of bodies, are designated atoms—the most convenient term, because the one most in use. We might call them particles, molecules, or chemical equivalents, as all these terms appear to be synonymous to Berzelius. The relative weights of these atoms represent chemical proportions. The fixed pro- portions, which had been recognised for weights, again appeared in gaseous combinations for volumes. Thus the theory of volumes and the atomic or corpuscular theory led to the same results, as far as the ponderable relations of elements in combinations are concerned: BERZELIUS—CORPUSCULAR THEORY. 47 what is called atom in one is called volume in the other. It would seem, therefore, as if we might assimilate the two notions, which indeed is necessary in the case of simple gases. Equal volumes of the latter contain the same number of atoms, under the same conditions of temperature and pressure. Berzelius observes that this law does not apply to compound gases; for, he says, it sometimes happens that a volume of a compound gas contains fewer atoms than an equal volume of a simple gas. Thus one volume of aqueous vapour contains one- half as many atoms (compound atoms, molecules) as one volume of hydrogen. Such was the manner in which Berzelius, about 1818, expressed the atomic hypothesis, which he founded partly upon chemical proportions and partly upon a peculiar conception of the law of volumes. This con- ception was not a very happy one. Not to mention the difficulty which he created by applying the same term, atoms, to the ultimate indivisible particles of simple bodies and to the complex molecules of compound bodies, a confusion which had been avoided by Avogadro and Ampére, Berzelius at this time introduced into the language of science a formula which long held its ground, and which must now be considered as erroneous—namely, the proposition that equal volumes of simple gases contain the same number of atoms. We shall presently reconsider this point. We must here draw attention to the influence which the discoveries of Gay-Lussac exercised upon Berzelius in his attempt to bring the atomic hypothesis into harmony with the facts relating to the combination of gases. It is a 48 THE ATOMIC THEORY. remarkable fact that neither Dalton nor Gay-Lussac accepted the views of the Swedish chemist. The author of the atomic theory obstinately maintained his first idea of deducing atomic weights solely from the ponderable relations of elements in combinations. Gay- Lussac, again, confined himself to the immediate con- sequences of his discovery, not without forcing them to some extent, in certain cases, by hypotheses upon the forms of condensation of the combining gaseous elements. He and Berzelius expressed the composition of bodies in volumes, the latter admitting that the relative weights of these volumes represented atoms, Gay-Lussac refusing to consider these weights as anything more than ponderable ‘ relations,’ and inclining rather to the views of Davy. The latter, deviating to an equal extent from the profound conceptions of Dalton, and with the idea of completing them by the discoveries of the French chemist, confined himself strictly to established facts and to the consideration of ‘proportional numbers.’ After the ingenious but ignored attempts of Avogadro and Ampére, and the unfruitful effort of Berzelius, Dalton’s conception would have been sentenced to sterility and oblivion, had it not happened that, at the period of which we are speaking, fresh discoveries and new ideas drew attention to it. We allude to Prout’s hypothesis, to the discovery of the law of specific heats, and to the discovery of isomorphism. CHAPTER III. PROUTS HYPOTHESIS—LAW OF SPECIFIC IIEATS— ISOMORPHISM. DULONG AND PETIT—MITSCHERLICH. Ee We must first return to Prout’s hypothesis, not that it is of such great importance from our present point of view, but because it preceded the important discoveries which we shall presently mention. The anonymous author of a memoir which appeared in 18151 upon the relations between the densities of bodies in a gaseous state and the weights of atoms, tried to prove that the densities. of oxygen, nitrogen, and chlorine are integral multiples of that of hydrogen, and that the atomic weights of certain elements are similarly integral multiples of that of hydrogen. Amongst these elements we meet with some metals, the atomic weight of which had been determined by the author or by other chemists by the following excellent process: the quan- tities of metal were determined which, combined with - oxygen, formed quantities of oxides capable of neutra- 1 Annals of Philosophy, vol. vii. p. 111. 50 THE ATOMIC THEORY. lising the same quantity of an acid. The results appear in the following table :— H= 1 Ca = 20 = 6 Na=24 N= 14 Fe=28 P= 14 Zn = 32 O- 8 K =40 S = 16 Ba=70 Cl= 36 I =124 We will make no remark upon the author’s con- siderations concerning the relations which may be traced between these numbers and those whichexpressthe atomic weights of other elements determined with less accu- racy. These considerations were obscure and erroneous. The important point was raised by Prout in 1816, in a work to which he appended his name. ‘ It is very advisable,’ he remarked, ‘to adopt the same unit for specific weights and atomic weights, and to take as this unit the weight of one volume of hydrogen. The same numbers will thus give the densities of gases and the atomic weights, or a multiple of these weights. If, proceeds the author, ‘ these numbers are whole numbers, the fact under consideration may be interpreted by ad- mitting that hydrogen is the primordial matter which forms the other elements by successive condensations. The figures expressing these condensations—that is to say, the densities—would at the same time give the number of volumes of primordial matter condensed into a single volume of a given element, and the weight of this volume, expressed by a whole number, would represent the atomic weight of the element.’ PROUT’S HYPOTHESIS. 51 The determinations of atomic weights, aud even those of the densities of gases, were too inaccurate, at the time of which we are speaking, for Prout’s hypothe- sis to be taken into serious consideration. It was a conjecture. It has,as we know, been lately again taken up with great energy by Dumas.! But the accurate determinations of a number of atomic weights by Stas, notably those of chlorine, potassium, sodium, and silver, by confirming or slightly rectifying the results formerly obtained by Marignac, have entirely annihilated the celebrated hypothesis in question. Unsuccessful attempts have been made to revive it, by taking as the unit, not the atomic weight of hydrogen, but the half or the quarter of this weight. There are well-known atomic weights, particularly that of potassium, which are not a multiple of the fraction 3, nor even of 4. If, however, we retained the idea, which is, moreover, striking and profound, of a primordial matter the sub-atoms of which were grouped in different num~ bers to form the chemical atoms of hydrogen and the various simple bodies, and attributed to these sub-atoms 1 Dumas made a communication to the Académie des Sciences (January 14, 1878) relative to the atomic weight of silver, dis- cussing the error which had arisen in the determination of this atomic weight, from the property which metallic silver possesses of retaining about ;2,; of oxygen, if the latter has not been care- fully expelled by heating it in vacuo to 600° C. Dumas main- tains the number 108, which he had previously adopted. He remarks that other atomic weights, described as forming excep- tions to Prout’s hypothesis, might probably be included in the rule, if, in the process of weighing, account were taken of errors similar to those which he had pointed out in the case of silver. UNIVERSITY OF ILLINOIS LIBRARY 52 THE ATOMIC THEORY. a weight inferior to a quarter of that of hydrogen, assigning to it, for example, +, of that weight, such an hypothesis would, I say, escape all experimental verifi- cation, as the differences which it would then be our object to establish between the atomic weights of the various simple bodies would fall within the limit of errors of observation. Such an hypothesis, though rea- sonable, ceases to be legitimate, and positive chemistry for the present must abandon this theory of Prout’s, this dream of the ancients upon the unity of matter and the compound nature of chemical elements. Does this mean that Prout wrote in vain? By no means. His idea gave rise to valuable work and important dis- cussions, and the example which he set of referring atomic weights to that of hydrogen has been followed, for all chemists have adopted this unit. But at that time, as in our own, Prout’s conception produced no argument in favour of the atomic theory, and has exercised no influence upon the development of that theory. 1Ale The discovery which we are about to mention is, on the contrary, one of the foundations of the atomic theory. It draws our attention to the relation which exists be- tween the atomic weights and the specific heats of solid elements. In a memoir published in the ‘ Annales de Chimie’ in 1819, Dulong and Petit gave the specific heats of a great number of solid bodies, particularly metals, and remarked that these specific heats were generally inversely proportional to the atomic weights. DULONG AND PETIT. 53 Their observations are summed up in the following table :— | Product of the | i Relative Weight of each Atom Weights of multiplied by the | Atoms.* | corresponding Capacity. : | pester aes tthe | Bismuth .. 0°0288 13:30 | 0°3830 | Lead i - 0°0293 139624 0°3794 Gold ; : 0°0298 12°43 03704 Platinum . ‘ 00314 11°16 0°3740 Tim) |, ‘ ; 00514 7°35 0°3879 Silver : - 0°0557 6°75 0°3759 Zine. : ; 0:0927 4-03 0°3736 Tellurium. . 0°0912 4°03 0°3675 Nickel = r 0°1035 3°69 0°3819 Iron . A : 0°1100 3°392 O°3731 Cobalt. . 0°1498 2-46 03685 | Sulphur .. 0:1880 2-011 | 0°3780 This table contained several errors, which were cor- rected by V. Regnault. Thus the specific heats attri- buted to tellurium and cobalt were much too high. On the other hand, the atomic weights adopted for these elements were too low. Disregarding these inaccuracies, which were after- wards corrected, we find that the atomic weights adopted by Dulong and Petit for a large number of metals differ materially from those admitted at the same period by Berzelius. The atomic weights of zinc, iron, nickel, copper, lead, tin, gold, and telluritim were half those which Berzelius attributed to the same metals 1 Referred to that of oxygen taken as unity. To compare these numbers with those of Berzelius on p. 62, it will therefore only be necessary to multiply them by 100. 54 THE ATOMIC THEORY. in their principal oxides, which he then erroneously regarded as dioxides.' But Dulong and Petit justly remark that the ordi- nary methods of determining atomic weights by chemi- cal proportions often give a choice between several numbers. ‘There is always, they say, ‘something arbitrary in the determinations of the specific weight of elementary molecules (atomic weights); but the un- certainty only rests between two or three numbers, which always bear to each other a very simple ratio.’ In this case we should prefer the number which agrees with the law of specific heats. Moreover, the determi- nations adopted by Dulong and Petit are in accordance with the most firmly established chemical analogies. They only apply to a limited number of simple bodies. ‘Still the mere inspection of the numbers obtained points to a relation so remarkable in its simplicity as to | be at once recognised as a physical law, susceptible of being generalised and extended to all elementary sub- stances. In fact, the products in question, which express the capacities for heat of atoms of different nature, are so nearly the same for all, that we cannot but attribute these very slight differences to inevitable errors, either in the determination of capacities for heat or in the chemical analyses.’ This is quite true; the errors have been corrected, and the exceptions disap- 1 Tf. we substitute for the double atomic weights of Berzelius those which Dulong and Petit calculated from the specific heats, the following formulz of Berzelius, ZnO, FeO,, NiO,, CuO,, PbO,, SnO,, AuO,, will become ZnO, FeO, NiO, CuO, PbO, SnO, AuO. That the latter are correct Berzelius did not hesitate to ac- knowledge. LAW OF SPECIFIC HEATS. 55 peared one by one, But this result was only obtained upon the completion in some points of the great re- searches commenced in 1840 by V. Regnault upon specific heats, which has only very recently been accom- plished. A reform in the system of atomic weights was also necessary, a reform which has taken place slowly and by degrees. Dulong and Petit recognised the importance of their discovery and did not exaggerate it. They brought to light a great law of nature, which they expressed in the following striking form: ‘The atoms of all simple bodies have precisely the same capacity for heat.’' This simple statement was of the greatest value to the idea of atoms, which until then rested upon purely chemical considerations, for we here meet with a physical relation between atoms, to which another physical relation was soon added—that, namely, existing between the density of gaseous bodies and the weight of their ultimate particles. In both cases the true formula had been wanting. Dulong and Petit readily discovered it as far as concerns the specific heats of atoms ; Berze- lius had been less fortunate in his attempt to define the volumes occupied by the latter in gases. III. Those were times of great activity, and the fertile discovery which we have just mentioned was soon fol- lowed by another, which exercised a manifest influence 1 Loe. cit., p. 405. 56 THE ATOMIC THEORY. upon the development of chemical theories. In the month of December 1819 E. Mitscherlich made known the law of isomorphism. The experiments which he was making upon phosphates and arsenates led to the discovery. He first established the fact that these salts resemble each other in composition, if this composition: is represented in ‘ proportions,’ phosphoric acid consisting of one proportion of phosphorus and five proportions of oxygen, and arsenic acid of one proportion of arsenic and five proportions of oxygen. This being granted, he observed, further, that the phosphates and arsenates of the same bases, combined with the same quantity of water of crystallisation, possess the same crystalline form. There is, therefore, a correlation between analogy of composition and identity of crystalline form, and it is this which constitutes the discovery. After having esta- blished this correlation for salts of the same base formed by two different acids, Mitscherlich observed it again in analogous salts formed by the same acids united with different bases. Thus potash and ammonia on the one hand, and baryta, strontia, and oxide of lead on the other, form, with the same acids, salts analogous in composition and identical in crystalline form. The same identity of forms is found in the carbonates of lime, iron, zinc, and manganese. Mitscherlich was continually adding to his first examples. He showed the identity of the forms of certain sulphates of the magnesian group, which crystallise with the same quantity of water, such as the orthorhombic sulphates of magnesia, zinc, and nickel, and the clinorhombic sulphates of iron and cobalt; and the insignificant errors MITSCHERLICH—ISOMORPHISM. 57 which crept into these early experiments had at least the advantage of strengthening his conviction. Though he was wrong in affirming that the different forms of sulphate of iron and sulphate of zinc are due to a dif- ference in the quantities of water of crystallisation, he shows that in magnetite and gahnite, which belong to the group of the spinels, the ferrous and zincic oxides form isomorphous combinations with ferric oxide.. He pre- pared iron alum, and showed its isomorphism with ordinary alum, &c. The first definition of the law of isomorphism—that acids of analogous composition and bases of analogous composition, the former with the same base, the latter with the same acid, form salts of identical crystalline form—was not absolutely correct. Mitscherlich himself recognised this subsequently. His discovery of di- morphism revealed the fact that the same substance can erystallise in two different forms. It must, therefore, be also possible for two ‘substances of different nature, but of analogous composition, to crystallise in two dif- ferent forms. . Facts of this nature, which are exceptions to the law, may be regarded as cases of dimorphism. Mit- scherlich also observed that substances which are really isomorphous, which combine and replace each other in the same crystal in every proportion, do not always present a perfect identity of form in isolated crystals, the number and value of the faces and angles being — liable to slight variations, though the general form of the crystal remains unchanged. Such is the discovery of Mitscherlich, and we must 58 THE ATOMIC THEORY. now describe the influence which it has exercised upon the development of the atomic theory. Mitscherlich himself admitted, in his first memoir, that the similarity of properties in compounds of analogous composition and identical form could scarcely be attributed to identity of crystallisation, but that the explanation must be sought for in a primary and mysterious cause, to which must be referred, on the one hand, the fact of combination by fixed ‘ volumes’ (or atoms), and on the other the resemblance of crystals. This primary cause is the atomic structure of the bodies. A similar atomic constitution not only deter- | mines the analogy of. chemical properties, but also the similarity of physical forms. Mitscherlich thus declares himself a supporter of the ideas of Berzelius upon the constitution of bodies. Following the example of the latter, he expresses it first in ‘ volumes’ and afterwards in atoms.. The memoir published in the ‘ Annales de Chimie et de Physique’ of 1821 bears the following significant title: Upon the identity of erys- talline form wm various substances, and the relation between this form and the number of elementary atoms in the crystals. We have already mentioned the restrictions which Mitscherlich was obliged to place upon his original idea. In a work presented to the Stockholm Academy in 1821 he endeavours to give them precision. He there proposes the following questions :—Do compounds formed by different elements with the same number of atoms of one or several other elements possess the same crystalline form? Is identity of crystalline form determined by the nwmber of atoms ISOMORPHISM. 59 alone, and is this form independent of the chemical nature of the elements? The replies to these questions are not as absolute as the principle. For the form to remain unchanged in analogous compounds, the elements which replace each other must be mutually isomorphous, as phos- phorus and arsenic, or barium, strontium, and lead. Having further observed that certain salts, such as the acid phosphate of soda, can crystallise in different forms, although in the two cases the composition is identical, he attributes this dimorphism to a different arrangement of atoms. Thus the idea that chemical com- pounds are formed of atoms, and that the number and arrangement of these atoms exercise an influence upon the physical form of crystals, had evidently made an impression upon him; and the idea is natural, although it is founded upon a comparison some terms of which are wanting. A crystal may be compared to an edifice of definite form. We see its production, growth, and modification. Is it not natural to suppose that this form is due to the accumulation and arrangement of materials, which we call atoms? We are doubtless using figurative language when we compare the mole- cular edifice, and its construction from these invisible materials, to a monument of human architecture, which rises piece by piece before oureyes. But the necessities of the case are so perfectly answered by this repre- sentation that it has passed at once into the language of our general explanations and demonstrations. However this may be, the atomic theory evidently exercised an influence upon the conception of Mitscher- 60 THE ATOMIC THEORY. lich and upon the manner in which. he stated his discovery. The same number of elementary atoms, he said, combined in the same manner, produce the same erystalline form, and this form is independent of the chemical nature of the atoms and determined solely by their number and arrangement. In spite of necessary restrictions and established exceptions, so great a law could not but act as a solid support to the atomic hypothesis, which had contributed such precise and simple terms for the enunciation of that law. Hays But this is not all. Mitscherlich’s discovery was the cause of important changes introduced by Berzelius into the system of atomic weights which he had esta- blished in 1813, and into the notation of which they are the origin. He had previously fixed the atomic weight of chromium and iron by attributing to chromic acid the composition CrO, and to ferric oxide the composition FeO, He now halves the atomic weight of chromium, attributing to chromic acid the formula CrO,, which makes it agree with anhydrous sulphuric acid, SO,. Chromium oxide now assumes the composition Cr,O,, and, on account of the isomorphism recognised between chromium oxide and ferric oxide (chrome alum and iron alum), the latter oxide becomes Fe,OQ, and ferrous oxide FeO. Thus was Berzelius forced to abandon an opinion which he had long entertained, one which Dalton, moreover, had never admitted—namely, that a BERZELIUS'’S SYSTEM OF ATOMIC WEIGHTS. 61 binary compound (that is to say, a combination of two elements) must always contain a single atom of one or the other element. The existence of sesquioxides, R,O,, was finally admitted. But the halving of the atomic weights of chromium and iron occasioned other changes. The chemical analogies and isomorphism of ferrous oxide, FeO, with lime, magnesia, and oxide of zinc made it necessary to attribute to these oxides, and to the strong bases in general, the composition of protoxides, RO, and consequently to halve the atomic weights of a great number of metals, as indeed had already been done by Wollaston, Dulong and Petit (see p. 53). The old formule of the sulphates of iron and zinc— FeO,, 280, + 14H,0; ZnO,, 280, + 14H,O— became, therefore— - FeO, $0,+7H,0; ZnO, SO, +7H,0. These lower atomic weights agreed, moreover, with the law of specific heats. Berzelius draws atten- tion to this fact, and in future, in the determination of atomic weights, follows three principles which mutually support each other. 1. The law of volumes. He steadily maintains the propositions which he had previously stated—namely, that equal volumes of simple gases contain an equal number of atoms. This proposition was soon to be refuted by the experiments of Dumas and Mitscherlich. 2. The law of Dulong and Petit. This law is sub- ject, it is true, to some exceptions, but is of great as- sistance in certain cases, where it enables us to control 62 THE ATOMIC THEORY. other determinations. The experiments of Regnault diminished the number of these exceptions, but it is only very recent investigations which have caused their final disappearance. 3. The law of isomorphism. We have seen in the preceding pages the assistance which Berzelius obtained from this law in the determination of atomic weights. We here give the list of atomic weights as given by the great Swedish chemist in 1826, and repeated by hin without alteration in 1835. Atomic Weights Symbols | referred to Atomic Weights referred to Hy- Oxygen as 100 drogen as 1 {a ae ee gen ede ho 100 | 16-02 Hydrogen ; H 6°2398 1 Carbon . : | C 76°44 12-26 Boron ; ; B 136°2 21°82 Phosphorus . i). eee 196°14 31°44 Sulphur : 8 201°17 32°24 Selenium ; Se 494°58 79°26 Iodine . : I 789°75 126°56 Bromine ‘ Br 489°75 78°40 Chlorine ‘ Cl 221°33 35°48 Fluorine : F 116°9 18°74 Nitrogen : N 88°52 14:18 Potassium , K 489-92 78°52 Sodium : Na 290:90 46°62 Lithium F L 80°33 12°88 Barium : Ba 856°88 137°32 Strontium ; Sr 547°29 87:70 Calcium ‘ ; Ca 256°02 41:04 Magnesium . : Mg 158°35 25°38 Yttrium : ¥ 402°51 64°50 Glucinum ; Gl 331°26 53°08 Aluminium . : Al 1 hp ES i, 27°44 Thorium : Th 744-90 119°30 Zirconium ; Zr 420-20 | © “6734 Silicon . P Si 277°31 44°44 Titanium * Ti 303°66 48°66 Tantalum ; Ta 1153-72 18490 Tungsten pAaseaal 118300 — | 189-60 BERZELIUS’S SYSTEM OF ATOMIC .WEIGHTS. 63 Atomic Weights Atomic Weights Symbols referred to referred to Hy- Oxygen as 100 drogen as 1 Molybdenum 598°52 95°92 Vanadium 856'89 137°32 Chromium 351°82 56°38 Uranium 2711°36 434°52 Manganese 345°89 55°44 Arsenic 47004 75°34 Antimony 806°45 129°24 Tellurium 801°76 128°50 Bismuth 886°92 142°14 Zine 403-23 64°62 Cadmium 696°77 111°66 Tin 735°29 117°84 Lead . 1294:50 297°46 Cobalt . 368°99 59°14 Nickel . 369°68 59°24 Copper . 395°71 63°42 Mercury 1265°82 202°86 Silver 1351-61 216°60 Gold 1243-01 199-20 Platinum 1233°50 197°70 Palladium 665°90 LOG TZ Rhodium ., : 651°39 104:40 Iridium 1233°50 197°68 Osmium 1244°49 198-44 The atomic weights of Berzelius are referred to oxygen as 100. Dividing the numbers which express the atomic weights by 6°2398, the atomic weight of hydro- gen, we obtain the numbers given in the second column, which are referred to hydrogen taken as unity. The comparison of these numbers with those adopted at the present day, which will be presently in leads to two important remarks. In the first place, the system of atomic weights which met with the approval of Berzelius is very similar to that which is adopted at the present day. With the exception of a few modifications which have been added 64 . THE ATOMIC THEORY. to it,! and which do not affect the ruling principles and general features of the whole, we only discover one important difference between the two systems. This difference arises from the atomic weights of the alkaline metals, and of silver, which are twice as great as those which chemical analogies and the law of Dulong and Petit have now forced us to adopt. We shall presently return to this point. Secondly, in examining these numbers of Berzelius, we are struck by their accuracy. Most of these numbers only differ in the decimals from those which we now adopt as true. Such is the result of the enormous amount of labour expended by the Swedish chemist upon the determination of the atomic weights. It is a lasting monument which he has raised to science and his own glory. Nevertheless Berzelius never succeeded in persuad- ing all chemists to adopt his system of atomic weights. Dissentient voices are always to be heard. Gay-Lussac and Wollaston, following the example of Dalton and Thomson, adhered to the atomic weights derived solely from the consideration of the equivalent quantities which enter into combination. Gmelin adopted the same ideas in the several editions of his great work, and contributed greatly in the course of time towards the introduction of the equivalent notation. Berzelius made a concession upon one point to all these opponents. He introduced the idea of double atoms and applied it to certain gases, such as hydro- gen, nitrogen, chlorine, bromine, and iodine, the atomic weights of which were only half those admitted by 1 Among others the atomic weights of uranium, silicon, &c. BERZELIUS’S SYSTEM OF ATOMIC WEIGHTS. 65 other chemists. These double atoms were supposed to enter into combination in pairs, and every pair repre- sented precisely what others termed ‘the proportion’ or ‘equivalent.’ Water was therefore composed of a double atom of hydrogen united to one atom of oxygen, and this combination was represented by the symbol HO. Hydrochloric acid and ammonia were formed, the first of one double atom of hydrogen united to one double atom of chlorine, the second of one double of nitrogen united to three double atoms of hydrogen. The formule HO, HE€l, HN, were really equivalent to the formule H,O, H,Cl,, H,N,, but remind us of the notation HO, HCl, H,;N, employed by Gmelin and others. It was, in fact,astep backwards. In admitting double atoms Berzelius unnecessarily doubled a number of formule ; and if itis true that H,O, H,Cl,, represent, from acertain point of view, equivalent quantities of water and hydrochloric acid, it is equally true that these formulz do not represent true molecular magni- tudes. Gerhardt subsequently showed that if a molecule of water, occupying two volumes of vapour, is repre-~ sented by the formula HO, a molecule of hydrochloric acid occupying two volumes of vapour should be repre- sented by the formula HCl, and a molecule of ammonia by H,N. It is true that the formula H,Cl, corresponds to the formula PbCl,, ZnCl,, CaCl,, and KCl,, by which Berzelius represented the chlorides of lead, zine, calcium, and potassium. But we now know that the molecules of all these chlorides are not, strictly speaking, equiva- lent, and that if the three first are true the fourth must be halved. The law of specific heats forces us, in fact, 4 66 THE ATOMIC THEORY. to halve the atomic weight of potassium, and conse- quently to represent its chloride by the formula KCl, which answers to HCl. The latter formula represents two volumes of vapour, as do the formule of water, H,O, and of ammonia, H,N. All these inaccuracies which we have pointed out in Berzelius’s system of atomic weights and notations arose from an erroneous conception of the law of volumes. Instead of regarding as equidistant, and equally distri- buted in equal volumes of gases or vapours, the particles of the second order, or the molecules of simple and com- pound bodies, as Avogadro and Ampére had done, and later Gerhardt, Berzelius only considered the primordial atoms of certain simple gases, holding that they alone, and not the ‘compound atoms,’ are’ distributed in equal numbers in equal volumes. We know now that this is an erroneous idea, and that the hypothesis of Avogadro and Ampére, long forgotten, but restored to its due place of honour by Gerhardt, applies to the single molecules or particles of the second order, which, whether simple or compound, constitute the ponder- able matter of gases and vapours. CHAPTER IV. SYSTEM OF CHEMICAL EQUIVALENTS—EQUIVALENT NOTATION. iB Tue interpretation which Berzelius had given of the law of volumes formed, as we have seen in the pre- ceding pages, one of the foundations of his system of atomic weights and of his notation. This foundation was destroyed by the researches of Dumas, and subse- quently of Mitscherlich, upon vapour densities com- menced in 1827. Dumas noticed that the vapour density of mercury is sensibly equal to 100, hydrogen being taken as unity. The vapour densities of mer- cury and of oxygen are as 100:16 or as 50:8. -If the atomic weights were proportional to the densities, 8 of oxygen should combine with 50 of mercury to form mercuric oxide. This is not the case; mercuric oxide is composed of 8 of oxygen and 100 of mercury, and it is the latter number which Berzelius had adopted for the atomic weight of mercury. If equal volumes of oxygen and of mercury vapour contain the same num- ber of atoms, their densities should be in the ratio of 8 to 100, or, in other words, the density of mercury vapour 68 THE ATOMIC THEORY. is only half what it should be. We have here evidently a well-marked exception, or, better, a manifest contra- diction between the facts and. the principle admitted by Berzelius. Other exceptions may be mentioned. The vapour densities of sulphur and phosphorus deter- mined by Dumas in 1832 were found to be, in the first case, three times as great, and in the second twice as great, as those indicated by theory. Chemical con- siderations have caused a composition, expressed by the formule H,S and SQ,, to be attributed to sulphuretted hydrogen and sulphuric anhydride. From these formulz the ratio between the atomic weights of sulphur, oxygen, and hydrogen is expressed by the numbers 32:1 :16, and the densities should be in the same ratio. Now, the vapour density of sulphur taken at about 560° is 96, hydrogen being taken as unity. From this density a quantity weighing 32 in the mole- cule of sulphuretted hydrogen would not represent an atom of sulphur, but 3 of an atom, and the formula of sulphuretted hydrogen, expressed in conformity with the law of volumes, would be H,S3, which is inadmis- sible. From the formulz PH, and P,O,, adopted for phos- phoretted hydrogen and phosphoric anhydride respec- tively, the relation between the atomic weights of phosphorus, hydrogen, and oxygen should be expressed _ by the numbers 31:1:16. Now, the vapour density of phosphorus is equal to 2x 31=62. If, therefore, the density of sulphur vapour is three times greater than that indicated by theory, that of phosphorus is twice as great. The case is the same with that of arsenic, from BERZELIUS'S PRINCIPLE OF NOTATION. 69 an experiment of Mitscherlich, who also confirmed, in 1833, the results obtained by Dumas upon the vapour of mercury, sulphur, and phosphorus. We here, therefore, meet with a serious difficulty. For its solution two courses are open to us: we must either maintain the principle of the equality of the number of atoms in equal volumes of gases or vapours, and determine to assign to mercury, sulphur, phosphorus, and arsenic atomic weights which shall conform to the vapour densities, although they are less probable, and consequently to give their compounds the formule Hg,O, H,S3, P5H, ; or else it will become necessary to sacri- fice the principle under discussion, in order to enable us to adopt the atomic weights, HgO, indicated by chemical analogies and the law of specific heats. The atomic weights of mercury, sulphur, phosphorus, and arsenic being, therefore, 200, 32, 31, 75, referred to hydrogen as unity, the preceding formulze become HgO, H,S, PH, and AsH,. It is the latter course which chemists have adopted, since they were properly unwilling to neglect more evident analogies. But the adoption of these atomic weights involves the following consequences :— 1. The vapour of mercury, the density of which is only half that required by the atomic weight assigned to mercury, evidently contains half the number of atoms contained in an equal volume of hydrogen. 2. The vapour of sulphur, which at 500° is three times as dense as it should be from the atomic weight assigned to sulphur, contains, at this temperature, three times the number of atoms contained in an equal volume of hydrogen. 70 THE ATOMIC THEORY. 3. The vapours of phosphorus and arsenic, which are twice as dense as they should be from their atomic weights, evidently contain twice as many atoms as an equal volume of hydrogen. The atomic constitution of gases or of elementary vapours is not, therefore, always the same, as Berzelius for a long time supposed. If we compare gases or ele- mentary vapours, as far as concerns the number of atoms which they contain, to the vapour of mercury, which contains the least, we shall have the result that, if mercury vapour contains in a certain volume one atom, hydrogen, oxygen, nitrogen, chlorine, bromine, and iodine contain 2, phosphorus and arsenic contain 4, while sulphur at 500° contains 6. The relations be- tween the number of atoms contained in equal volumes © of gases or of vapours may be obtained by dividing the density of the gas or vapour by the corresponding atomic weight. We shall thus obtain the following results :-— Densities divided by Atomic Atomic Weights. Number of : Atcms in Weights Number of ; Atoms in 2 Volumes Unit of Volume Densities referred to Hydrogen Or Mercury Hydrogen Oxygen Nitrogen Chlorine Bromine Iodine . Phosphorus . Arsenic , Sulphur at 500? NGI Nl ll ell ll ll oll ek) SH EH DO ND DOD ww Ne roe) BERZELIUS'S PRINCIPLE OF NOTATION. 71 We have, therefore, to distinguish monatomic, diatomic, tetratomic, and hexatomic gases. Gmelin has already introduced into science a similar distinction, which has now become so important. At p. 54 of the first volume of the fourth edition of his treatise he gives a table analogous to the preceding, with some dif- ferences due to the different. atomic weights adopted. Those used in our table are those of Berzelius (p. 62), which are now adopted for the respective elements. With Gmelin and other chemists who soon followed his example it was different. As we have already remarked, the former maintained the proportional numbers which he designated in the first editions of his classical treatise by the erroneous term of * Mischungsge- wichte,’' and which he referred to hydrogen as unity, following the example of Dalton. In the fourth edition of his work he returns to the term atomic weights, but the numbers thus designated were identical with the proportional numbers or equivalents. iN The system of chemical equivalents and the notation derived from them gradually prevailed over the system of atomic weights and the notation of Berzelius, and are still preferred by some French chemists. It will therefore be useful to explain the principles upon which this equivalent notation rests, and particularly the arguments used by Gmelin against Berzelius in the question which forms the chief point of the discussion— 1 Literally ‘mixing weights,’ instead of ‘combining weights. fe THE ATOMIC THEORY. viz. the atomic weights of hydrogen, nitrogen, phos- phorus, arsenic, chlorine, bromine, and iodine, which are half the proportional numbers or equivalents. 1. The atomic weights are deduced from the densi- ties of the gases, and are founded upon the hypothesis that these gases contain an equal number of atoms in equal volumes. Now, this hypothesis is contradicted by experiment, as far as concerns the vapours of sulphur, phosphorus, arsenic, and mercury. Gmelin also re- marked that certain gases — hydrochloric acid, for example—contain in equal volumes only half as many atoms as chlorine and hydrogen.! There is, therefore, no reason for the adoption of the halved atomic weights of Berzelius, and his double atoms are the true atoms—that is to say, the equiva- lents. 2. The small atoms of which he speaks never enter singly into any combination. Neither do they ever enter into combination in uneven numbers, such as 3, 5, 7, &c., but always in even numbers, such as 2, 4, 6. Thus water contains H,O, and hydrochloric acid H,Cl,, and ammonia H,N,. The atomic weights of hydrogen, chlorine, and nitrogen ought, therefore, to be doubled, so as to give the above compounds the simple formule HO, HCl, and H.N. 3. We should admit that heterogeneous atoms unite in the simplest proportions, and the value of the atomic weights ought to be doubled-in order to repre~ sent these simple relations. ‘Thus, when a metal only combines with oxygen in a single proportion, it should 1 HCl was then called an atom of hydrochloric acid. THE OBJECTIONS BROUGHT FORWARD BY GMELIN. 73 be supposed that the combination takes place atom with atom, unless isomorphism points to the contrary; and when a metal forms several combinations with oxygen, the strongest base ought to be supposed to be a com- bination of one atom of metal with one atom of oxy- gen. 4. The sum of all the atomic weights of an acid should represent the weight of the acid which saturates a quantity of base containing one atom of oxygen. Thus 40 parts of sulphuric acid saturate 111°8 parts of oxide of lead, which contain 8 of oxygen (one equiva- lent) and 103°8 of lead (one equivalent). These 40 parts consequently represent the sum of the equivalents of oxygen (24=3 x 8) and of sulphur (16); 16 is there- fore the equivalent of sulphur, the formula of sulphuric acid being SO,. The question of equivalence is here very clearly put, almost in the terms used by Dumas in 1828 in the first volume of his celebrated treatise on chemistry applied to the arts. 5. The same formule ought to be assigned to iso- morphous compounds as well as to compounds of the same order formed by simple and similar bodies, such as cobalt and nickel. We will briefly examine into the value of these argu- ments,’ reserving the development of some of the points here mentioned for the following chapter. 1, Gmelin justly observed that there are exceptions to the principle laid down by Berzelius of the equality of the number of atoms in equal volumes of a gas or a No objection can be made to the last principle (No. 5), which is followed by all notations. 74 THE ATOMIC THEORY. vapour, and that consequently the weights of equal volumes do not always represent the atomic weights. Thisisadmitted. But when he remarks that one volume of hydrochloric acid gas only contains half the number of atoms contained in one volume of chlorine or of hydrogen, he evidently confuses atoms with molecules. It is now admitted that equal volumes of these gases contain the same number of molecules, and we may remark that, in the present case, they also contain the same number of atoms, as is shown by the following formule :— H,=2 vol. Cl, =2 vol. HCl=2 vol. 1 molecule of 1 molecule of 1 molecule of hydro- hydrogen. chlorine. chloric acid. 2. It is not correct to say that the small atoms of Berzelius do not enter into combination in uneven numbers. Ifa molecule of water is represented by the formula H,O=2 vol., a molecule of hydrochloric acid is represented by the formula HC]l=2 vol., and a molecule of ammonia by the formula NH,. The double formule of Berzelius, H,Cl, and N,H,, did not represent the true molecular masses; they were double the true number and ought to be halved, as was first proposed by Gerhardt. The small atoms of Berzelius, H=6-24, Cl= 221°3, N= 88:5, therefore represent the true atomic weights of these elements, referred to oxygen as 100, and the ratio between these numbers is identical with the ratio between the numbers 1 : 35:5 : 16, which are now accepted as the atomic weights of these ele- ments. 3. The statement is correct that heterogeneous atoms generally unite together in very simple propor- INCONSISTENCIES IN THE EQUIVALENT NOTATION. 75 tions. This fact becomes evident if we allow ourselves to be guided in determining atomic weights and in constructing formule, not only by chemical considera- tions, but also by the great physical laws which have been described—namely, the law of volumes, of specific heats, and of isomorphism. Purely chemical considera- tions might lead us into error. Thus it is not correct to say that the strong bases ought always ‘to contain one atom of metal and one atom of oxygen. Lime, baryta, strontia, cupric oxide, mercuric oxide, &c., con- tain, it is true, 1 atom of metal and 1 atom of oxygen; but oxide of silver, which is a strong base, contains 2 atoms of metal for 1 of oxygen, the atomic weight of silver being determined by the law of specific heats. As far as concerns oxide of silver, therefore, we make a mistake if we invoke analogy in order to connect it with the preceding oxides in respect to its atomic constitution. 4. The principle of equivalence made use of by Dalton, Wollaston, Gay-Lussac, and Gmelin for the determination of equivalents (which Dalton and Gmelin called atomic weights) would be admirable if it could be applied rigorously either to elements or to com- pounds. But we now know that all atoms are not equivalent, and that the case is the same with molecules and with the reactions to which they give rise. ; Atoms differ in their combining or substituting value —in their valency, as it is called—molecules in their state of condensation and their degree of saturation, and reactions in the greater or less extent of their com- 76 THE ATOMIC THEORY. plexity. As we have remarked above concerning oxides, it is impossible to cast all this in the same mould. To return to the exact point of the discussion, it is impossible to consider a molecule of nitric acid ‘and of phosphoric acid as equivalent; and if, in conformity with the rule laid down by Gmelin, 14 is the equivalent of nitrogen because nitrate of silver contains 14 parts of nitrogen for 108 of silver, 10°5 should be the equiva- lent of phosphorus, for it is the weight of phosphorus contained in a quantity of phosphate of silver contain- ing 108 parts of silver. Now, all chemists admit that the equivalent of phosphorus is 31:4; but then we must no longer consider a molecule of nitric acid as equivalent to a molecule of phosphorie acid, for if the former saturates a quantity of oxide of silver containing 1 atom of silver, the latter saturates a quantity of oxide of silver containing 3 atoms. In fact, the discovery of polybasic acids proved a serious difficulty to the theory of equivalence ; it showed that chemical molecules are not equivalent, as was shown for atoms by the law of volumes. Moreover, Gmelin felt that he had met with a difficulty, for he mentions polybasic acids as forming an exception to the theory of equivalence. It is some- times said—I do not know for what reason—-that ‘exceptions prove the rule; in the present case they have become so numerous and so striking that they have overthrown it. The discovery of polybasic acids has, in fact, been supplemented by other discoveries, and they have completely modified the old ideas upon the equivalence of molecules and of reactions. But INCONSISTENCIES IN THE EQUIVALENT NOTATION. 77 this is not the proper place to develope this point, and we will merely add a remark which seems important. Dalton and Gay-Lussac alone applied true principles to the determination of equivalents. Dalton attributed to phosphorus the atomic weight 10°3; it represents the quantity of phosphorus which combines with 1 part of hydrogen: to carbon the atomic weight 4°3 (instead of 6); it represents the quantity of carbon which unites with 1 of hydrogen to form bicarburetted hydrogen. Gay-Lussac started from another point of view. Con- sidering ordinary phosphate of soda as neutral, he admitted in this salt the presence of one equivalent of base and consequently one equivalent of sodium. He therefore expressed its composition by the formula PO,3.Na0 + Aq,'! and attributed to phosphorus the proportional number 15:7. The quantity of neutral phosphate of soda which is proportional or equivalent to a molecule of nitrate of soda, NO,.NaO, or of silver, NO,.AgO, ought, in fact, only to contain 1 atom of metal, like the latter. Applying the same principles in other cases, he wrote ferrous oxide FeO and ferric oxide Fe20. Ferrous sulphate, SO,.FeO, was strictly equivalent to ferric sulphate, SO,.Feg0. Berzelius, on the contrary, who had at last decided 1 P=15'7; O=8. At this time no account was taken of basic water. Gay-Lussac therefore involuntarily.committed an error in the determination of the equivalent of phosphoric acid. In fact, the quantities of phosphate of soda and of nitrate of silver which enter into reaction, and which are strictly equivalent, are 4(PO,Na,H) and NO,Ag, and the quantity of phosphorus in 4(PO,Na,H) is 10°5. This is the number of Dalton. 78 THE ATOMIC THEORY. to admit the existence of sesquioxides, proved that they unite with 3 atoms (molecules) of acid. He con- sequently represented ferrous and ferric sulphates by the formule SO,.FeO and 3S0,.Fe,0,. Is it not evident that he was less consistent than Gay-Lussac, and that these formule do not represent equivalent quantities? It is only a strange abuse of language, not to say a logical error, to consider as equivalent a mole- cule of ferric oxide, which saturates 3 molecules of sulphuric acid, and a molecule of ferrous oxide, which only saturates 1 molecule. Formule analogous to those of the sulphates of the sesquioxides, such as those of the phosphates and of several other compounds, which are now distinguished by the name polyatomic, reveal, there- fore, serious inconsistencies in the equivalent notation, and we must choose between such inconsistencies and the graver inconvenience of misrepresenting reactions by referring them to strictly equivalent proportions. This point will be developed in the following chapter. The preceding discussion renders it sufficiently evident that the system of chemical equivalents, and of the notation derived from them, introduced by Dalton, Wollaston, Davy, Gay-Lussac, and Gmelin, were based upon too narrow a foundation for the enlarged edifice of chemistry. Our present system of atomic weights and our notation rest upon a wider foundation. Their establishment has required the numerous efforts which have been perseveringly maintained for a period of thirty years. CHAPTER V. PRESENT SYSTEM OF ATOMIC WEIGHTS. GERHARDT AND LAURENT—CANNIZZARO, ye Tue equivalent notation of the English chemists and of Gay-Lussac, which was adopted by Liebig and defended by Gmelin in 1843, had, at the period of which we are speaking, gained the almost unanimous approval of chemists; they were struck with the excep- tions presented by the law of volumes as it was then interpreted, by the useless complication which the con- ception of the double atoms of Berzelius had introduced into a large number of formulz, and they were satisfied with the more simple expressions which the notion of equivalents offered for chemical reactions and com- binations. The law of volumes was entirely sacrificed. The equivalents of hydrogen, nitrogen, chlorine, &c., corresponded to two volumes, whilst that of oxygen only constituted one. The formule of water, HO, of sulphuretted hydrogen, HS, of protoxide of nitrogen, 80 THE ATOMIC THEORY. NO, expressed two volumes; those of hydrochloric acid, HCl, of ammonia, NH,, of phosphoretted hydrogen, PH,, &c., represented four. Gerhardt was the first to draw attention to these errors, and to the necessity of considering as equivalents quantities of water, ammonia, hydrochloric acid, &c., corresponding to equal volumes. Regarding water, H,O, as formed of two atoms or volumes of hydrogen and as occupying 2 volumes, if one atom of hydrogen occupies one volume, he compares it to hydrochloric acid, HCl, formed of one atom or volume of hydrogen and of one atom or volume of chlorine, and occupying 2 volumes; to ammonia, NH,, formed of one atom (volume) of nitrogen and of 3 atoms (volumes) of hydrogen, and occupying 2 volumes. In the same manner the for- mule N,O, NO, CO, CO,, CH,, C,H,, which correspond to 2 volumes, represent molecules (Gerhardt still used the term equivalents) of protoxide of nitrogen, dioxide of nitrogen, carbonic oxide, carbonic acid, and of marsh gas and olefiant gas. The atomic weights on which the preceding formulze are founded are the same as those of ‘Berzelius, Le. O=100; H=6°25, N=88," C= 75, But the formule of hydrochloric acid, H,Cl,, of ammonia, N,H,, of marsh gas, C,H,, of olefiant gas, C,H,, which Berzelius had employed were halved and made to represent 2 volumes. Here lies the true progress. It will be interesting to recall the considerations which led Gerhardt to propose ue reform in the nota- tion of Berzelius. for Regarding a molecule of water as formed of 2 atoms of hydrogen and 1 atom of oxygen, and carbonic acid as GERHARDT’S NOTATION. 81 containing 1 atom of carbon and 2 atoms of oxygen, he was struck, in the attentive study of the reactions of organic chemistry, by the fact that in none of these reactions, represented by the formulz and equations of Berzelius then in use, were quantities of water and car- bonie acid corresponding to H,O and CO, set free, but that the quantities formed were never less than those corresponding to the formulze H,O, and C,0,. We may therefore conclude, he says, that an error has been committed in the construction of organic formule, for it would be strange if no redaction should give rise to the formation of a single molecule of water or a single molecule of carbonic acid. This is the error: organic formulze are twice as great as they should be, and must be halved, as well as the atomic weights of metals. These two facts are correlative, and it was precisely those high atomic weights attributed by Ber- zelius to the metals which gave to organic compounds formule double what they should be. Thus amongst the organic combinations with which we are most fully acquainted we must reckon the acids ; their mole- cular magnitude is determined by their capacity of saturation, and we admit that a molecule of acid saturates a molecule of basic oxide—that is to say, a quantity of base containing one atom of metal. Thus, for example, the formula of acetic acid is constructed by combining it with oxide of silver and analysing the acetate of silver. The composition of this salt, contain- ing one atom of silver, is represented by the formula C,H,AgO,, derived from the atomic weights C=75, H=6°25, O=100, Ag=1351°6, which are those of Berze- 82 THE ATOMIC THEORY. lius. But upon halving the atomic weight of silver we obtain Ag=675°'8; the preceding formula will become C,H,Ag,O,; and there is no reason why we should not halve this, for we must admit that the monobasic acetic acid only contains in its salts a single atom of metal. The true formula of acetate of silver and acetic acid are therefore C,H,AgO, and C,H,0,. But why must we halve the atomic weights of metals in this manner? In order that their oxides may be comparable to water. If the latter is formed of 2 atoms of hydrogen, we may reasonably attribute to protoxides a similar composition, and represent them by the formula M,O instead of MO. Oxide of potassium and oxide of silver being, therefore, K,O and Ag,O, the atomic weights of potassium and silver will be 245 and 687°5 '—that is to say, the half of those attributed to them by Berzelius, the atomic weights of hydrogen and oxygen being 6°25 and 100. Applying the same considerations to the other protoxides, Gerhardt also halved the atomic weights of the metals which they contain. We shall presently see that in this he went too far; but this reasoning was perfectly correct as far as it concerned acetate of silver, and nothing could be more legitimate than the halving of the formula of acetic acid, the unnecessary complication of which he was the first to show. And this change demanded others. It is clear that the several monobasic acids, the alcohols, ethers, amides, &c., must be represented by formule which harmonise with that of acetic acid. 1 The number 687°5 is deduced from a determination of Erdmann and Marchand (Précis de Chimie organique, t. i. p. 54). GERHARDT’S NOTATION. 83 This led to an important reform in the notation of organic compounds, which reform extended even to in- organic compounds. JBerzelius had represented hydro- chloric acid by the formula H,Cl,, because 100 being the atomic weight of oxygen, this quantity of hydro- chloric acid was necessary to saturate a molecule of oxide of silver containing 1351°6 of silver and 100 of oxygen. The formula H,Cl, is therefore in harmony with the formule KCl,, AgCl,, PbCl,, which represent the composition of the protochlorides. But when the atomic weights of the metals were halved, it was con- sidered advisable to attribute to all these chlorides the more simple formule HCl, KCl, AgCl, PbCl. The reform which Gerhardt introduced into notation necessitated certain modifications in the existing ideas concerning the constitution of salts. It can now no longer be said that a molecule of acetate of silver contains a molecule of anhydrous acetic acid and a molecule of oxide of silver, or that hydrated acetic acid contains a molecule of anhydrous acid and a molecule of water. The double formule favoured these interpretations, while the simple formule cannot be divided in such a manner. Though C,H,Ag,O, might be decomposed into C,H,O,+Ag.0, and C,H,O, into C,H,O,+H,0, the formule C,H,AgO, and C,H,O, could not be divided so as to give anhydrous acid and oxide of silver, or anhy- drous acid and water. Yet the relations between acetic acid and acetate of silver are very simple, and correctly defined when we say that acetate of silver consists of acetic acid in which one atom of hydrogen is replaced by an atom of silver. The molecule of acetic acid may, 84 THE ATOMIC THEORY. then, be regarded as a group of atoms in which an atom of hydrogen, which is called basic, can be replaced by an. atom of metal, in the same manner as, in a different class of facts, the three other atoms of hydrogen may be replaced by three atoms of chlorine. This is an important consequence of Gerhardt’s notation, to which I thought it well to draw attention in passing, for these ideas upon the nature of salts were opposed by their great author to the dualistic theory, and are in harmony with the proposition which was at that time supported by Dumas, Laurent, and the advo- cates of the substitution theory, which teaches that chemical combinations form a whole, a unit. ‘This was at that time—perhaps improperly—called the unitary system. But to return to the point under discussion: I have just mentioned Laurent, and we should notice the fact that he was the first adherent of Gerhardt’s system of atomic weights and notation. I think it will be also interesting to recall some of the ideas which he then published in connection with this notation. We admit that oxide of potassium is formed of 2 atoms of potassium and | atom of oxygen, but caustic potash or potassium hydrate cannot be regarded as con- taining the elements of oxide of potassium + the elements of water. The molecule of potassium hydrate may be compared on the one hand to that of oxide of potassium, on the other to that of water, and is derived in a manner from the latter by the substitution of an atom of potassium for an atom of hydrogen. Thus water, potash, and anhydrous oxide of potassium LAURENT’S IDEAS. 85 are compounds of the same order containing respec- tively a single atom of oxygen combined either with 2 atoms of hydrogen, 2 atoms of potassium, or with 1 atom of potassium and 1 atom of hydrogen. The metallic hydrates are, therefore, compounds of the same order as the oxide, and cannot be represented as con- taining an anhydrous oxide + water. But there are also organic hydrates and oxides; and if we admit, in aleohol and ether, the existence of an ethyl group, so termed by Berzelius, we shall observe the same relations between water, alcohol, and ether as those which exist between water, potash, and oxide of potassium. Alcohol becomes ethyl hydrate, and ether ethyl oxide. The following formule, in which Et represents the ethy! group O,H,, will show these analogies :— H,0O, water. _ H,O, water. KHO, potassium hydrate. EtHO, alcohol. K,0, potassium oxide. Et,0, ether. This grand generalisation was afterwards extended by Gerhardt, who had first discovered the acid chlorides and anhydrous monobasic acids, to the acids. Upon comparing Ac(l, chloride of acetyl, with EtCl, chloride of ethyl, and hydrochloric acid, the same kind of rela- tions were discovered between acetic acid and anhydrous acetic acid as those between alcohol and ether. The salts and ethers of acetic acid can, as Williamson has shown, be added tv this synoptic table, which formed the basis of the celebrated idea of considering hydro- chloric acid and water as types :— 86 THE ATOMIC THEORY. HCl, hydrochloric actd. H,0, water. | EtCl, chloride of ethyl. AcHO, acetic acid. AcCl, chloride of acetyl. AcKO, acetate of potassium. KCl, chloride of potassium. AcEtO, acetic ether. Ac,O, acetic anhydride. It is from the new notation that these views, which embrace the discoveries of Williamson on etherification and those of Gerhardt on anhydrous acids, derive their simple and striking forms. The molecules of all the bodies just mentioned are comparable, under the con- dition that they are represented, in accordance with the principles developed by Gerhardt, by formule which represent the true molecular magnitudes. And it is important to remark that all these formule corre- spond, in the case of volatile compounds, to 2 volumes of vapour. ‘ We halve,’ he says, ‘ organic and mineral formule, so as to express their equivalent by 2 volumes.’ ‘ Equivalent’ is used instead of ¢ molecule,’ and from the preceding proposition we conclude that equal numbers of the molecules of gaseous or volatile compounds are contained in equal volumes of gases or vapours. This is the law of Avogadro and Ampére, which reappears as a guiding star upon the horizon after a long eclipse. And yet we cannot say that Gerhardt, at this period at at: least, gave himself up entirely to its guidance. The considerations by which he was principally influenced were rather of a purely chemical character—those which we have alluded to above. They were correct, and were found to agree with an equally correct idea which had been forgotten. The distinction between two species of minute particles, moleculesand atoms,which Avogadro GERHARDT’S NOTATION. 87 and Ampére had introduced without effect into science, and which Dumas had endeavoured to reproduce in his ‘Chemical Philosophy,’ was probably mentally clear to Gerhardt, though as yet it had not appeared in his writings. The word ‘ equivalent ’ was sometimes synony- mous with the term ‘molecule, sometimes with ‘atom’ or ‘volume.’ To quote his own words, ‘ Therefore,’ he says in p. 51 of his ‘ Précis, ‘ volumes, atoms, and equi- valents are synonymous in the case of simple bodies. It therefore follows that the densities of simple gases are proportional to thew equivalents. These pro- positions were not new, but they were inaccurate. These inaccuracies soon disappeared, and the distinction between molecules and atoms appeared clearly in the classic ‘ Traité de Chimie organique.’ Gerhardt’s system of atomic weights, which was immediately adopted by Laurent, gradually gained the approval of a great number of chemists. His works upon the theory of types, the discovery of the anhydrides and the chlorides of the monobasic fatty acids, gave him great authority, which profited him but little personally, but which will always be connected with his name. The simplicity of the new notation gave great clearness to the explanation of new facts and ideas. In England Williamson, Odling, Brodie, Frankland, Hofmann, Gladstone, Roscoe, and others successively adopted this notation. The new German school, which was then under the brilliant direction of Kekulé and Baeyer, adopted it at once, as also has been the case with the greater number of Russian and Italian chemists. In France Chancel has always made use of 88 THE ATOMIC THEORY. it, and I myself did so in my memoir upon the glycols in 1858. | I. The commencement of the year which I have just mentioned was, however, marked by the introduction of an important change. Cannizzaro proposed once more to double the atomic weights of a great number of metals. We must now point out the facts and follow the course of ideas which have proved the reform introduced by the illustrious Italian to be legitimate and gained for it the almost unanimous approbation of chemists. Gerhardt’s atomic weights were not. true equivalents, and molecules which occupy the same volume in a gaseous state are not always compounds of the same degree or the same order ; for Gerhardt afterwards referred these compounds to three different types—the hydrogen or hydrochloric acid type, the water type, and the ammonia type. That the molecules of chemical compounds differ from each other in their type—that is to say, in their degree of complication or in their manner of condensa- tion (which, moreover, the discoveries of Gay-Lussac had already indicated)—and consequently that molecules belonging to different types are not strictly equivalent, was an idea which was gaining ground in science. Correctly speaking, it was not at that time perfectly new; since the admission of the existence of sesquioxides, such as alumina and ferric oxide, it had been found that their capacity of saturation was three times greater than that of the protoxides; the sesquioxides are polyacid bases. CANNIZZARO’S REFORM. 89 On the other hand, Graham had already made the great discovery of polybasic acids. But other facts were soon added to the preceding, which introduced into science, if not the fact, at least the clearly defined notion of polyatomic compounds. I allude especially to the works of Berthelot upon gly- cerine, which produced such a number of important results, to which, I believe, I was the first to give their true interpretation in the order of ideas which we are now discussing. I must also mention Berthelot’s work upon the sugars and my own researches upon radicals and glycols, in which I endeavoured to define the part played by radicals in polyatomic compounds. These researches have introduced into science the idea that all chemical molecules are not mutually equivalent as far as their molecular complication is concerned, or, to use the phraseology of that time, ‘the degree of condensation affected in them by matter.’! In order to define the differences which they present in this relation, they were referred to more or less condensed types. Thus, to take a few examples, the constitution of nitric, sulphuric, phosphoric, acetic, and oxalic acids were represented by the following formulz :— H H H TYPE ef O+ Vi, | UTYPE 1} 9 TYPE Hs} 0; (NO x (S0,)" (PO)” iu 0 He, O, 1 O; Nitric acid. Sulphuric acid. Phosphoric acid. C,H,OY C,0,)” (CHLOY} (.09"! o, Acetic acid. Oxalic acid. 1 Annales de Chimie et de Physique, 3e série, t. xliv. p. 308. 5 30 THE ATOMIC THEORY. Similar formule represented the constitution and the increasing complication of the molecules of potash and ferric hydrate, for example, and of those of alcohol, olycol, and glycerine :— H H H TYPE a O TYPE ae co TYPE His} 0, nL RY" co : eho Potash. Ferric hydrate. (C,H,y (C,H,)" (C,H, y” 2 a oO 2 7 : 0, 3 “Hr, O, Alcohol. Glycol. Glycerine. These typical formuls had an advantage. They clearly indicated the fact that not only inorganic or organic radicals, but even simple bodies are capable of replacing 1, 2, or 3 atoms of hydrogen, and consequently differ in their substituting value. A distinction was therefore drawn between the monatomic, diatomic, and triatomic radicals. And as these radicals are in a manner nothing more than the representatives of the elements themselves, the distinction was extended to the latter. We shall presently develope this idea, that the power of combination or substitution with which radicals are endowed is essentially connected with that of the elements which they contain. But for a moment we must be contented with remarking that there is a gap between potash, which contains monatomic potassium, and ferric hydrate, which contains triatomic iron.! This 1 The formula} O;, which has been proposed by Odling, clearly expresses this idea of triatomic iron. Fe’ here takes the place of H, in three molecules of water, Hi} Oz. 8 CANNIZZARO’S REFORM. 9] gap has been, thanks to Cannizzaro, in a great measure filled up. This eminent chemist has doubled the atomic weights of a great number of metals, to bring them into harmony with the law of Dulong and Petit and the law of Avogadro. ‘These metals have been regarded, there- fore, as diatomic. Their oxides have become RO. Their R hydrates, rt | 0, answered to the hydrates of the 2 diatomic radicals—for example, to ethylene hydrate or glycol—which is given in the preceding table. We must not forget the influence which the discoveries of organic chemistry, and the interpretation given to them, have exercised upon the general conceptions of chemistry, and even upon the progress of mineral chemistry. We shall, therefore, return to this point in treating of atomicity. We here give the list of atomic weights now adopted by the majority of chemists. And, in order that the changes which the new discoveries and the progress of the theory have successively introduced into the system of atomic weights may be appreciated, we have, in the following table, marked elements with a distinctive sign. ‘Those which are printed in ztalics represent the elements to which Berzelius and Gerhardt attributed the same atomic weights, which they now retain; those which are marked by an asterisk have retained Gerhardt’s atomic weights ; those, finally, which are marked by two asterisks are the metals whose atomic weights were halved by Gerhardt and doubled again by Cannizzaro, | these double numbers being, moreover, those of Ber- zelius (see p. 62) :— 92 THE ATOMIC THEORY. Hydrogen Aluminium*. Antimony* . Arsenic Barium** Bismuth** Boron* , Bromine Cadmium** , Cesium Calcium** Carbon Cerium Chlorine Chromium Cobalt** Copper** Didymium . Erbium Fluorine Gallium Glucinum Gold** Indium Todine . Iridium** Iron** Lanthanum Lead** Lithium* Magnesium** Manganese** Mercury** . Molybdenum** _Nickel** Niobium Nitrogen Osmium** Oxygen 5 Symbols. H _—— Al Sb As Ba Bi B Br Cd Cs Ca Atomic Weights. 1 275 122 74:9 (75) 137-2 198°6 15-96 (16) TABLE OF ATOMIC WEIGHTS. 93 Atomic Symbols. Weights. Palladium** é ; oro 106°2 Phosphorus ‘ : ; ae 31 Platinum . . : ' * PG 196°7 Potassium* : ‘ Tet ase 39°137 Rhodium** : F 2 . Rh 104°2 Rubidium* ? : . = lepiiy 85'2 Ruthenium 3 : : =. 10 103°5 Seleniwm . ‘ , : . Se 78 Silicon* . i é : . Si 28 Silver* . : : ; . Ag 108! Sodium* . : : s NS 23°043 Strontium** . - : @.5r 87°2 Sulphur . - F : eet 31°98 (32) Tantalum . f : ped be 182 Tellurium / : - owa8 128 Thallium . : F “ ee El 203°6 Thorium . ; of eg bin) 233°9 Tin** : p : : oe oil 117°8 Titanium** . : ; pod byl 48 Tungsten** é ; ee Wi 184 Uranium . , : : 2 Us 120 Vanadium : we ee inc 51:2 Yttrium . : ’ : ae ¥ 89°6 Zinc** ; : ; ee ris 64:9 Zirconium ; : P me As Hye The limits which are imposed upon us by the character of this work make it impossible to mention the methods which have been employed in each parti- cular case for the determination of the atomic weights given in the preceding table. We must refer our readers for these details to the article upon Atomic Weights in the ‘ Dictionnaire de Chimie pure et appliquée’ 1 We have retained 108 as the atomic weigiie of silver, founding our opinion upon a recent observation of Dumas. Stas gave the number 107:93. The atomic weights of chlorine, bromine, and iodine being dependent upon that of silver, we have also retained the round numbers 35:5, 80, 127, as their atomic weiglits. 94 THE ATOMIC THEORY. We would especially draw the attention of the reader to the methods employed by Stas in the determination of the atomic weights of oxygen, sulphur, chlorine, bromine, iodine, nitrogen, potassium, sodium, lithium, and silver, giving results the accuracy of which is un- surpassed. We must also mention the labours and analyses of Marignac, and to the names of the two chemists we have just mentioned must be added the great name of Berzelius. However, setting aside the question of practical chemistry, to which we have just alluded, we must confine ourselves to the theoretical discussion which has justified the adoption of the new system of atomic weights. We shall endeavour to show that the atomic weights given in the preceding table are in harmony—first, with the law of Avogadro and Ampére; secondly, with the law of Dulong and Petit; thirdly, with the law of iso- morphism. We shall then devote a chapter to the proof of the fact that the chemical and physical properties of elements are dependent upon the atomic weights. We shall prove, lastly, that the notation which is derived from the present system of atomic weights ascribes to compounds their true molecular magnitudes, and allows a correct representation of chemical reactions. BOb The new system of atomic weights is founded upon the law of volumes, and 1s in harmony with the hypothesis of Avogadro and Ampére. The $ law,’ as it is generally called, of Avogadro and LAW OF VOLUMES. 95 Ampére may be enunciated as follows: Equal volwmes of gases or vapours! contain the same number of molecules. We have here two things, a group of facts and an hypothesis. The facts are a result, or rather a development, of the laws of Gay-Lassac. Gay-Lussac had shown—first, that gases combine in simple volumetric relations; secondly, that there is a simple relation between the volumes of the combining gases and that of the product of the combination. To these two laws may be added a third. There is a very simple relation between the volumes of all compound gases thus formed, and the hypothesis of Avogadro and Ampére consists in the assertion that all these com- pound gases occupy the same volume, and that the matter thus condensed into the same volume exactly represents the ultimate particles of the compounds— that is to say, the molecules. The accompanying table willexplain our meaning :— 2 vol. hydrogen + 1 vol. oxygen give 2 vol. water. 2 vol. chlorine + 1 vol. - »» 2 vol. hypochlorous anhydride. 2vol.nitrogen +1vol. ,, ,, 2 vol protoxide of nitrogen. 1 vol. sa + 1 vol. » gives 2 vol. dioxide of nitrogen. 1 vol. chlorine + 1 vol. hydrogen ,, 2 vol. hydrochloric acid gas. 1 vol. nitrogen + 3vol. ,, » 2 vol. ammonia. 1 vol. carb, oxide + 1 vol. chlorine ,, 1 vol. oxychloride of carbon. 1 vol. ethylene + 1 vol. » 33 1 vol. ethylene chloride. We here have clear examples of the two laws of Gay-Lussac (see p. 34), as well as of the third law of volumes. Between the volumes of the compound gases we have the very simple relation 2:1. The 1 Under the same conditions of temperature and pressure. 96 THE ATOMIC THEORY. : hypothesis of Avogadro consists in the assertion that this relation is still more simple, that it is 2 : 2, for the smallest quantity or the ultimate particle of oxychloride of carbon and of ethylene chloride which can be formed does not occupy 1 volume, but 2 volumes. This is au hypothesis, if you will, but one the truth of which is easily demonstrated, for experiment shows that the smallest quantity of carbonic oxide which enters into reaction occupies two volumes, which contain a single volume of oxygen; it shows, moreover, that the ultimate particle or the molecule of oxychloride of carbon cor- responds to the ultimate particle or molecule of carbonic acid gas, which occupies two volumes. These considerations apply to ethylene chloride and to other compounds. Consequently it is better to express the formation of oxychloride of carbon and of ethylene chloride in the following manner :— 9 vol. carbonic oxide + 2 vol.chlorine = 2.vol. oxychloride of carbon, 2 vol. ethylene oxide + 2 vol. chlorine = 2 vol. ethylene chloride. The two volumes thus formed represent the mole- cules of gases or vapours, and we are therefore led to give the following form to the statement of the law of Avogadro and Ampére. The molecules of compounds which are gaseous or volatile without decomposition occupy two volumes, if an atom of hydrogen occupies one volume. This pro- position holds good in the case of by far the greater number of volatile compounds, under the condition that their true molecular weights are attributed to these compounds, The proofs are so abundant that it is impossible to quote all the examples, and we must confine ourselves i lM ~ 7 LAW OF AVOGADRO. 97 to giving a list of the groups of compounds which obey the law in question. Water and its analogues, sulphuretted hy- q H,O =2 vol. rogen, &c. . . Hydrochloric acid and its analogues . HCl = 2 vol. Ammonia and its mineral and organic analogues; substitution derivatives of ammonia; organo-metallic radicals ae = 2 yok _of the type RX, CLO *~= 3 val. ClO; = 2 vol. N,O = 2 vol. NO = 2 vol. Oxides and anhydrides of chlorine, nit- | NO, = 2 vol. rogen, sulphur, and carbon : bt BO; = 2 Vol: SO, = 2 vol. CO = 2 vol. CO, = 2 vol. LCOS = 2 vol. (CH, = 2 vol. CH. = .2 vob. C.He = 2-vol. Allhydrocarbons. . . . ae i" ; sie CHa = Zi vol: C,,Hig = 2 vol. Oi," =" 2 vol: SiCl, .= 2 vol. PCl, =. 2 vol. PCl, = 2Zivol. ASCI,..., = 2\vol. SbCl, = 2 vol. POC], = 2 vol. . = 2 vol. Chlorides, bromides, and iodides of the metalloids and metals - “heen = 2 vol. COCL =. 2 vol. HeCl. = 2 vol SnCl, = 2 vol. ALC), = 2.vol | 2 vol. &c. &e. ra 2 @ a l 98 THE ATOMIC THEORY. Mercuric sulphide. ‘ : , . * Het ©* ta vor, Alcohols, glycols, phenols. : 5 C,H,O = 2:vol: Their anhydrides, ) ethyl oxide. ; (CH). 0 fa ok such as Vanek oxide . C370 = 2 vol. Aldehydes and aldehyde : » O,H,O)- ='2 vol. acetones iste : ; SA SMe 4 8) = 72 vol, Organic acids, such as acetic acid « “COCO ee avo, Their anhydrides lec anhydride pr i yligus a eee Vor succinic anhydride . C,H,O, = 2 vol. ethylene acetate (C,H,0,),.C,H, = 2 vol. ethyl oxalate . ©,0,(C,H,;), = 2 vol. &c. &c. Therlethers ethyl acetate . C,H,0,(C,H,;) = 2 vol. 7 > such as This table is undoubtedly very much abridged, but it is evident that it embraces a vast number of mineral and organic compounds, and it is difficult to imagine how, in the presence of such a wealth of facts and proofs, accumulated by the labours of the last fifty years, some chemists should still refuse or hesitate to believe the law of Avogadro and Ampére. It is useless for them to bring forward some cases which apparently form exceptions, and which we shall presently mention and discuss. In fact, we may say that the other physical and chemical laws of which we have spoken—the law of Dulong and Petit and that of isomorphism—do not rest upon such a number of imposing facts, and consequently upon such a solid foundation, as the law of Avogadro and Ampére. When a theoretical idea is true, the exceptions which are at first admitted gradually disappear, either because the new observations are more accurate than the old, or from a more correct interpretation of the facts. It also sometimes happens that these exceptions give rise to interesting developments of the theory and to a more extended generalisation. LAW OF AVOGADRO. 99 That it has been so in the case before us we will now proceed to show. ‘ I. Thirty years ago ordinary ether was represented by the formula C,H,O, which answered to two volumes, while the formula of alcohol, C,H,O.HO, answered to four volumes of vapour. Here was an exception to the law of volumes. Williamson came forward and showed that the old formula of ether should be doubled. The doubled formula, C,H,,O,, which in the new notation becomes C,H,,0, corresponds to that of alcohol, C,H,O, both representing two volumes of vapour. It is unne- cessary to insist upon the proofs which Williamson has given in his masterly memoir, and which are well known to all chemists—namely, the existence of mixed ethers, and the perfect agreement between the physical proper- ties of these ethers and those of ordinary ether, under the condition that the latter is regarded as a double _ molecule of the form (C,H;,),0. II. According to Gerhardt’s notation, which is still applied to organic compounds, monatomic hydrates do — not contain the elements of water, but merely the residue OH. Thus acetic acid is acetyl hydrate, C,H,0.0OH, and it is obviously impossible to separate from this formula the elements of water, H,O, which could be done with the old formula of Berzelius C,H,O, =C,H,O0,.H,O, or with the formula in equivalents C,H,0,=C,H,0,.H0. Thus it was the opinion of Gerhardt that the anhydrides of monobasic acids could not exist, and he had the singular fortune to discover them himself. But at the same time he showed, in striking confirmation of his ideas and formule, that in 100 THE ATOMIC THEORY. order to lose water the molecules of acetic acid must act in pairs, one of the molecules furnishing an atom of hydrogen, the other the residue OH. The anhydride formed, (C,H,0),O, or acetyl oxide, answers to two volumes of vapour. III. An analogous case occurred with the hydro- carbons called alcohol radicals, methyl, ethyl, &c. These are imaginary forms, said Laurent and Gerhardt, and have no separate existence. Kolbe and Frankland isolated them, but showed that their formule must be doubled.?- Free ethyl is not composed of two atoms of carbon and five of hydrogen, as the group C,H, in ethyl hydrate or alcohol, C,H,O.OH, but of C,H,, =(C,H,),, and this doubled formula corresponds to two volumes of vapour. The result of this is that the molecular weights of volatile compounds are accurately given by their densities. And if we refer these densities. to that of hydrogen taken as unity, we have only to multiply the numbers obtained by 2 to find the weight of the mole- cules compared with that of an atom of hydrogen=1. This is a general rule. The density referred to hydro- gen is the weight of one volume. The molecular weights are the weights of two volumes, for molecules occupy two volumes if an 1 Subjoined is the equation which expresses this dehydration of acetic acid— ~- — i) + 2s of 2 Two molecules of acetic acid. Acetic anhydride. 2 «Mémoire sur une nouvelle Classe de Radicaux organiques,’ Ann. de Chim. et de Physique, 3° sér., t. xliv. p. 275. ia Se Oe eee, ee LAW OF AVOGADRO. 101 atom of hydrogen occupies one; we must, therefore, multiply densities by 2 in order to obtain molecular weights. The atomic weights of a certain number of metal- loids and metals may be calculated from the molecular weights. Thus the atomic weights of phosphorus, arsenic, antimony, carbon, silicon, titanium, tin, mer- cury, and lead may be calculated from the molecular weights of the corresponding chlorides and ethides. For example— The molecular weight of chloride of silicon (ob- tained by doubling its vapour density) is 170, and analysis shows that 170 parts of chloride of silicon con- tain 142=4 x 35°5 of chlorine and 28 of silicon. The vapour density and analysis of chloride of silicon assign, therefore, to this body the formula SiCl,, and to silicon the atomic weight 28, for we have reasons for the belief that the molecule of chloride of silicon only contains a single atom of silicon. , The vapour density of zinc ethyl doubled = 123, the density of hydrogen being = 1. Now, analysis shows that these 123 parts of zinc ethyl contain two ethyl groups, which weigh 58, and 65 parts of zinc. 65 is the atomic weight of zinc, the composition of zinc ethyl being expressed by the formula Zn(C,H,),. The number 65 (64'9) is, moreover, confirmed by the law of specific heats. The molecular weight of mercuric chloride, calcu- lated from its vapour density, is 271, and analysis shows that these 271 parts of mercuric chloride contain 2x 35°5=71 of chlorine and 200 of mercury. Hence 102 THE ATOMIC THEORY. the simplest composition which can be assigned to mer- curic chloride is represented by the formula HgCl,, Hg being an atom of mercury. The atomic weight of mer- cury is thus fixed at 200, a number which agrees with the law of specific heats. According to the invaluable experiments of H. Sainte-Claire Deville and Troost, the vapour density of ferric chloride assigns to this compound the molecular weight 325. Now, 325 parts of ferric chloride contain 213=6 x 35°5 of chlorine and 112 (111°8) of iron. Are one or more atoms of iron represented by these 112 parts? In this case we should no longer prefer the simplest hypothesis, as in the preceding cases. The law of specific heats attributes to iron the atomic weight 56 (55:9); we must, therefore, admit that ferric chloride contains two atoms of iron, and six of chlorine, and that its composition is represented by the formula Fe,Cl,. These examples show the use which may be made of the law of Avogadro and Ampére in the determination of molecular weights and in settling atomic weights. We also see the assistance which chemists derive from the law of Dulong and Petit, when they have to choose between several molecular formule for a given compound, and consequently between several atomic weights for the same element. The considerations mentioned above apply to a great number of cases. That this is so will be seen from the following table,’ which shows the part played by the law of volumes, firstly in the determination of 1 Abridged from a more complete table which I have given in my Lecons de Philosophie chimique, 1864. Hachette. LAW OF AVOGADRO. 103 molecular weights, and subsequently in that of atomic weights. The experimental densities given in the third column are referred to that of air taken as unity. To refer them to the density of hydrogen we have only to multiply them by the number 14°44, which expresses the relation which the density of air bears to that ot hydrogen. The figures in the fourth column express the double densities referred to hydrogen, and conse- quently the weight of two volumes, 1 standing for the weight of one volume of hydrogen. They were obtained by multiplying these densities by 28°88. They are the same as the molecular weights given in the fifth column. Lastly, the sixth column! gives the molecular composition: it shows the weights of the elements contained in the molecule, and consequently the atomic weights, or in some other cases a multiple of these weights (see the remarks upon the atomic weight of iron, p. 102). The atomic weights thus ob- tained from the molecular weights are printed in large figures. 1 T have followed the example of Lothar Meyer 'n adding this column, : THE ATOMIC THEORY. 104 uesoip{y Jol x g=¢E woqIvd JO ZT eurpor Jo LET uasorpdy Jo TL outpor Jo LETS ussoip{y JjoT x g =e wWOqIVd JO ZT eulmoig JO O§ uosoipsy Jo, x ¢ = € | WOqIBd JO ZT suTIOTYO JO G.GE uasorpsy Jo T | ouTLo[YO JO G.GE f uas{xo JO 9T x Z = BE) wunrtueyes Jo BL J uwosfkxo JO 9T X & = QF inydyus jo SE uadfx0 JO OT X BZ = BE imydns jo Z&. uasorpsy Jo Z inyd{ns jo ZB uasorpsy JO % uasfxo Jo OL J OMoa]OPL Oy} FO worqtsoduroy GPL 86T SYUESTO AK IT Noe[O]L ——— I¥L| §88-F | T0AZ = VHO|* °° eprpor péqo zy Sol] StF | TOAG = THe, * ; poe orpormpéy, 6-€6 | 896-8 | ‘loa g = ig*HO | * = *— oprumorq [AqQey Log | ge2't | ‘toag = 1O°HO |* +s opraoqyo pAuOPy 9¢ LIG-L | “TOA G = TOH |". * pov o1sopqooipé HH 9IT| €0-4F | ‘ToAg = “099 eprpéque snorusyeg | 8.6L | $918 | ‘ToAg = *Os epupéyur ormyding 6-49 | LFBE | ‘ToAs = “OS |* *_~—- eprxorp epradyng FE | SIGLI | ‘OAS = SH |* * oprydins uesoipéy SI | ¢629-0 | ‘ToAg = O°H * ae ee eats A) BRAY ieee 2 VRE 2ESo| ary oF ~~ t| po«togor SB[NULLO if Sorpog JO soury yy 4 Sg | sorsueg Bins 105 PRESENT SYSTEM OF ATOMIC WEIGHTS. uesorpéy JOT X $= FU] uoqivod Jo SL J uasoipdy JOT x ST = ST uwoqivo JO ZIT X 9 = ZL p oruesie Jo Gs, oruesiv JO JZT x § = 18E\ otuesie Jo GF, f auTIoTYO JO g.cg x § = ¢-90T | otuasiv Jo GL f uasoipsy JoT xX €=E oluesie Jo GL eulIopyo Jo g.gg xX € = ¢-90T uwasfxo JO 9[ snioydsoyd jo Tg oulLOTYo Jo g.cg xX E = ¢-90T | snxoydsoyd jo TE f uesoipsy joT x ¢ = El snioydsoyd jo Tg J uasoipsy JOT x ¢ = ¢ | wogIed JO ZI ussoIjlU JO FL uesoipsy Jo. x g¢= El ussoiyIU JO FL J uasAxo JO QT | uoZorj1U JO PLS uas{xo0 Jo 91 | UdSOINIU JO FL x Z = 8aJ uasoipsy Jo, xX ¢ = | woqivd Jo Z% eul0ouy JO T.6 J . er 91 o9T 99F g- [81 SL g-€91 g-LET ik L-4& T-91 691 6-F9F 6-I8T SLL T-€9T 6-9ET GTE 61-1E L0-LT 86-64 L-4F &-FE 699-0 19-7 ST-9T 9008-9 £69-6 §-¢ GPL-Y FSI 80-T T6¢-0 860-1 LOST 98T-1 "TOA G "TOA G ‘JOA Z ‘JOA G ‘JOA Z “TOA G JOA G ‘JOA Z ‘JOA Z [OA G “TOA G ‘TOA & ‘JOA G ll ll Svs YSIVv]{ * guisivpAqyerqy, * @pIpor oluesly Oplloyyo oTMesIy QuUISIV eprmo[yoAxo snioydsoyg apoyo snroydsoyg sutydsoyg * gutuv_syj}oyy * eIMOulLy OPIXOIp UISOIJIN eprxojoid usSo0I4yIN epliong [AqIeW THE ATOMIC THEORY. 106 uasoipfy JOT x 0% = 06 | Toag = = 'CH*0)us | Se AES moqivd JO ZT X 8 = 96 8-§€6| 9-184} 120-8 | uty} JO 8.LTL SUHOTYO JOSE FS BYE NR ocag] yy. 6 | JOAZ = Noug |} * ° oepraoryo oruue ary Jo S-LITS | 8:93] 2896] 661-6 | T0Az tous PITOTYO oruuEyS uasoipséy JOT X 0% = 02 woqt¥d JO ZI xX 8 = 96 FIT] T-SPI] €0-9 | TAZ = 'CHOVIS |: * — * Oply}o too;]Ig WOdTTIS JO 83) oullony JO 1-61 X F = F-9L\ yeaa ie: i Ger, uoomszo ges] PPOT} 0] 29-8 | Toss AIS | oplony Woorrs eullo[yo JO ¢.cg xX F = orl | 2 4 Pal f [ moons yo gZs| OL) @-TLT) 686-9 | ‘Toag 10!S opHOTYyo WooI!s inydtns Jo Ze x % = F9 | m z ome. woqrteo JO BT 92 | FOL | SFOS | ‘OAS gO aprydustp uwoqie9 aUOTYO Jo g.gg x B= TL) | ‘ uaskx0 JO9T ¢; 66 | &86 | 668-E | ‘TOA = 1000 |* = *: SpHopyoAxo woqreD woqivo Jo SL J | uasfxo0 Jo 9T x 3 = 2S | ; ‘ — : : : soqasagesTi(| 1% | LPR | 608-T° |M10s.ge= 00 oplxoIp woqreD uasfx0 JO OT : ee ; ; ; uoqavo jo BT S| 86 | 648 | 496.0 | Toag = 00 aprxojord woqrep ia ae