HISTORY OF CHEMISTRY. LECTURES ON THE -HISTORY OF THE DEVELOPMENT OF CHEMISTRY SINCE THE TIME OF LAVOISIER BY DR A. LADENBURG, PROFESSOR OF CHEMISTRY IN THE UNIVERSITY OF I5RESLAU. ^Translates from tbe Second German EMtfon LEONARD DOBBIN, Ph.D., LECTURER ON CHEMICAL THEORY AND ASSISTANT IN CHEMISTRY IN THE UNIVERSITY OF EDINBURGH. ( With additions and corrections by the author?) PUBLISHED BY THE ALEMBIC CLt/B. Edinburgh Agent: WILLIAM F. CLAY, 18 TEVIOT PLACE. London Agents: SIMPKIN, MARSHALL, HAMILTON, KENT, & CO. LTD. 1900. AUTHOR'S PREFACE. IN placing these lectures before a wider section of the public, I consider it essential to indicate the point of view from which they have been prepared. I regard them as an attempt to follow the development of our ideas of to-day from those that were formerly current. Hence I have only gone back as far as Lavoisier ; because our science assumed a new aspect in his hands, and because it may be held that, as regards develop- ment, we are still passing through the epoch inaugurated by him. It has been my wish to arrange the matter of the lectures in such a way that the student may be enabled to obtain a survey of this portion of the history of chemistry with little trouble, and at the same time so that it may serve as a guide for those who may desire to engage their attention more par- ticularly with special investigations in this department. On this account I have expressed myself as concisely as possible, whilst, on the other hand, I have supplied moderately complete references to the original literature in connection with the subjects treated of. A twofold result appears to me to be V* PREFACE. attained in this way, inasmuch as the reader is placed in the position of being able to form an opinion forthwith regarding the value of the narrative, and to correct errors and omissions, while the labour of subsequent investigators is lightened. While I could scarcely consider it possible to give an absolutely accu- rate representation of the period in question, with its great wealth of discovery, still it has been my aim to furnish a useful contribution towards the history of the chief chemical facts and theories. It is almost unnecessary to say that the book has no preten- sions to completeness. I only felt justified in taking notice of those investigations and ideas which have exercised an influence upon the further development of chemistry, whereas I have at most merely referred to other investigations which, in my opinion, will still exert such influence. An objective treatment of the subject appeared to demand that it should be handled in this way. I have not hesitated to carry the history of the develop- ment of chemistry down to the present day, although the difficulty of the task has been greatly increased by my doing so. It is certainly in this part in particular that many correc- tions will still be necessary before the end in view is attained. How different the latest phases of our science will appear to subsequent investigators ! And yet the opinion of a contem- porary is not without value also, when it is moderate and free from prejudices or special leanings. This is exactly what I have PREFACE. Vll striven to attain. If I have not always been successful in doing so — if here and there I may have underestimated the merits of some and unduly asserted those of others-— this has been unintentional. If I have been severe in my judgment, I have at least been free from any personal feeling, and it has always been the matter alone that I have attacked. Should I have approached in some cases too closely to the limits of historical accuracy, or should I not have succeeded in representing fairly the claims of every one, I am fully prepared to rectify my error as soon as I am convinced of it. If my colleagues are interested in the subject, and assist me with their knowledge and advice, it may soon be possible, perhaps, to obtain an objective picture of the chemical theories of the last hundred years. I desire that this book may be regarded as an attempt in that direction, and that it may be judged indulgently. A. LADENBURG. KIEL, December 1886. AUTHOR'S PREFACE TO THE ENGLISH EDITION. THIRTY years after the appearance of the first edition of this book, an English translation of it is now being prepared. I regard this as a favourable indication of the permanent value of the book, since it is evident that the standpoint then adopted is intelligible at the present day and is still unsuper- seded. Moreover, it may be concluded that the exposition of the subject is not marred by national prejudices. In order to keep pace with the constant progress of the science, two new lectures have been added to the original fourteen. One of these appeared fourteen years ago, upon the j ,f,dT publication of the second German edition of the book ; whilst the other — the sixteenth — is here published for the first time. The English edition is a faithful translation, and, so far as I am able to judge, it is written in a good style. For these features my best thanks are due to the translator. I venture to express the hope that the book will find friends amongst the English-speaking peoples, and that it may contribute to stimulate interest in the history of our science. A. LADENBURG. GRASSENDALE, SOUTHBOURNE-ON-SEA, September 1899. TRANSLATOR'S NOTE. THE translator wishes to express his sincere thanks to Professor Ladenburg for the very cordial manner in which he agreed to the preparation of this translation of his well-known history ; as well as for his kindness in specially writing an additional lecture for this edition, thereby bringing the latter up to date, and for the great care he bestowed upon the revision of the proof-sheets. He further wishes to thank a number of friends to whom he is indebted for advice and assistance upon a variety of points. \ TABLE OF CONTENTS. LECTURE I. PAGE Introduction — The Phlogiston Theory in its First and in its Later Acceptations — Chemical Knowledge of the Phlogistians — Fall of the System - I LECTURE II. Revolution of the Views regarding Combustion — Priestley — Scheele — Lavoisier — Indestructibility of Matter • 15 LECTURE III. Chemical Nomenclature — Tables of Affinity — Berthollet's Views — Controversy regarding Constant Composition - 31 LECTURE IV. Richter's Investigations — Dalton's Atomic Theory— Gay-Lussac's Law of Volumes - — Avogadro's Hypothesis — Wollaston's Equivalents - - 47 LECTURE V. Davy's Electro-Chemical Theory— Discovery of the Alkali Metals — Discussion regarding their Constitution — Does Hydrochloric Acid contain Oxygen ? — Hydrogen Theory of Acids 67 LECTURE VI. Berzelius and his Chemical System— Dulong and Petit's Law — Isomorphism — Prout's Hypothesis — Dumas' Vapour Density • Determinations— Gmelin and his School - 85 XIV CONTENTS. LECTURE VII. PAGE Organic Chemistry at the commencement of its development — Attempts to determine the Elementary Constituents of Organic Compounds— Isomerism and Polymerism — Views regarding Constitution — Radical Theory - - - 108 LECTURE VIII. Further Development of the Radical Theory — Views concerning Alcohol and its Derivatives — Phenomena of Substitution — Dumas' Rule — The Nucleus Theory — The Equivalent of Nitrogen « 130 LECTURE IX. Graham's Investigation of Phosphoric Acid — Liebig's Theory of Polybasic Acids, and his Views with respect to Acids in general — Adoption of the Davy-Dulong Hypothesis— Discovery of Trichloracetic Acid — Attack upon the Electro-Chemical Theory — Replies of Berzelius — Copuke •' — 150 LECTURE X. Influence of the School of Gmelin — Theory of Residues — Coupled Compounds — Gerhardt's Determination of Equivalents — Dis- tinction of Atom, Molecule, and Equivalent by Laurent — New Characteristics of Polybasic Acids — Molecules of the Elements - 173 LECTURE XL Reasons for the Assumption of the Divisibility of Elementary Mole- cules— Fixing of the Molecular Weights, by Williamson, by Means of Chemical Reactions — Theory of the Formation of Ether — Fusion of the Radical Theory with Dumas' Types — Substituted Ammonias — Polyatomic Radicals — Gerhardt's Theory of Types and System of Classification 197 LECTURE XII. Mixed Types — Relation between Kolbe's Views and the Copulae of Berzelius — Radicals containing Metals — Conjugated Radicals — Kolbe and Frankland and the Views regarding Types— Poly- basicity as an Evidence for the Accuracy of the new Atomic Weights — Discovery of the Polyatomic Alcohols and Ammonias 220 CONTENTS. XV LECTURE XIII. PAGE Ideas regarding the Types — Elucidation of the Nature of the Radicals by the Valency of the Elements — Quadrivalence of Carbon — Specific Volume — Constitutional Formulae — Separa- tion of the Ideas of Atomicity and Basicity — Isomerism amongst Alcohols and Acids — Physical Isomerism — Unsaturated Sub- stances - - 249 LECTURE XIV. Theory of the Aromatic Compounds — Determination of Position of Substituted Atoms or Groups — Quinones — Artificial Dyes — Ring Compounds — Constitution of the Alkaloids — Syntheses — Condensation Processes - 272 LECTURE XV. The Fundamental Conceptions of Chemistry — Phenomena of Dis- sociation— Abnormal Vapour Densities — Constant or Variable Valency— The Doctrine of Valency in Inorganic Chemistry — The Periodic Law — Later Development of the Doctrine of Affinity — Spectrum Analysis — Synthesis of Minerals — Con- tinuity of Matter in the Liquid and Gaseous States — Liquefac- tion of the so-called Permanent Gases— Thermo-Chemistry — Electro-Chemistry — Photo-Chemistry — Molecular Physics — Morphotropy - - - 297 LECTURE XVI. The Doctrine of Phases — Van der Waals's Equation — Theory of Solution — Electrolytic Dissociation — Electro-Chemistry — At- tainment of High Temperatures — Low Temperatures — The New Elements in the Atmosphere— The Chemistry of Nitrogen — Transition Temperature — Stereo - Chemistry — Racemism — Syntheses in the Sugar and Uric Acid Groups — lodoso-Com- pounds— Terpenes and Perfumes— New Nomenclature - - 333 INDEX OF AUTHORS' NAMES - 357 INDEX OF SUBJECTS ... , 363 NOTES RESPECTING THE REFERENCES TO JOURNALS, ETC. IT is not anticipated that the abbreviations employed for the titles of journals, etc., will, as a rule, present any difficulty. The following unfamiliar abbreviations may be explained : — A. C. R. = Alembic Club Reprints. E. (following a reference to a foreign book) = English Translation. (This contraction is only employed in the cases of a few well-known translations.) In cases where journals have been issued in two or more series, references to volumes belonging to the second or any subsequent series have the series indicated by prefixed numerals enclosed within square brackets. I ... OF THE "7* UNIVERSITY HISTORY OF CHEMISTRY. LECTURE I. INTRODUCTION — THE PHLOGISTON THEORY IN ITS FIRST AND IN ITS LATER ACCEPTATIONS— CHEMICAL KNOWLEDGE OF THE PHLO- GISTIANS— FALL OF THE SYSTEM. '"T^HE value of historical narratives is undisputed. This value X no doubt varies with the subject matter which is dealt with ; but the history of human actions and of human know- ledge always forms one of the most interesting inquiries. If we are adherents of the Darwinian theory, and grant to this theory a warrantable latitude, the importance of a retrospect of bygone centuries is thereby enhanced. We are then obliged to recog- nise a steady progress of development ; history is no longer a mere enumeration of isolated facts in chronological order, as these succeeded one another fortuitously, but it embraces the development of the human mind and of human civilisation. It shows us the results of the influence which the most varied causes have exercised upon the most different natures, and may perhaps at some time enable us to discover the laws which regulate these results. From this point of view it cannot be denied that the development of the present condition out of any former one becomes of increased importance ; and hence the interest which the thinking public has taken in Buckle's " History of Civilisation " is easily understood. I do not, however, go so ar fas to assert that this actual standpoint is necessary in order to lend due importance to the representation of the past. The facts cannot be overlooked, A 2 HISTORY OF CHEMISTRY. [LECTURE 1. that knowledge itself affords a certain satisfaction to the human intellect, and that every one eventually seeks to draw lessons for the present from the destinies of nations in former times. The most pronounced opponents of Darwin, for example, must admit that a connection exists between the outstanding character and the fate of a nation, and even they will attribute the success or the non-success of great undertakings to material causes and circumstances. Assuming as a basis, then, the standpoint mentioned above, it may be asserted that a historical account of any science possesses an interest extending beyond that particular science. In a comparative study of the history of all intellectual sciences, certain general tendencies of speculation may perhaps be recog- nised which were predominant at particular times, and owed their existence to real, definite circumstances. In this respect the history of philosophy, in particular, is of importance for early times ; while for modern times the historical exposition of the natural sciences, in my opinion, possesses just as great, and probably even greater, importance. The subject matter treated of in the present work may hence find an application some day : it may be regarded as one of the many preparatory studies which will be required if the question of writing a history of the development of the human intellect should ever arise. If we limit our view, however, and inquire as to the interest which the historical representation of a science possesses for that science ; or, what concerns ourselves still more closely, if we merely consider the advantage which accrues from it for the study itself, or for the student, the points of view which then become paramount are entirely different. ' A retrospect of the past, especially in the exact sciences, alone affords a proper comprehension of what is accepted to- day. It is only when we are acquainted with the theories which preceded those accepted at present, that the latter can be fully understood; because there is almost always an inti- mate connection between them. It might appear in our science (where any final result is arrived at by the test of experiment) LECTURE I.] HISTORY OF CHEMISTRY. 3 that the earlier views, which give expression to a limited num- ber of facts only, must not merely be supplanted by the later theories, which deal with a larger class of phenomena, but that they must lose their importance altogether. For the most part, however, this is not the case. On the contrary, a certain connection between successive hypotheses can very frequently be observed. When the general development is followed up, the effects of the earlier ideas can be recognised in the later views, and it is in this way that the latter first come to be properly understood. The abandonment of a theory is not always accompanied by a revolution. Such, indeed, is scarcely conceivable in the higher stages of the development of science ; and even when new modes of explanation are proposed, traces of former opinions may still be recognised in the direction which these take. But quite apart from this real advantage of the study of history, which thus, in my opinion, leads to a clearer under- standing of our present position, yet another advantage may be adduced which is perhaps of still greater value to the student : namely, the accurate estimation of the value of theories. An examination of the past shows us the muta- bility of opinions ; it enables us to recognise how hypotheses, apparently the most securely established, must in course of time be abandoned. It leads us to the conviction that we live in a state of continuous transition ; that our ideas of to- day are merely the precursors of others ; and that even they cannot, for any length of time, satisfy the requirements of science. We learn from any historical exposition that our natural laws are not incontrovertible truths or revelations, but that they can be regarded as the expression, for the time being, of a certain series of facts, which are thereby summarised and, as we say, explained in the most practical way for us. We recognise that these laws do not originate suddenly in the head of a single individual, like Minerva in the head of Jupiter. It is only slowly that the fundamental ideas which underlie them mature, and that the requisite facts are ascertained by the labours of many ; until, at last, the law common to them all is 4 HISTORY OF CHEMISTRY. [LECTURE I. announced by some one, or often by several persons simul- taneously. Further, by the study of history our faith in authority is diminished — a faith which produces pernicious effects by obstructing the way for any original development of the individual. On the other hand we also learn, it is true, that theories are necessary for further development ; and that, although the actual teachings of science may lie in the facts, the real intel- lectual significance can only be acquired by connecting isolated observations by means of hypotheses ; so that the present posi- tion consists far more in the mode of explaining observations than in the observations themselves. When the point of view which I regard as essential for our subject is thus made clear, it will be understood that I direct my attention principally to theories, and only take cognisance of those experimental investigations which have contributed to the establishment or the overthrow of general views. The early history of our science has been treated most excellently, and in detail, by Hermann Kopp ; and for this reason I confine myself to the last hundred and twenty years ; that is, to the period of modern chemistry, or to that of quan- titative investigations.1 I must not omit, . however, to give a short description of the views which prevailed in chemistry prior to Lavoisier. The influence of the Greeks upon art and literature, on their reawakening after several hundred years' sleep, is so well known that it need not surprise us if we recognise a similar influence in science also. The four elements of Empedocles, water, earth, fire, and air, which, in Aristotle's system, are repre- sentative of the four cardinal properties, moist, dry, hot, and cold, are quite familiar. I attach great importance to finding fire here amongst the elements, and to seeing it regarded as a material substance. As we shall learn in what follows, the first chemical theories have reference to the phenomena of combus- 1 Kopp's Entwickelung der Chemie in der neueren Zeit (1873) appeared five years after the publication of the first German edition of this book. LECTURE I.] HISTORY OF CHEMISTRY. 5 tion, and the phlogiston theory becomes more comprehensible when we minutely study the views of the Greeks and the Romans. Amongst these peoples, combustion is already looked upon as consisting in the separation of the material of fire ; and Pliny regarded the easy inflammability of sulphur as a proof of its being largely composed of a fire material.2 At a later date sulphur was itself assumed to be the fire material ; and from that view the hypothesis that all metals contained sulphur, unquestionably arose. These few words concerning the chemical theories of the ancients appear to me sufficient in order to understand Becher and his follower, Stahl. Both of these based their views upon those of the Greek and of the Roman philosophers ; in the same way that we find so many imitators of Greek art at the same period, that is, in the seventeenth century. A difference may, it is true, be pointed out between them ; namely, that only the latter intentionally and knowingly followed in the footsteps of the ancients, whilst the former declared themselves to be their opponents. Thus Becher says : " A good peripatetic is a bad chemist." He replaces the four elements of Empedocles by three others : the verifiable, the inflammable, and the mercurial earths.3 It is not my business here to inquire whether it was Becher or Stahl who thought and wrought most with respect to the phlogiston theory. Still I will not omit to draw attention to the great modesty of Stahl, who wished that his own services should be attributed to his teacher and friend Becher : " Becheriana sunt quae profero" 4 Such examples are rare. The adherents of the phlogiston theory regard combustion as consisting in a decomposition : " only compound substances can burn ; " these all contain a common principle which Becher calls terra pinguis and Stahl calls Phlogiston. During the combustion this principle escapes and the other constituent of the substance remains behind. 2 Kopp, Geschichte. 3, 102. 3 Ibid. I, 179. 4 Ibid. I, 188. 6 HISTORY OF CHEMISTRY. [LECTURE I. This theory was applied to all combustible substances. Thus, according to StahFs views, sulphur consists of sul- phuric acid and phlogiston ; a metal, of its metallic calx (of its oxide we should say) and phlogiston. According to Stahl, sulphur was not identical with phlogiston, but, as with Pliny, it was rich in the principle of inflammability, which he did not know in a separate state. Soot appeared to be the substance richest in phlogiston ; in fact as almost pure phlogiston. It was in consequence of this that the conversion of the metallic calx into the metal by heating it with soot succeeded so well ; for the soot handed over its phlogiston to the metallic calx, so that a metal was produced again. In his experimentum novum Stahl tries to prove that the phlogiston in soot and in sulphur is identical. He shows here, how a sulphate can be converted by means of charcoal into liver of sulphur, from which sulphur is precipitated by the action of an acid. From the reduction of the metallic calces by means of soot, Stahl further infers the identity of the phlogiston of the metals with the inflammable principle in soot and in sulphur ; and thus he arrives at a proof that there exists only one such principle, which he calls simply Phlogiston (from <£Xoy«rTo Alembic Club Reprints, No. 7, 5. LECTURE II.] HISTORY OF CHEMISTRY. 17 experiments were independent of and nearly simultaneous with those of Priestley.2 Both chemists employed almost similar methods for its preparation. They obtained it from mercuric oxide, pyrolusite, minium, nitre, etc. Lavoisier also wrote a treatise on oxygen, but Priestley states that he had previously informed Lavoisier of his discovery, although the latter makes no mention of this.3 It is to be deplored, but unfortunately it seems to be established, that Lavoisier repeatedly tried to appropriate to himself the merits of others. I do not enter further into this matter here, because I regard it as inessential in the history of the development of chemistry. A man's own period is concerned with his personal qualities, and history with his works. Lavoisier paid with his life both for faults which he committed and for faults which he did not commit. His own time judged him. Posterity may regard him with admira- tion and indulgence. The different views which were held with respect to oxygen by its discoverers are of interest to us. Priestley, the worshipper of chance, who asserts that his greatest discoveries are due to the latter only, and for whom every new experiment is a source of new surprises,4 describes in detail how he discovered oxygen and studied its properties. He recognises that combustion proceeds better in this gas than in any other, and assumes, further, that atmospheric air owes its property of supporting combustion and respiration, to the presence in it of the gas which he has discovered. He finds, moreover, that it is absorbed by nitric oxide, whence he derives a method of determining the quantities of oxygen in mixed gases. What does he conclude from all this, however; how does he explain these phenomena ? According to him, when a substance burns its phlogiston must be able to separate from 2 Nordenskjold (see Carl Wilhelm Scheele, Nachgelassene Briefe und Aufzeichnungen, Stockholm (1892), xxi.) even endeavours to prove that the priority belongs to Scheele rather than to Priestley, but with this I do not agree. 3 Priestley, The Doctrine of Phlogiston established and that of the Composition of Water refuted. Northumberland (1800), 88. 4 Priestley, Experiments, etc., 2, 29, 39, 42, etc.; A.C.R. 7, 5, 12, 15, etc. B 1 8 HISTORY OF CHEMISTRY. [LECTURE II. it.5 But in order that this may take place the phlogiston must find another substance with which to unite. Combustion is possible in air ; therefore air can take up phlogiston, but that only to a certain extent, for after some time it becomes incap- able of supporting combustion any longer. It is then saturated with phlogiston. Substances burn in the oxygen gas discovered by Priestley better than they do in air : it is dephlogisticated air (a name which Priestley proposes for the new substance) or air deprived of phlogiston, and it is thus better fitted for taking up phlogiston than ordinary air is. The nitrogen, on the other hand, which remains behind after the oxygen of the air has been absorbed (and with regard to which Priestley is aware that it neither supports combustion nor respiration) is air saturated with phlogiston, or, phlogisticated air. With Priestley the exist- ence of oxygen was no argument against the assumption of phlogiston, which he defended till the end of his life. Thus we find him at the beginning of the present century, when the majority of chemists had given up the phlogiston theory, addressing letters to the French Academy from America (whither he had withdrawn, chiefly on account of his political opinions), in which he requests refutation of his views.6 This was not difficult to give, and although it was refused him by the learned French Society, I must not omit to point out what is fallacious in his mode of regarding the matter. " When a substance burns in air, the latter becomes phlo- gisticated " — if we burn phosphorus, we obtain phosphoric acid (or phosphorous acid), while nitrogen, the phlogisticated air, remains behind. But if we burn a candle, or coal, we obtain a mixture (consisting of nitrogen and carbonic anhydride), part of which can be absorbed by means of alkali, thus exhibiting a phlogisticated air, possessed of properties different from those of the preceding one. If we burn phosphorus in dephlogisticated air, nothing at all remains behind — the phlogisticated air van- ishes. The contradictions to which Priestley's system leads, become manifest when it is applied to the facts known even at 5 Kopp, Geschichte. I, 242. 6 Dumas, Le9ons. 115. Priestley, The Doctrine, etc. x. and xii. LECTURE II.] HISTORY OF CHEMISTRY. 19 that period. Priestley failed to recognise this, because his general chemical knowledge was small ; 7 because he placed no value upon the results obtained by others ; and because he uni- formly defended, with the most dogged persistence, the ideas which he once adopted. What, on the other hand, were Scheele's theoretical views ; how did he regard oxygen ? Scheele — the ideal of a pure ex- perimental chemist, the discoverer of numberless substances, a man who carried out the most difficult investigations with the most slender resources, who possessed in the highest degree the faculty of observation, so that an error can scarcely be found in any of his very numerous researches ; — Scheele, who does not, as happens to-day even with the best and most capable observers, overlook the half of the points, but grasps the phenomena in their entirety and examines them one by one, and for whom every new experiment forms a mine of great discoveries — what intellectual progress did he introduce into our science ? I must, unfortunately, reply that this was very small. His general views are so confused that I only enter with reluctance upon the task of giving an outline of them. Scheele laid down his views chiefly in a small work on " Air and Fire." The principal difficulty in giving an account of his opinions is due to the fact that phlogiston, the basis of them, is an unknown substance, to which he can assign every possible property ; so that sometimes he endeavours to identify it with an element which is known to us, while at others he seeks to place it side by side with the medium which the physicists call ether. In consequence, it often seems as if Scheele adopted the hypothesis of Cavendish and Kirwan, and by phlogiston understood hydrogen,8 and yet, on the other hand, this is not in accord with many of his other views. With him phlogiston is, generally speaking, a subtle substance weighing but littlej and concerning which he assumes that it is capable of penetrating the walls of his vessels. He regards oxygen as a compound of water with a hypothetical saline substance,9 in which compound 7 Kopp, Geschichte. I, 239. 8 Ibid. I, 262. 9 Ibid. I, 261, 20 HISTORY OF CHEMISTRY. [LECTURE II. there is, according to him, very little phlogiston. During com- bustion, the phlogiston of the combustible substance escapes, along with this saline substance from the oxygen, in the form of heat and light ; the other constituent of the combustible sub- stance— the metallic calx, for instance — remains behind, united to the water of the oxygen. With Scheele, hydrogen is almost pure phlogiston, which, however, contains a small quantity of that hypothetical substance (matter of heat) which is also present in oxygen. When hydrogen is burned with oxygen, the water of the latter separates, and the matter of heat from the oxygen unites with the hydrogen — the compound of phlogiston with little of the matter of heat — and produces heat and light. Thus it was only necessary to add some of that hypothetical substance to hydrogen in order to convert it into heat or light. Scheele's views are at variance with all the relations by weight, about which he troubled himself very little. In accordance with his views, the metallic calx, for instance, should weigh less than the metal plus the oxygen consumed ; since the phlogiston of the former escapes, in combination with the matter of heat from the latter, as heat and light. The assumption of a ponderable matter of heat, which plays a great part in his arguments, was at variance with the earlier experiments of Boerhave (compare p. 8), so that Scheele, in his theoretical views, came nearer to those who would retain Stahl's doctrine at any cost, than to those who desired an explanation of the observed facts, free from preconceived ideas. I leave these, however, and I do so all the more willingly, because I now wish to turn to the ideas and observations of Lavoisier, which are accessible and comprehensible to every one, and constitute the basis of the chemistry of to-day. It is not requisite that I should enumerate and describe all the researches of this accomplished investigator ; this would exceed the claims which any one could make, in such a his- torical sketch as I intend to give. On the other hand, the im- portance of the philosopher with whom I have now to deal, requires that he should be treated of apart from his con- temporaries. LECTURE II.] HISTORY OF CHEMISTRY. 21 The characteristic which distinguishes his researches from those of most of the other chemists of his time, is the sys- tematic consideration of the quantitative relations, which con- stitutes, in his hands, a decisive criterion with respect to phenomena. Before Lavoisier's time — and I recall this in- tentionally— Rey,10 and after him, Hooke n and Mayow, had turned their attention to the increase in weight on combustion. The views which they stated, come very close to the correct explanation of the process — Mayow approaching nearest to the truth. In Mayow's opinion the real substance which affects combustion is the " nitro-aerial spirit " which is present in the air and unites with the metal during calcination. (The name of this substance is intended to recall its occurrence both in nitre and in the air.) For any process of combustion there are requisite, according to him, not only inflammable particles (which he designates " sulphureous particles "), but also the presence of this nitro-aerial spirit, the taking up of which explains the increase in weight.12 The establishment of the phlogiston theory, which occurred at this period, and the acceptance that it met with, show to how small an extent these views were then understood, or, indeed, definitely established. Notwithstanding this, the priority in the mode of explaining the process of combustion cannot be claimed for Lavoisier. At the same time, the latter did not obtain his views from the chemists above mentioned, whose works were not widely dis- seminated, and were disregarded. But what places Lavoisier before any of them is the fact that he not only stated, as they had done, an idea which could be employed to explain some phenomena, but that, with the balance in his hand, he also vindicated, by means of a series of brilliant investigations, the . universality of the principle of the conservation of matter. He thus proved that he possessed not only a speculative head, but that he was also a scientific thinker and worker, who tested his views by means of intelligently conceived experiments, and, from these, further created new ideas. 10 Alembic Club Reprints, No. II. n Ibid. No. 5. la Kopp, Geschichte. 3, 135. 22 HISTORY OF CHEMISTRY. [LECTURE II. It cannot be asserted, at least I have not been able to read it out of Lavoisier's works, that he stated the principle of the indestructibility of matter as an axiom. But he recognised the truth of the law, for why, otherwise, should he have had con- structed for one of his first investigations, " On the Conversion of Water into Earth," a balance which surpassed in accuracy, everything that was known at the time in such instruments ? He recognises the truth of the principle, but he does not state it. Proof lies with experiments, and not in words, and so he holds back until a suitable opportunity : just as he held back the attack upon the phlogiston theory until he saw that the moment had arrived when, with a single blow, he could over- throw the house of cards, held together as it was, by decaying preconceived notions only. Accordingly we find his ideas about this fundamental doctrine expressed in his works only here and there, where it is necessary for him to furnish, at once, grounds for an opinion, the experiments in support of which are not yet completed. An instance is furnished, for example, in his first treatise on the composition of water (which substance he finds to consist of hydrogen and oxygen), where he would like to prove that the weight of the water produced by the union of hydrogen and oxygen is equal to the sum of the weights of the two gases employed — a point which he had not, at that time, established by experiment. He there states that this necessarily follows, since the whole is equal to its parts,13 and nothing but water is produced by the combustion. At that time the priority of this discovery was disputed, and had, with- out injustice, been ascribed to Cavendish and not to Lavoisier. As it appears from a letter of Blagden's u and also from a letter from Laplace to De Luc,15 Lavoisier was acquainted with Cavendish's investigation prior to his own experiments, and hastened to publish his results. It is in this way that we obtain knowledge of a fundamental doctrine which had long been clear to him, but which was only at once adopted by a 13 Lavoisier, Oeuvres. 2, 339. 14 Crell's Annalen. 1786, I, 58. 13 Kopp, Beitrage. 3, 271. LECTURE II.] HISTORY OF CHEMISTRY. 23 very few chemists. At a later date Lavoisier expresses himself still more distinctly, stating that the substances employed and the products obtained can be brought into an algebraical equa- tion ; and that if any term is unknown it can be calculated.16 This is the first idea of those equations which we now employ daily. Still we do not foresee the course of the development of this great thinker's ideas, if we follow him, in a general way at least, from his first experiments onwards. It may be said, how- ever, that the development of his views is that of the chemistry of his period. One of Lavoisier's earliest investigations deals with the supposed conversion of water into earth.17 He points out the inaccuracy of this supposition, which was generally held at the time. It is interesting to follow him in his experiments. He seals up a quantity of water in a glass vessel, which was known at that time as a pelican, and is so arranged that a glass tube which is fused on to its neck above, leads the condensed water back again into the body of the vessel. He weighs this empty, and then with water in it, after closing the single opening by means of a glass stopper. He then distils the water for a hundred days. The formation of earth appears after a month, but he proceeds with the distillation until the quantity formed seems to be sufficient. He now weighs the apparatus again, and finds it to be just as heavy as before ; whence he concludes that no fire material has penetrated it ; for otherwise, he con- siders that the weight must have been increased. He next opens it, weighs the water along with the earth, and finds its weight increased, but that of the glass diminished. This leads him to the assumption that the glass has been attacked by the water, and that the formation of earth is not a conversion but a decomposition. His conclusions accord exactly with his ex- periments ; yet he does not permit himself to be blindly led by them. Thus he finds the increase in the weight of the water to be a few grains more than the decrease in the weight 16 Dumas, Le9ons. 157. 17 Lavoisier, Oeuvres. 2, I. 24 HISTORY OF CHEMISTRY. [LECTURE II. of the glass. Another person might perhaps have concluded from this that there had been a production of matter ; Lavoisier, however, explains it as an error of experiment. Although this is a bold view for the period, yet it shows that he is guided by profound ideas, and understands how to criticise his own experi- ments. All later investigations have confirmed the accuracy of his explanation. Scheele,18 on his part, was occupied about the same time, with similar investigations, and arrived at the same results ; but the mode in which the Swedish chemist conducts the experiment is very different. He analyses the earth and finds that it consists of the same substances as the glass in which the water was heated. A later paper of Lavoisier's treats of the increase in weight on combustion. As early as 1772, he hands in to the French Academy a sealed paper, in which he shows that the products of the combustion of phosphorus and sulphur are heavier than the substance burned. This he attributes to an absorption of air (of air, because oxygen had not then been discovered).19 In an investigation on the calcination of tin,20 he causes this ope- ration to take place in closed vessels ; these he weighs before and after, without observing any difference ; whence he con- cludes that no fire material is taken up. He shows, further, that the metal has increased in weight by just as much as the air has lost. Oxygen is discovered shortly afterwards, whereupon Lavoi- sier repeats the experiments of Priestley and of Scheele : but his conclusions are totally different from those of the two other chemists. He is already prepared for this discovery, and, with him, it becomes the foundation of a new theory. Oxygen he at once recognises as that part of the air which unites with the combustible substance during its combustion ; and he calls it " air eminemment respirable" In the same paper he shows that fixed air is a compound of carbon with this air, and that the latter is also contained in nitre.21 18 Dumas, Le9ons. 129. 19 Lavoisier, Oeuvres. 2, 103. 20 Ibid. 2, 105. a Ibid. 2, 128. LECTURE II.] HISTORY OF CHEMISTRY. 25 Some time afterwards, in 1777, he advances a complete theory of combustion.22 He says : — 1. Heat and light are disengaged during every combustion. 2. Substances burn only in pure air. 3. This is used up in the combustion, and the increase in weight of the substance burned is equal to the loss in weight of the air. 4. The combustible substance, by its combination with pure air, is usually converted into an acid ; the metals, on the other hand, are converted into metallic calces. In a paper on the composition of nitric acid, Lavoisier tries to prove, in the case of this acid also, the last of the above statements, which is of importance later when we come to deal with theories of acids.23 He there shows that this acid con- tains oxygen, while he is not aware that it contains nitrogen. The latter fact was discovered a few years afterwards by Caven- dish, on passing electric sparks through mixtures of oxygen and nitrogen,24 whereby nitric acid was obtained. Lavoisier at this time groups together the facts that carbonic acid (carbonic anhydride) consists of carbon and oxygen, sul- phuric acid (sulphuric anhydride) of sulphur and oxygen, phos- phoric acid (phosphoric anhydride) of phosphorus and oxygen, and nitric acid of " air nitreux " (nitrous air, i.e. nitric oxide) and oxygen. He further shows that an acid is obtained by treating sugar with nitric acid (that is, by supplying oxygen), and from this he concludes that Priestley's dephlogisticated air must contain the acidifying principle (Principe acidifiant — principe oxygine)^ From this time onwards, he regards all acids as consisting of a basis, or radical, and of \h\s principe oxygine. His " air pur " on the other hand, besides this acidifying prin- ciple, contains also the " matiere de chaleur" (matter of heat). It is certainly remarkable to find even Lavoisier speaking of a fire material, which he afterwards designates " calorique" and of which I shall explain the signification. 22 Lavoisier, Oeuvres. 2, 226. w Ibid. 2, 129. ^ Kopp, Geschichte. 3, 231. 25 Lavoisier, Oeuvres. 2, 249. 26 HISTORY OF CHEMISTRY. [LECTURE II. The " matiere du feu " is not possessed of weight. Lavoisier shows this by burning phosphorus in closed vessels, whereby heat is liberated but no loss of weight takes place.26 Further, he causes water to freeze in closed vessels, and here, likewise, he finds no change of weight. Since he is aware, from his own experiments, that heat is disengaged in the process, he con- siders that he is justified in assuming that heat has no weight. A better notion of what Lavoisier calls " matiere du feu " will be obtained by my stating his views on the constitution of matter, which I take from his " Reflexions sur le Phlogistique." 2T According to him, matter consists of small particles which do not touch one another, since, otherwise, a diminution of volume by lowering of temperature could not be explained : 2S the matter of heat exists between these particles. The hotter a substance is the more of the matter of heat does it contain. In the investigations into the specific heats of various sub- stances, carried out along with Laplace, Lavoisier further proved that, for a like increase in temperature, substances do not take up like quantities of the matter of heat.29 Into the considera- tion of these experiments, as well as into those on the heat of combustion,30 I do not enter here. Lavoisier knows that, by the addition of heat, ice is first converted into water, and the water then into steam. Hence gases contain most of the matter of heat. This is what we should understand when he says that his " air pur" consists of the acidifying principle and the matter of heat. During combustion the former unites with the combustible substance, and the matter of heat is liberated. It produces heat and light. The following statement is very characteristic of Lavoisier's standpoint : 31 — " Heat is the energy which results from the im- perceptible movements of the molecules of a substance ; it is the sum of the products of the mass of each molecule into the square of its velocity." Here we find him in complete agree- 126 Lavoisier, Oeuvres. 2, 618. ^ Ibid. 2, 623. 28 Compare also Lavoisier, Traite de Chimie. I , etc. a Lavoisier, Oeuvres. 2, 289. 30 Ibid. 2, 318 and 724. 31 Ibid. 2, 286. LECTURE II.] HISTORY OF CHEMISTRY. 27 ment with the fundamental doctrines of the mechanical theory of heat. His views with respect to the heat disengaged during combustion, although not quite accurate, are also of great importance. He considers 32 that when a solid substance (phosphorus) burns in a gas (oxygen), and the product of the combustion is solid (phosphoric anhydride), the disengagement of heat is due to the condensation which the gas has undergone, in order that it may become solid. If the product is gaseous (carbonic anhydride), he attributes the disengagement of heat to the alteration of the specific heat. He advances the general view that the heat of combustion must be greatest when two gases unite to form a solid substance. How correctly he under- stood the application of these fundamental ideas, is shown by his mode of explaining the lowering of temperature produced by dissolving salts in water. Lavoisier assumes, as we do, that it is the change of state of aggregation which occasions the absorption of heat.83 He shows, further, that the evolution of heat which occurs on mixing sulphuric acid with water, is accompanied by a decrease in volume, and that both maxima coincide ; so that theory and experiment agree. I must not, however, enter too deeply into these matters, which belong, partly at least, to physics ; and, therefore, I return to his purely chemical investigations. Lavoisier adheres to Boyle's definition of an element,34 which we retain to-day. With him, an element is any sub- stance which cannot be further decomposed.35 What the signi- ficance of this definition is, and of what importance this idea of an element has become for the whole of natural science, has been specially pointed out by Helmholtz.36 The metals were first regarded as elements by Lavoisier.37 In a long paper he contests the prevailing view, which assumed the existence of phlogiston in the metals. These interesting 32 Lavoisier, Oeuvres. 2, 647. w Ibid. 2, 654. 34 Kopp, Geschichte. 2, 275. 35 Methode de Nomenclature Chimique. Paris (1787), 17. 36 Tageblatt der Naturforscherversammlung in Innsbruck. 37. 37 Lavoisier, Oeuvres. 2, 623. 28 HISTORY OF CHEMISTRY. [LECTURE II. disquisitions, which contain the annihilation of the preceding system, only appear towards the end of his short and brilliant scientific career. At first there was no explanation forthcoming for a series of phenomena which were in agreement with Kirwan's phlogiston theory. I refer to the behaviour of the metals towards acids ; to the hydrogen liberated ; and to the reduc- tions carried out by Priestley by means of hydrogen. It is only after the composition of water has been ascertained by Cavendish, Watt, and Lavoisier,38 that Laplace arrives at the idea (as Lavoisier relates 39) that on dissolving metals by means of acids, water is decomposed — that the hydrogen is, therefore, evolved from the water, while the oxygen of the latter unites with the metal to form an oxide. The phenomena of reduction now become clear also. The hydrogen unites with the oxygen of the metallic oxide, forming water, and the metal remains behind. Lavoisier tries to prove all these points by a series of excellent experiments. His investigations upon the decom- position of water, are, especially, of extreme interest.40 He passes water vapour over weighed iron turnings, heated to red- ness, and collects the hydrogen in an eudiometer. In this case also, he weighs everything — the water, the gain of the iron, and the hydrogen. In this way, he succeeds in finding out the quantitative composition of water; and the latter, along with the quantitative composition of carbonic anhydride, which he determines somewhat later,41 constitute the starting point for his researches on organic analysis.42 I wish to say a few words at least with regard to these researches ; since, even although the numbers obtained are not very exact, the methods are so important that I cannot pass them by unmentioned. Lavoisier places a weighed piece of charcoal in a dish, under a bell-jar containing a measured volume of oxygen, and 38 With respect to the part played by each in this most important dis- covery, compare Kopp, Die EntdeckungderZusammensetzungdes Wassers. Beitrage. 3, 235. 39 Lavoisier, Oeuvres. 2, 342. 40 Ibid. 2, 360. 41 Ibid. 2, 403. 42 Ibid. 2, 586. LECTURE II.] HISTORY OF CHEMISTRY. 2Q standing over mercury. Besides the charcoal, the dish con- tains a trace of phosphorus and tinder. By means of a bent, red-hot, iron wire, he ignites the phosphorus, which communi- cates the combustion to the tinder, and thus to the charcoal. After the combustion of the latter has ceased, he takes out the dish and weighs it, and so finds the quantity of the charcoal burned. He measures the volume of gas in the bell-jar, absorbs the carbonic anhydride by means of potash, and measures again. In this way he obtains the volumes of the carbonic anhydride produced and of the oxygen employed in the com- bustion ; and hence, all the data necessary to calculate the composition of carbonic anhydride. This he makes use of in carrying out the analysis of organic substances, such as spirit of wine, oil, and wax. He had satisfied himself, at an earlier date, that water and carbonic anhydride are alone produced by the combustion of these substances, whence he had quite correctly concluded that they contain carbon, hydrogen, and oxygen only. In the estimation of their quantitative composition, Lavoisier employs an appa- ratus similar to that indicated above. For example, he places a spirit lamp, which he weighs before and after, under the bell-jar, and burns some of the spirit: he also determines the quantities of carbonic anhydride produced and of oxygen employed, and from these data he is able to calculate the composition of the spirit. With this I shall close the consideration of Lavoisier's chemical investigations. A superficial estimate of his merits is all that I have been able to give. It is only by a minute study of his works that any complete idea of his significance, and any proper understanding of how much our science owes to his great intellect, can be obtained. There are, however, certain directions of his activity that I have not even mentioned, as, for example, the researches on respiration, about which I still wish to say a few words. Priestley already knew that oxygen was necessary for respiration.43 Lavoisier shows how it is used up in the lungs, in the formation of carbonic anhydride and of 43 Compare p. 17, 30 HISTORY OF CHEMISTRY. [LECTURE II. water, and how this process, which he properly classes as one of combustion, furnishes to man the heat necessary for his existence.44 He demonstrates that the expired carbonic anhy- dride derives its carbon from the blood itself; that in the pro- cess of respiration we thus, to a certain extent, burn ourselves, and would consume ourselves if we did not replace, by means of our food, that which we have burned. Since he next finds, by special experiments, that in strained activity the breathing is hastened, and that the consumption of carbon is thereby increased, he arrives at the conclusion that the poor man, com- pelled as he is to work, consumes more carbon than the indolent rich man ; but that the latter, by an unfortunate circumstance in the division of worldly possessions, can satisfy his smaller requirements — and that too by means of better food — much more readily than the poor workman can. He therefore calls upon society to remedy this evil by means of its institutions, to improve the lot of the poorer classes, and in this way to smooth away, as far as possible, those inequalities which apparently are established in nature. He closes this ingenious treatise with the words : 45 — " It is not indispensable, in order to deserve well of humanity, and to pay his tribute to his country, that a man should be called to those public and pompous functions which co-operate in the organisation and regeneration of empires. The physicist, in the quiet of his laboratory and of his study, can also exercise patriotic functions ; he can hope to diminish by his labours the many ills which afflict the human species, and to increase human pleasures and prosperity. And if he should only con- tribute, by the new methods which he may have shown, to the lengthening of the mean age of man by a few years, or even by a few days, he also may aspire to the glorious title of benefactor of humanity." His own time rewarded him badly for his endeavours. — Four years later, in 1794, he was guillotined by order of the Revolutionary Committee. 44 Lavoisier, Oeuvres, 2, 331. 45 Ibid. 2, 703. LECTURE III. CHEMICAL NOMENCLATURE— TABLES OF AFFINITY — BERTHOLLET'S VIEWS— CONTROVERSY REGARDING CONSTANT COMPOSITION. IT will readily be understood why a new era dates from Lavoisier, and why the latter is designated the reformer of chemistry, if we consider what were the theoretical views held before his time, and what they were at the period of his death. Lavoisier lived to have the satisfaction of seeing his views generally recognised, in France at least ; they gradually gained ground also in England and in Germany, where his works were translated, so that it may be justly said that phlogiston, at the beginning of the present century, had disappeared from scientific works. Lavoisier not only overthrew the old theory, but it is his chief merit that he introduced a new one in its place, and it is perhaps advisable to state here the most important heads of his theory. 1. In all chemical reactions it is the kind of matter alone that is changed, whilst its quantity remains constant ; conse- quently, the substances employed and the products obtained may be represented by an algebraic equation in which, if there is any unknown .term, this may be calculated. 2. In the process of combustion the burning substance unites with oxygen, whereby an acid is usually produced. In the combustion of the metals, metallic calces are produced. 3. All acids contain oxygen, united, as he expresses it, with a basis or radical which, in inorganic substances, is usually an element, but in organic substances is composed of carbon and hydrogen, and frequently contains also nitrogen or phosphorus. If we contrast these three statements with the views of the phlogistians, i.e., with the theories which prevailed prior to 32 HISTORY OF CHEMISTRY. [LECTURE III. Lavoisier, we shall appreciate the reformation introduced by him into chemical science. The direction of chemical thought was entirely changed, and the facts hitherto ascertained appeared in a new light. It was necessary, in a sense, to translate them in order to understand them, and as it was recognised that a new language was required for their proper comprehension, the need of a system of chemical nomenclature made itself apparent. I pass over all the attempts which had been' made prior to this period to secure a uniform mode of expression, as these did not lead to any results worth mentioning, and as they occurred during a period to which I can only slightly refer. I wish, however, to mention that Bergman repeatedly approached the French chemists with a view to securing uniformity in the naming of substances. In 1782 Guyton de Morveau, probably stimulated by Bergman's suggestions, travelled to Paris, and laid before the Academy there a pro- posed system of chemical nomenclature. This system con- tained much that was new and good ; but it could not secure the approval of the principal chemists of the period, as it assumed the existence of phlogiston, which even at that time was vigorously contested by Lavoisier. The latter afterwards succeeded in convincing Guyton of the accuracy of his new views. Guyton agreed to reconstruct his system, and in 1787, in conjunction with Lavoisier, Berthollet, and Fourcroy, he published the Nomenclature Chimique. As this system em- bodies the principles and constitutes the basis of the chemical nomenclature now employed, I cannot pass it by without men- tion, and I shall therefore state at least its more important features. In doing so, it will frequently be necessary to em- ploy French words. For practically all of these there are exact English equivalents. Substances are all divided into elements and compounds. Amongst the former there are included all those substances which could not then be further decomposed, and these are classed under five headings. Of these headings the first em- braces those bodies which are of very common occurrence, and LECTURE III.] HISTORY OF CHEMISTRY. 33 whose behaviour seems to indicate that they are not decom- posable. Examples of these are : — i, Heat (Calorigue) \ 2, Light ; 3, Oxygen ; 4, Hydrogen ; 5, Nitrogen (Azote}. The second class contains the acidifiable bases, such as sulphur, phosphorus, carbon, &c. The third embraces the metals ; the fourth the earths ; and the fifth the alkalies, which, as is well known, had not at that time been decomposed. The names of the substances belonging to the second, third, and fourth classes are for the most part unchanged ; the alkalies are called potash, soda, and ammonia.1 For all these substances, which, with the exception of ammonia, were regarded as ele- ments, the authors observed the principle of designating each by a single word. The radicals constitute an appendix to the elements. These are substances which they regard as decomposable, but which exhibit certain resemblances to elementary bodies. Next come the binary substances, consisting, as they do, of two elements. The acids occur in this class. According to the theory of Lavoisier, the acids all contain oxygen. Their names are in each case composed of two words, of 'which the first is common to them all and indicates their acid character (acide\ while the second is a specific name indicating the element or radical occurring in each. Thus we have acides sulfurique, carboniquc, phosphorique, nitrique, etc. Two acids containing the same element or radical are distinguished by the different termination of the specific name ; that containing the smaller proportion of oxygen receiving the termination eux, whereby such names as acides sulfureux, nitreux, etc., are obtained.2 Hydrochloric acid is called acide muriatique^ and the existence of oxygen in it is assumed ; while oxygen is sup- posed to be present in still greater quantity in chlorine — the acide muriatique oxigene? The names of the binary substances of the second group, i.e., of the basic compounds containing oxygen, are formed in a manner exactly similar. For these the general designation 1 Nomenclature Chimique. 67. a Ibid. 85-86. 3 Ibid. 87. C 34 HISTORY OF CHEMISTRY. [LECTURE in. oxides is introduced, and to this word the specific name is appended in the genitive ; for example, oxide de zinc, oxide de plomb, etc. The remaining binary compounds are distinguished as sulphur, phosphorus, carbon, etc., compounds, and receive the class names : sulfures, phosphures, carbures, etc. Compounds of the metals with one another are called alliages (alloys), the expression amalgames being retained, how- ever, for mercury alloys. Amongst the ternary compounds, the salts alone need be mentioned. They obtain their class names from the acids from which they are derived, and are called accordingly : sulfates, nitrates, phosphates. The termination ate becomes ite when the salts are derived from the acid poorer in oxygen instead of from that richer in oxygen. The name of the base is appended ; for example, sulfate de zinc, de baryte, etc. When the salt has an acid reaction, the word acidule is employed ; on the other hand they call basic salts sursature de soude, etc.4 Relatively few double salts were known at that time. The designation intro- duced for these was not very convenient ; for example, tartar emetic was called " tar 'trite de potasse tenant d'antimoine " 5 (tartrate of potash containing antimony). This general review may suffice. Berzelius, as is well known, considerably extended these beginnings of a rational nomen- clature, and I shall refer to some of his improvements and expansions when considering his period. On comparing the science of to-day with what I stated in the preceding lecture regarding Lavoisier's views, it will be possible to judge of the extent to which the latter have been retained. Lavoisier's theories required modification on several points ; but on others his ideas were attacked without result, since it has been necessary to return to them again. Thus Lavoisier's theory of acids is now abandoned by the majority of chemists. The introduction of the new views only took 4 Nomenclature Chimique. 93 and 97. 5 Ibid. 52 ; on p. 235 it is called " tartrite de potasse antimonie " (antimoniated tartrate of potash). LECTURE III.] HISTORY OF CHEMISTRY. 35 place, however, long after his death, and therefore I postpone the consideration of this matter to a later lecture. We must occupy ourselves here with another attack upon Lavoisier, which was eventually decided in his favour; and is of im- portance on this account, that through it a strict separation of mixtures from compounds was first brought about. This attack had to do with the question whether chemical combinations are possible in all proportions, or whether sub- stances can combine in certain fixed proportions only. The latter view, as is evidenced by many of his investigations, was assumed by Lavoisier ; and indeed it seems to have been accepted as self-evident by all the chemists of his time, without having been proved. But a book appeared in 1803 which attracted the greatest attention in the scientific world, both on account of its contents and of the form in which these were set forth ; and in this book, amongst other things, the constancy of chemical proportions was denied, on the ground both of theoretical speculation and of experimental investigations. The work to which I refer is Berthollet's Statique Chimique, and if I am to render intelligible the importance of the attack which it contains, I must give at least a slight sketch of Berthollet's extremely interesting general theoretical ideas. I extract these from the work just mentioned, and from some scattered essays by its author upon the same subject.6 Berthollet's book will always be of importance in chemistry, chiefly because the fundamental doctrines to which the author subordinates all chemical reactions, are the principles of me- chanics and of physics ; and these must necessarily possess a value in chemistry. And even although many of Berthollet's conclusions do not harmonise with experiment, and have long since been disproved, still this does not damage the basis of his conceptions. The work as a whole is chiefly directed against the false view which had been adopted with regard to the affinities of substances ; and against the misuse which was made at the 6 Ann. Chim. 36, 302 ; 37, 151 and 221 ; 38, 113 ; 39, 3. 36 HISTORY OF CHEMISTRY. [LECTURE III. time of so-called tables of affinity. The latter were tables which purported to express the strength of the affinity of substances, and they had been drawn up by a large number of chemists. The earliest originated with Geoffroy,7 and dated from the year 1718. It consisted of various tables in which the other sub- , stances were so arranged with respect to a particular one, that each preceding substance always decomposed the compound which the next succeeding one formed with this particular substance. Thus, for example, his table for acids in general ran : fixed alkali ; volatile alkali ; earths ; metals. The con- struction of such tables was one of the chief occupations of chemists in the middle of the eighteenth century. With them they associated the erroneous view that the affinity of one sub- stance towards another is invariable ; and it was only by degrees that chemists were convinced that this was an error. In 1773 Beaume pointed out that the affinities were different at ordinary and at very high temperatures (in the wet and dry ways), and that for each substance it would thus be necessary to construct two tables which should express its behaviour towards all other substances under these two different sets of conditions.8 Berg- man undertook this task,9 and it is truly astonishing what enor- mous pains he took in carrying it out. For each substance he constructed two tables, in which he compares its behaviour towards fifty-eight others, and these latter were so arranged that each preceding substance decomposed the compound which the next succeeding one formed with the particular substance concerned. From these tables it was, seemingly, possible to foretell the results of all reactions ; hence they were held in great esteem. When a new substance was discovered, such a table of affinity was at once constructed for it, and even Lavoisier respected this usage on the occasion of his investi- gation of oxygen, although he pointed out at the time that a similar table would really be required for every degree of temperature.10 7 Kopp, Geschichte. 2, 296. 8 Ibid. 2, 299. 9 Ibid. 2, 301. 10 Lavoisier, Oeuvres. 2, 546. LECTURE III.] HISTORY OF CHEMISTRY. 37 Berthollet, however, is the first to point out the error to which chemists had committed themselves in drawing up these tables. He destroys their significance by advancing the doc- trine that the effective action of a substance is related to its mass. Berthollet illustrates, especially by reference to salts, the laws in accordance with which chemical compounds are formed. He assumes that the same chemical effect is always connected with the neutralisation of any given quantity of a particular base (or acid). He represents this effect as the product of the affinity A, and the saturating capacity S (that is, for example, the quantity of acid necessary to neutralise a unit of weight of alkali). This gives — AS= Constant A — Constant ~~S~ that is, the affinities of two acids are inversely proportional to their saturating capacities,11 or just the reverse of what Berg- man had regarded as correct.12 But, according to Berthollet, the quantity Q of any sub- stance present exercises, quite generally, an influence upon the chemical action, which he considers to be proportional to the product of Q into the affinity of the substance. He calls this product the chemical mass.13 In the case of acids, the chemi- cal mass may also be stated as proportional to the product of the saturating capacity S into <2, the quantity present, as Berthollet likewise points out.14 The effects produced by affinity do not, however, depend exclusively upon the chemical mass : they are varied besides by the state of condensation of the substance concerned, and are subject, therefore, to the physical conditions of the experi- ment (to the pressure, temperature, etc.). As regards the state of condensation of matter, it is, according to Berthollet, a con- 11 Berthollet, Statique Chimique. I, 71 ; E. I, 45. 12 Kopp, Ges- chichte. 2, 314. I3 StaL Chim. I, 72 : E. I, 45. 14 Ibid. I, 16 ; E. I, xxiv. 38 HISTORY OF CHEMISTRY. [LECTURE III. sequence of the two opposed forces, cohesion and elasticity. The predominance of the former brings about the solid state, and that of the latter, the gaseous state; while in liquids a balance exists between the two. If all acids were in the same state of condensation, then that one should be regarded as the strongest, of which the smallest quantity is necessary to saturate a given weight of a base ; or, as we should now say, of which the equivalent is smallest. Berthollet applies these principles especially to simple and to double decompositions. According to him, when we add an acid to a dissolved salt, a partition of the base takes place between the two acids in the ratio of their affinities, that is, in the ratio of their masse chtmiytje^ Both salts and both acids are thus present in the solution ; yet this only holds when both salts possess approximately the same solubility, because a balance is then established which is dependent not only upon the strengths of the acids, but also, and in particular, upon the quantities of each present. He further draws attention to the fact that conviction as to the accuracy of this view cannot be obtained by evaporating the solution and permitting the salts to crystallise out, because as soon as the quantity of water is no longer sufficient for complete solution, the phenomena which take place depend chiefly upon the forces of cohesion and crystallisation, that is, upon the different solubilities of the substances.16 Thus, upon crystallising out after mixing a solution of potassium nitrate with sulphuric acid, the more sparingly soluble salt potassium sulphate is alone obtained ; whereas, according to Berthollet, potassium nitrate and potassium sul- phate are both present in the solution. When one salt is much more soluble than the other, it is the less soluble one that is principally formed ; and if one is quite insoluble, there is no partition, but complete decomposi- tion. In this way Berthollet explains, for example, the com- plete precipitation of barium nitrate by means of sulphuric 15 Stat. Chim. I, 75 ; E, I, 49. 16 Ibid. I, 82 ; E. I, 54, LECTURE III.] HISTORY OF CHEMISTRY. 39 acid. The barium sulphate, in consequence of its insolubility, is removed from the reaction ; a constantly progressing parti- tion taking place until the whole of the barium sulphate is precipitated.17 A similar thing occurs in the cases of volatile acids or bases. In this case also a partition takes place, in the ratio of the masse chimique ; but as one product (carbonic anhydride, for example) escapes, the decomposition proceeds to the end.18 Nevertheless, it is only in cases of extreme preponderance of cohesion (insolubility) or of elasticity (volatility) that com- plete decompositions are observed. Cases of partial decompo- sition are far more frequent. Thus, according to Berthollet, calcium salts cannot be completely precipitated by means of oxalic acid.19 His views regarding double decomposition are similar. Four salts are, in general, produced in these cases, and the formation of two only is confined to cases where the cohesion, or solubility, is totally different. This furnishes the explanation of the so-called reversible reactions. Amongst these reactions there are, in the first place, the various crystallisations which may be obtained at different temperatures from the same mixture of salts, if these salts possess solubilities which vary greatly from one another with changes in temperature. Berthollet adduces several examples of this kind,20 and I shall mention one of them. If a solution contains soda, magnesia, sulphuric acid, and hydrochloric acid, Glauber's salt crystallises out from it at a very low temperature (o° C.), whereas sodium chloride is ob- tained on evaporation. Hence, at o°, magnesium sulphate and sodium chloride must change into sodium sulphate and mag- nesium chloride, whilst at high temperatures the reverse takes place. In the same way Berthollet is also able to explain the phe- 17 Stat. Chim. I, 78 ; E. I, 51. 18 Ann. Chim. 36, 314. 19 Stat. Chim. i, 78-79; E. i, 51-52. 2° Ibid. I, 101-102 and 129-130; E. I, 7 1 -72 and 395-396- 40 HISTORY OF CHEMISTRY. [LECTURE III. nomena which are denominated, according to Bergman's views, by the affinities " in the wet and dry ways." Thus, for example, dissolved silicates are decomposed by almost all acids, whereas, on the other hand, silicic acid drives most acids out of their salts at a red heat. But Berthollet goes still further. Cohesion determines not merely the nature of the compound formed, but also the pro- portions in which combination takes place. His conception of a chemical compound is not associated with the constancy of proportions which had been assumed prior to his time. On the contrary, there exist, in his view, chemical compounds with all possible proportions ; 21 and it is only special reasons, such, for example, as considerable condensation on combination (that is, change in the cohesion of the constituents) which occasion constant proportions. Thus hydrogen only combines with oxygen in a definite proportion because water, the product of the combustion, is liquid, and the contraction which takes place presents too great an obstacle to the production of other compounds.22 But if, on combination, there is no change in cohesion, or only a slight change, then compounds are formed with variable proportions. As examples of these he mentions alloys, glasses, and solutions. He says that in these cases the limits are determined solely by the quantities required for mutual saturation, but that, between these limits, the most varying proportions occur.23 It will be observed that Berthollet thus classes solutions and alloys amongst compounds, and it will now be understood how he was able to distinguish amongst them some with vary- ing proportions. But it is much more remarkable that Ber- thollet also assumes variable proportions amongst the oxides. In one of his essays upon the laws of affinity,24 in which he speaks of metallic precipitations, he assumes, in accordance with his principles, that the two metals distribute themselves over the oxygen : there are thus formed, according to him, oxides con- 21 Stat. Chim. i, 373 ; E. I, 282. t£i Ibid. I, 367 ; E. I, 276. *» Jbid, I* 373-374 5 E- I, 282-283, 24 Ann, Chim. 37, 221. LECTURE III.] HISTORY OF CHEMISTRY. 41 taining different quantities of oxygen. He afterwards develops his views upon this point still more clearly.25 He says : " I have to show, then, that the proportions of oxygen in the oxides depend upon the same conditions as those in other compounds; that these proportions may vary progressively from the limit, at which combination becomes possible up to that at which it attains the highest degree." The limits them- selves he regards as determined by circumstances of cohesion. — In the same way he believes in the existence of salts with vary- ing quantities of base. If the base is precipitated, by means of an alkali, from a salt with an insoluble base, he supposes that a certain quantity of acid is precipitated along with the base, and that this quantity is variable.'26 In short, compounds with con- stant proportions are, according to Berthollet, exceptions, and the proportions in which substances combine are, as a rule, dependent upon the conditions of the experiment. If we summarise Berthollet's views once more, we may say that affinity appeared to him to be a force identical with gravity,27 whose phenomena are more varied only because it sets the molecules themselves into motion, and its effects are dependent upon the size and shape of the particles. In apply- ing these physical principles to chemical reactions, he arrives at the conception of chemical mass, which he defines as the product of affinity and quantity present. The chemical effects depend upon the magnitude of this chemical mass, and upon the cohesion of the substance, that is, upon its solubility and its greater or less volatility. This then leads him further to two general conclusions : — 1. Tables of affinity are useless, since in them affinity is assumed to be constant and independent of physical conditions, 2. There are compounds with varying, and progressively increasing proportions of their constituents. The first of these conclusions is adopted generally, and we find that the tables of affinity disappear soon after the appear- 23 Stat. Chim, 2, 370; E. 2, 316. 2(! Ibid. I, .86; E. I, 58. '•* Ibid, I, i ; E. x, vii 42 HISTORY OF CHEMISTRY. [LECTURE III. ance of Berthollet's Statique Chimique. The second, on the other hand, meets with vigorous contradiction ; Proust, a fellow-countryman of Berthollet's, especially opposing the views there stated. Thus there arises that renowned controversy between these savants which is remarkable not only for the talent which the opponents exhibit, but also for the extreme politeness which is observed on both sides. Berthollet was at that time held in high esteem by the scientific world. The sagacity which he had manifested in a high degree in the working out of his book was justly admired. It will be understood, then, that it was no small undertaking to attack views which he had stated with much confidence, and' had endeavoured to prove by experiments. At the same time I may state here that the experimental part, especially, of the Statique Chimique leaves much to be desired. When Berthollet asserts that in the oxidation of the metals oxygen compounds are formed with widely varying proportions, the reason of this is that he analysed the crude product directly, and did not first try to satisfy himself that he was not dealing (as was generally the case) with a mixture. If, in addition to this, the backward condition in which quantitative analysis stood at that time is taken into consideration, it will then be understood how Berthollet arrived at these erroneous results. Proust, on the contrary, proceeded very cautiously. He endeavoured to find proofs of the purity of his substances, and bestowed the greatest care upon the determination of the con- stituents. He thus succeeded in discovering the hydroxides, which had hitherto been entirely overlooked.28 and were re- garded as oxides containing a special proportion of oxygen. We are indebted to Proust for investigations of the most of the« metals, which he usually published under the title Faits pour servir a I'histoire, etc.29 Further, he wrote detailed treatises upon the sulphur and oxygen compounds,30 in which 28 Ann. Chim. 32, 41 ; Journ. de Phys. 59, 347. 24) Journ. de Phys. 51, 173 5 52, 409 ; 55, 325» <157 ; Ann. Chim. 32, 26 ; 38, 146 ; 60, 260, etc. 30 Journ. de Phys. 59, 321 ; Sulphur compounds, ibid, 53, 89 ; 54, 89 ; 59' 265- LECTURE III.] HISTORY OF CHEMISTRY. 43 he shows that many metals form only a single oxide ; that many, however, form two ; and that in those cases where three oxides exist, the intermediate one can be regarded as a com- pound of the other two.31 In the same way, he endeavours to detect the error in Berthollet's view regarding the existence of sulphur compounds with variable proportions of sulphur. In all these investigations he emphasises the distinction between mixtures and compounds. He says that the latter are charac- terised by quite definite proportions, which hold for compounds occurring in nature as well as for those obtained in the labora- tory, and that this pondus nature lies just as little within the discretion of the chemist as does the law of affinity, which governs all combinations.32 But Berthollet also responds by means of facts. He examines the carbonates of the alkali metals,33 and finds that when the base is saturated by carbonic anhydride under pressure, crys- tals are obtained which differ in composition from the carbonates previously known. He shows that these give up carbonic anhydride on dissolving and heating, and yield salts differing again in composition. He contests the fact asserted by Proust,34 that by leading a trace of carbonic anhydride into an alkaline solution only a few molecules are saturated, while the others remain uncombined. According to Berthollet, a solution of this kind evolves carbonic anhydride on the addition of a drop of hydrochloric acid, and consequently contains a souscar- bonate^ that is, he considers that the trace of carbonic anhy- dride present is distributed over the whole quantity of the base. Rendered cautious by Proust's rejoinders and excellent re- searches, Berthollet now no longer assumes all possible propor- tions between oxygen and the metals as actually occurring. He limits these to a few ; yet, in his examination of the oxides of lead, he asserts that he has isolated four different stages of oxidation which are attained by heating the metal in air.30 All the same he has thereby moved a step nearer to Proust. 31 Journ. de Phys. 59, 260. 32 Ann. Chim. 32, 31. 33 Journ. de Phys. 64, 168. 34 Ibid. 59, 329. 35 Ibid. 64, 181, 3« Ibid. 6l, 352- 44 HISTORY OF CHEMISTRY. [LECTURE III. Still the struggle is not on this account at an end. Even yet Berthollet will not recognise the difference between mixtures and compounds which had been advanced by Proust. For both of these conceptions he demands sharp definitions.37 Now, as a matter of fact, Proust cannot furnish these defi- nitions ; still he shows how mixtures can be distinguished from compounds in special cases, and in doing so, he succeeds in disproving a great many of Berthollet's statements. Obviously I cannot follow this matter in all its details, and I shall, therefore, only show by a single example Proust's mode of adducing his proofs. Berthollet had previously asserted that, by treating mercury with nitric acid, a series of oxides was obtained, in which the proportion of oxygen increased steadily from a defi- nite minimum.38 Further, he had observed that these oxides, on treatment with hydrochloric acid, were converted into two chlorides, and had assumed that it was the insolubility of mer- curous chloride which caused the oxides to betake themselves from the stage of oxidation at which they stood to the two end stations.39 Proust considers that too much intelligence is, on this assumption, attributed to the oxides. He shows that in the dry way also, only two chlorides are formed, and that these cor- respond to the only two oxygen compounds, of mercury into which Berthollet's mixtures can be separated. Thus this controvers}7, which began in 1801, continues until 1807, but, about this time, the interest that the scientific world had at first taken in the two opponents diminishes consider- ably. The authority of Berthollet had made it possible that, in consequence of his attack, a doctrine which had previously been regarded as accurate a priori, should appear doubtful "to many. But the researches of Proust on the one hand, and those of Klaproth and Vauquelin on the other, had restored confidence in that doctrine. Berthollet's rejoinders began to lose their effects. He was forced to admit the existence, in ever widening classes of substances, of compounds with con- 37 Journ. de Phys. 60, 347. 3S Arm. Chim. 38, 119. 39 Proust, Jourru de Phys. 59, 335. LECTURE III.] HISTORY OF CHEMISTRY. 45 stant proportions ; which, as a matter of fact, he had never wholly denied. In 1809 he still regards compounds with variable proportions as possible,40 but in this opinion he stands isolated. Too much support has now come to the opposing side. Richter's investigations, carried out from 1791 to 1800, had at length become known ; Gay-Lussac's classical work on the pro- portions by volume in which gases combine was concluded ; Berzelius had published his first important papers ; and Dalton had formulated his atomic theory, which was irreconcilable with Berthollet's view, and was already beginning to constitute the basis of chemical considerations. Thus the controversy ends, apparently with the complete defeat of Berthollet. I have treated this subject at length, because I consider it very important. We have here to do with a general doctrine which constitutes one of the foundations of our theoretical considerations, and settles the distinction between mixtures and compounds. It is for the latter alone that our chemical laws hold, since mixtures are not subject to them. In any particular case, therefore, it is very important to know which class of substances is being dealt with. What, then, are our means of forming an opinion ? It is to be found stated in text-books, that compounds possess a homogeneous character, whereas mixtures can very frequently be mechanically separated into their ingredients. It is further stated there that in compounds the properties of the constituents have disappeared, whilst they are present side by side in mixtures. Finally the constancy of proportions is then adduced as characteristic, and I wish to direct attention to this whole matter. There are cases in which mixtures are, in their whole behaviour, no longer distinguishable from com- pounds. We then have recourse to analysis to solve the ques- tion. We prepare the substance in various ways, and observe whether it always possesses the same composition. We thus invert the doctrine discussed by Berthollet and Proust. 'The former regarded compounds with variable proportions as pos- 40 Mem. d'Arcueil. 2, 470. 46 HISTORY OF CHEMISTRY. [LECTURE III. sible, whilst the latter assumed that substances combine in a few definite proportions only. We call a substance a com- pound when it contains its constituents in invariable pro- portions. I do not know whether the difference between the two con- ceptions has been made clear. In order to appreciate the full bearing of the question, a person must himself have required to decide as to whether he had a mixture or a compound in his hands. Even yet we are without a definition which shall suffice for every case ; that is, without such a definition as Berthollet repeatedly demanded from Proust. It is true that we have certain means of judging with respect to chemical compounds, as, for instance, capacity for crystallising, and in- variable melting point in the case of solids, and constant boiling point in the case of liquids. Yet these are frequently insufficient. I need only recall the phenomena of isomorphism, when we must admit that mixtures also can crystallise. I men- tion the solutions of hydrochloric acid, hydriodic acid, etc., in water, regarding which Roscoe has proved that they are only mixtures (solutions), and we must admit that these likewise can possess a constant boiling point. In short, this distinction forms one of the most difficult and most important problems, and in point of fact it is often insufficiently attended to. In the study of chemical papers opportunity is often afforded of observ- ing how errors have arisen through neglect of this very matter, How often have formulae been advanced for substances and theoretical conclusions based upon their existence, before their compound character has been conclusively settled ! The pur- pose of the foregoing remarks is to serve as a warning against any such error, and I therefore hope to be excused for having, for a short time, quitted my proper theme. LECTURE IV. RICHTER'S INVESTIGATIONS— DALTON'S ATOMIC THEORY — GAY-LUSSAC'S LAW OF VOLUMES— AVOGADRO'S HYPOTHESIS— WOLLASTON'S EQUIVALENTS. I SHALL now proceed to describe the development of the atomic theory, up to the second decade of this century, in so far as it possesses scientific interest. It is beyond my inten- tion, in doing so, to enter into the hypotheses of the Greek: and Roman philosophers regarding the constitution of matter. That Leucippus and Democritus regarded matter as composed of ultimate particles, and that Lucretius afterwards expounded these views at length, merely shows to us, what we have long known, that there were men amongst the Greeks and Romans who might in every respect be placed beside our thinkers of to-day. These philosophers made use of the deductive method of reasoning ; they started from general principles, although their conclusions from them were not always in accord with observations. The latter were of relatively small importance, especially as at that time experimentation, or the art of in- vestigation under stated conditions, was practically unknown. For this reason, Democritus cannot be placed in front of Kant, who, starting from the opposite view — from the dynamic hypothesis — constructed the universe in perhaps just as logical a fashion. The expenditure of talent and sagacity which we observe on the part of the supporters of the two views, was in vain ; the observations by means of which such questions could be solved, were entirely wanting. The scientific development of the atomic theory depended simply upon the discovery of a series of facts which were con- nected together by it, and found in it a simple explanation. It is my business now to mention these experiments, and to 48 HISTORY OF CHEMISTRY. [LECTURE IV. give an account of the chemical researches which rendered the assumption of the atomic theory necessary, and which were brought to a conclusion by means of the theory. In the preceding lecture I dealt with one of these regu- larities; that is, with the law of definite proportions. This alone, however, would not have sufficed for that theoretical conception, and another law was also necessary, namely, the law of multiple proportions. The latter was propounded by Dalton in 1804, that is to say, prior to the close of the con- troversy regarding the constancy of the proportions by weight in which substances combine. I have intentionally departed from the chronological sequence of the facts, in order to secure their logical arrangement. The law of multiple proportions has no meaning, so long as the law of constant proportions is not proved. It includes the latter within it, and can only stand along with it. We may well be surprised that it should have been propounded at the very time when doubts were enter- tained as to the accuracy of the law of constant proportions. The explanation probably lies in the fact that Berthollet and Proust lived in France, whereas Dalton made his discovery in England, and withheld the publication of his investigations until 1808; whilst prior to that date the scientific world only obtained a short statement of the results of Dalton's experi- ments from Thomson's System of Chemistry, in which the investigations are mentioned. Dalton's theory, which very soon gained favour, undoubtedly exercised a decisive influence on the views of the chemists of the period with respect to con- stant proportions; and it is partly to be attributed to the labours of Richter, of Dalton, and of Wollaston, that Berthollet, if he did not exactly withdraw his previous assertions, at least did not further endeavour to bring about their acceptance. It may be that Dalton, as his biographer, Smith, asserts,1 had no knowledge, or only a very incomplete knowledge, of the work of Richter, which might otherwise have contri- buted materially to the establishment of the atomic theory. 1 Memoir of John Dalton and History of the Atomic Theory, 214. LECTURE IV.] HISTORY OF CHEMISTRY. 49 He may have arrived quite independently at those ideas which exercised the greatest influence upon the subsequent develop- ment of chemistry, but we must here take notice of all the facts which are of importance in this connection, and we must not overlook Dalton's predecessors. Almost simultaneously with the conception of the atom, that of the equivalent was also de- veloped. The recognition of the latter contributed to procure admission and general favour for the atomic theory, and it is therefore advisable, in my opinion, to deal with both ideas side by side as they arose chronologically. In doing so, it will be observed that Dalton's atom was introduced independently of the equivalent, but that Wollaston, in particular, endeavoured to replace the atom by the equivalent. This afterwards led to the identification of the two ideas, a fact which had a detri- mental effect upon the science. The first experiments which could have led to the establish- ment of equivalent quantities, were carried out by Bergman in the second half of last century.2 Bergman observed that neutral solutions of metals were precipitated by other metals, without the production of an acid reaction and without the evolution of gas. x\s an adherent of the phlogiston theory, he explains the observations quite correctly in accordance with the principles of the theory. He assumes that the precipitated metal has taken up just as much phlogiston as the precipitating metal has parted with; and in consequence of this view, he perceives a means of determining the quantities of phlogiston contained in different metals. The quantities of the dissolved and of the precipitated metals respectively, must stand to each other inversely as the supposed quantities of phlogiston con- tained in equal weights of these metals. Lavoisier, who repeats and extends Bergman's experiments a few years afterwards,3 recognises that they must show, in terms of his theory, the quantities of oxygen which combine with equal weights of the metals. Where Bergman had spoken 2 Bergman, Physical and Chemical Essays. Translated by E, Cullen, M.D., 2 (1784), 349 et seq, 3 Lavoisier, Oeuvres. 2, 528. D 50 HISTORY OF CHEMISTRY. [LECTURE IV. of the taking up of phlogiston, Lavoisier only needs to assume the giving up of oxygen, and vice versa : the inverse ratio of the quantities of the precipitated metal A and of the dissolved one B gives Lavoisier the relations of the quantities of oxygen com- bining with equal weights of A and B ; or otherwise, and more clearly expressed, the quantities of the precipitated and of the dissolved metal respectively, which the experiment furnishes directly, have the property of being able to combine with the same quantity of oxygen. The generalisation, in the latter form, was not emphasised, however, either by Bergman or by Lavoisier, otherwise it would probably have led to the notion of equivalence. This result did not follow, and even the ex- periments of both these chemists were but little heeded. It did not fare much better with the investigations of Richter, which were carried out between 1791 and 1802, and were sup- ported by far more observations. Richter was the first to state the law of neutrality, and to deduce accurate conclusions from it.4 This merit was formerly attributed erroneously to Wenzel, who arrived, however, at exactly the opposite result. The error which passed into many of the older text-books appears to have been caused by Berzelius,5 and was pointed out by Smith,6 and others.7 Richter observed that upon mixing solutions of two neutral salts, the neutrality is maintained, even when double decom- position takes place, and- from this he concluded that the quantities a and b of two bases, both of which are neutralised by the same quantity c of an acid, are both likewise neutralised by the same quantity d of another acid ; and, conversely, that the weights of two acids which are neutralised by the same quantity a of a base, require for neutralisation the same quantity 4 Richter, Ueber die neueren Gegenstande der Chemie. 5 Berzelius, Essai sur la theorie des proportions chimiques et sur 1'influence chimique de 1'Electricite, Paris (1819), 2. 6 Memoir of John Dalton, etc., 160. 7 Wenzel determined the proportions in which base and acid combine to form salts. He found, however, exactly the opposite of what Berzelius makes him say. Compare Wenzel, Ueber die Verwandtschaft der Korper, Dresden (1782), especially 450 et seq. LECTURE IV.] HISTORY OF CHEMISTRY. 51 b of another base. The mode of expressing this which Richter employs is remarkable. He says : 8 — " If P is the mass of a determining element, where the masses of the elements determined by it are a, b, c, d, e, etc., and <2 is the mass of another determining element, where the masses of the elements determined by it are a, ft, y, 8, e, and so on, but where a and a, b and ft c and y, d and S, e and e, always represent one element, and the neutral masses P+a and Q + P ; P+ b and Q + y ; P+ c and Q + a, etc., decompose by double affinity in such a way that the products obtained are again neutral, then the masses a, b, c, d, e, and so on, have ex- actly the same quantitative ratio amongst themselves as the masses a, ft y, 8, e, and conversely." I must observe that by the determining element, and the element determined, Richter understood the quantities of acid and of base that mutually neutralise each other. Richter well understood the importance of the foregoing statement. He remarks :9 "This rule is a true touchstone of the experiments instituted with regard to the relations of neutrality ; for, if the proportions ascertained empirically are not of the nature that the law of decomposition by double affinity requires, where the decomposition actually taking place is accompanied by un- altered neutrality, they are to be discarded without further ex- amination, since an error has then occurred in the experiments tried." Richter tabulated the quantities of the bases which are neutralised by the same weight of sulphuric acid, of hydro- fluoric acid, etc., and these he calls neutrality series, or series of masses ; 10 he also determined the quantities of the acids which are saturated by the same quantity of different bases.11 In doing this, he thought he had discovered certain regularities, but this subsequently proved to be fallacious. Thus, accord- ing to him, the series of masses in the case of the bases formed an arithmetical, and that in the case of the acids, a geometrical 8 Neuere Gegenstande (1795). 2, 66, 9 Ibid. 2, 69. 10 Ibid, 2, 70, n Ibid. 2, 92; 3, 176. 52 HISTORY OF CHEMISTRY. [LECTURE IV. series. He wished, in fact, to establish a regularity in chemical compounds, such as had been assumed in the distances of the planets from the sun, and, in order to accomplish this, he had probably corrected many of his results. Another department of Richter's " stochiometric investiga- tions " must be mentioned here ; that is, his work upon metallic precipitations. He determines the quantities of the metals as they mutually precipitate one another from their solutions, and employs the numbers obtained to ascertain the proportions of oxygen in the oxides. Here again, his way of expressing him- self leaves much to be desired. He says : 12 — " When an aqueous solution of a metallic neutral salt is so decomposed by an inflammable metallic substrate, that is, by another metal in the metallic state, that not only does the metal which was dissolved separate out in a wholly metallic state, but also that neither the dissolving acid solvent nor the water associated with it, is decomposed, then the masses of the vital air which must unite with equal masses of the metallic substrates, in order to make their solution in acids possible, are inversely proportional to the masses (or weights) of the separat- ing and of the separated metallic substrates from the metallic neutral salts." And at another place:13 — "The quantitative order of the specific neutrality of the metals towards vitriolic acid does not by any means follow the usual order in which one metal is separated -by another from the solution in the acid ; on the contrary, it is wholly analogous to the inverse quantitative order of the removal of the inflammable matter and of the respective combination with vital air." It is worthy of mention that Richter introduced the name Stochiometry, which signifies the measurement of the pro- portions in which substances combine. Fischer, on the other hand, deserves credit for having combined Richter's various tables into a single one. In this connection, he expresses himself as follows : 14 — " It is only 12 Neuere Gegenstande. 3, 83. 13 Ibid. 3, 127. 14 Berthollet, Statique Chimique. German translation with explanatory comments by Fischer, I, 135. LECTURE IV.] HISTORY OF CHEMISTRY. 53 necessary to determine the proportional quantities of one acid against the different alkaline bases ; afterwards, it is sufficient to ascertain the proportional quantity of a single compound of every other acid with one alkaline basis, when, by means of an easy calculation, the proportional quantities of the acids in all the other compounds may be determined." It may be said, indeed, that Fischer's table was the first table of equivalents. The numbers attached in it to the different bases, represent equivalent quantities, since these quantities are neutralised by the same quantity of acid ; and conversely with respect to the different acids. The conception of the equivalent was, in this way, established about the year 1803, although the word itself was not in use at that time. The discovery of the law of multiple proportions, and the formulation of the atomic theory by Dalton (which were first announced in the third edition of Thomson's System of Chemistry), both took place almost at this same time. I wish to deal, in a few words, with the questions of priority which were called forth by these important experiments and views.15 The idea of the atomistic point of view is an old one, and at the beginning of this lecture I have named a few Greek philosophers who advanced and upheld the theory. The opinions pass down through all the centuries ; they are con- tested ; but they always find supporters. The chemists of the eighteenth century appear to have embraced pretty generally the atomistic way of regarding matter. I may here adduce the views of Lavoisier with respect to the constitution of matter, which I have already stated at length ; and those of Berthollet, who frequently speaks of molecules. In one word, these were the opinions of the day, and they were preferred to the dynamic hypothesis, chiefly, it is almost certain, because the assumption of discreet particles of matter, separate from one another, fur- nished a simple explanation of the diminution of volume with the lowering of temperature. Dalton, on his part, did not claim that he had introduced 15 Compare the work by Smith, already mentioned. HISTORY OF CHEMISTRY. [LECTURE IV. these opinions into the science. In this connection he says, that the observation of the existence of different states of aggregation, has " led to the conclusion which seems univer- sally adopted, that all bodies of sensible magnitude, whether liquid or solid, are constituted of a vast number of extremely small particles, or atoms of matter bound together by a force of attraction, which is more or less powerful according to cir- cumstances, and which as it endeavours to prevent their separation, is very properly called in that view, attraction of cohesion"^ And with regard to the "agency of heat" he says: — "An atmosphere of this subtile fluid constantly sur- rounds the atoms of all bodies, and prevents them from being drawn into actual contact." 17 Dalton shows, however, in the course of his most interest- ing work how the relative weights of these particles can be ascertained ; and it will remain as his imperishable merit to have shown the possibility of determining the atomic weights. Higgins, it is true, tried to prove that he also participated in this important discovery ; 18 but even if it must be admitted that Higgins employed the atomic theory as early as lySg,19 still his way of expressing himself is not, by a long way, so clear and definite as Dalton's, and, so far as I know, he does not mention atomic weights. Dalton turned the atomic hypothesis to account as the basis of chemical considerations, after he had found that when two substances unite in several proportions, these proportions are always expressible in simple multiples by whole numbers. He examined the two hydrocarbons, marsh gas and ethylene, and found that, for the same weight of hydrogen, there was twice as much carbon combined with it in ethylene as in marsh gas. He then examined to see whether any such regularity was to be found in other compounds, employing in particular for this purpose the oxides of nitrogen, and he thereby ob- 16 Dalton, A new System of Chemical Philosophy, I, 141-142 ; Alembic Club Reprints. 2, 27. 17 New System. I, 143-144. 18 Experiments and Observations on the Atomic Theory, by W. Higgins (1814). 19 A Comparative View of the Phlogistic and Antiphlogistic Theories. LECTURE IV.] HISTORY OF CHEMISTRY. 55 tained confirmation of the law. The law is to this effect :— If two substances, A and B, form several compounds, of which the compositions are all calculated with respect to the same quantity of A, then the quantities of B combined with this, stand to each other in a simple ratio.20 Dalton sought in the atomic theory, an explanation of this law, which was simply an expression of the observed facts. According to the atomic theory, chemical compounds are formed by the arrangement, in juxtaposition, of atoms of the elements, these latter being incapable of undergoing any further decomposition. With regard to this Dalton says : 21 — " Chemi- cal analysis and synthesis go no farther than to the separation of particles one from another, and to their reunion. No new creation or destruction of matter is within the reach of chemi- cal agency." By the fact that Dalton assigns a definite unalter- able weight to the atom of every element, and admits the possibility of the combination of several atoms, his theory is brought into harmony with experiment, and becomes, indeed, a necessary consequence of it. According to the number of atoms which enter into combination, the resulting atom may belong to a different order. The atoms of elements are simple atoms, or atoms of the first order. When i atom of an element A combines with i atom of an element B^ i atom of the second order is produced. When 2 atoms of an element A combine with i atom of an element B, i atom of the third order is produced. When i atom of an element A combines with 2 atoms of an element B, i atom of the third order is produced. When i atom of an element A combines with 3 atoms of an element B, i atom of the fourth order is produced. When 3 atoms of an element A combine with i atom of an element B, i atom of the fourth order is produced, etc. 20 It appears that Dalton never stated the law in this general form. Compare Memoirs of John Dalton, by W. Henry, 79 et seq. 21 New System. I, 212 ; A.C.R. 2, 29. 56 HISTORY OF CHEMISTRY. [LECTURE IV. I have not been able to find any statement as to whether two atoms of one element can combine with three atoms of another, but it appears as if Dalton regarded this assumption as untenable. Compounds which are most simply regarded in this way, consist, according to him, of two composite atoms ; he is obliged, of course, to make the assumption that atoms of the higher orders are capable of combination with one another.22 I have pointed out, above, that Dalton's theory was in agreement with the facts ; I shall now explain how, from his experiments, he determined the atomic weights. In order that he might be able to do this, the first thing necessary was to settle the number of atoms in a compound. According to Dalton, this number is to be sought for, in general, in the simplest possible ratios. In estimating it, he starts from the following principles : 23 — 1. When only one compound of two elements is known, this is composed of an atom of the second order. 2. When two compounds are known, the one consists of an atom of the second, and the other of an atom of the third order. 3. When three compounds are known, one atom of the second and two atoms of the third order must, be assumed. How does Dalton now proceed to the determination of the atomic weights, i.e., the. relative weights of the smallest par- ticles? In the first place, he requires to choose a unit for comparison. As unit he assumes hydrogen with the atomic weight = i, and he refers all the other atomic weights to this. To fix the others, he then applies his first principle. At that time, only one compound each of oxygen and of nitrogen with hydrogen was known, viz., water and ammonia respectively ; therefore the atomic weights of oxygen and nitrogen can be de- termined directly from the composition of these compounds. In this way, Dalton finds them to be 5 and 7 respectively. He checks the numbers so obtained by the proportions of the 22 New System. I, 213-215 ; A.C. R. 2, 30-31. <23 New System. I, 214; A.C.R. 2, 30. LECTURE iv.] HISTORY OF CHEMISTRY. 57 oxygen and nitrogen in the oxygen compounds of nitrogen.24 He is acquainted with four of the latter. In nitric oxide he finds 7 parts of oxygen for 5 of nitrogen ; its atom is, therefore, the atom of the second order, derived from these elements. In nitric acid, according to his view, there are 14 parts of oxygen for 5 of nitrogen, or two atoms of the former gas for one of the latter. In nitrous oxide, 7 parts of oxygen are com- bined with 10 parts of nitrogen, and in this he therefore assumes two atoms of nitrogen and one of oxygen. Nitrous acid, however, is supposed to contain loj parts of oxygen for 5 of nitrogen, and in it he might have assumed two atoms of nitrogen and three of oxygen. He prefers, however, to regard this substance as a compound of nitric acid and nitric oxide. Further, he finds in ethylene 5.4 parts of carbon for i of hydrogen, and in marsh gas the same quantity of carbon for 2 of hydrogen. On this account, he regards ethylene as consist- ing of atoms of the second order, and assumes the atomic weight of carbon to be 5.4. Carbonic oxide likewise consists of atoms of the second order, since he finds in it 7 parts of oxygen for 5.4 of carbon, while carbonic anhydride has atoms of the third order, because it contains 14 parts of oxygen for 5.4 of carbon. But Dalton does not always adhere quite rigidly to his own rules. Thus he regards sulphuretted hydrogen as consisting of one atom of sulphur and three of hydrogen, and sulphuric acid, of one atom of sulphur and three of oxygen, whereby he is led to 13 as the atomic weight of sulphur. Hence there is, in my opinion, something haphazard in these atomic weight determinations, quite apart from the rules themselves. I shall return afterwards to the question whether the latter are justifiable or not. The numbers advanced by Dalton were thus relative in a twofold manner, if I may so ex- press myself; they were affected by two unknown constants. In the first place they were all determined with reference to an arbitrary standard, and in the second, they were only relatively 24 New System. I, 215 ; A.C.R. 2. 30-31. 58 HISTORY OF CHEMISTRY. [LECTURE IV. accurate with respect to this standard. The atomic weight of carbon had in reality only been found to be a multiple or a submultiple of 5.4. Dalton himself does not appear to have been aware of this arbitrary character of his numbers. In spite of this, his theory met with very general re- cognition, and chemists were astonished at the simplicity with which it explained all the regularities which had been dis- covered immediately before. With the rapid progress that the science was at that time making, some new support was necessary. In order to avoid being left behind it was essential to possess a general point of view from which the isolated facts and the different regularities could be conveniently surveyed. It was soon to be shown that this theory was capable of stand- ing the test ; that it was not only sufficient to connect the known phenomena, but that laws which were only discovered subsequently could also be explained by means of it. This holds especially for the law of gaseous volumes, which Gay- Lussac discovered in 1808, a few months after the appearance of Dalton's ingenious book. As early as 1805, Humboldt and Gay-Lussac, on the occasion of their joint investigation of the composition of the air, had determined anew the proportions by volume in which hydrogen and oxygen combine.25 They found slight differ- ences from the earlier observations, and arrived at the highly interesting result that water is produced by the condensation of 2 volumes of hydrogen and i of oxygen ; while Meusnier and Lavoisier26 had found 23 volumes of hydrogen and 12 of oxy- gen, and Fourcroy, Vauquelin, and Seguin27 had found 205.2 of hydrogen for 100 of oxygen. Three years later, Gay-Lussac extended his experiments to other gases.28 He had previously observed the law (often called after him) of the uniform expansion of gases with increase of temperature.29 He was also familiar with the so-called law of 23 Journ. de Phys. 60, 129. 26 Lavoisier, Oeuvres. 2, 360. ^ Ann. Chim. 8, 230 ; 9, 30. w Mem. d'Arcueil. 2, 207. 29 Ann. Chim. 43, 137- LECTURE IV.] HISTORY OF CHEMISTRY. 59 Mariotte, which was discovered by Boyle, and which states the relation between pressure and volume ; in short, he possessed all the data for reducing the results directly obtained to like pressure and temperature — a basis upon which to execute the experiments he had in view. His investigation is a model of accuracy, and, in this respect, it is very markedly distinguished from other experimental investigations of the period. The re- sults obtained are extremely simple ; and he states them some- what after this manner. Two gases always combine in simple proportions by volume, and the contraction which they under- go, and, therefore, also the volume of the product formed, stand in the simplest relation to the volumes of the con- stituents. Thus Gay-Lussac found, for example, that 2 volumes of carbonic anhydride are produced from 2 volumes of carbonic oxide and i of oxygen ; that 2 volumes of nitrous oxide are composed of 2 volumes of nitrogen and i of oxygen ; that equal volumes of nitrogen and of oxygen are combined in nitric oxide, while the product has the same volume as the two con- stituent gases separately ; and finally, that i volume of nitrogen and 3 of hydrogen are condensed to 2 volumes in ammonia. Gay-Lussac, who was well acquainted with Dalton's theory, shows at the end of his paper that the facts ascertained by him are in harmony with the theory ; that, by the assumption of a similar molecular condition in all gases, it will explain their analogous behaviour towards changes of pressure and of tem- perature ; and that his law of gaseous volumes is an important support of Dalton's view. It might be supposed that the latter would have been highly pleased by so unexpectedly brilliant a confirmation of his views. This was not so, however. In the second part of his "New System of Chemical Philosophy," which appeared in 1810, he practically regards Gay-Lussac's experiments as erroneous. I shall endeavour to explain the reasons that prompted him to do this, especially as it has been stated that Dalton wished, from jealousy or want of judgment, to contest Gay-Lussac's merits. In the first part of his book, Dalton had already speculated 60 HISTORY OF CHEMISTRY. [LECTURE IV. as to the volume relations of the gases. He says there : — 30 " In prosecuting my enquiries into the nature of elastic fluids, I soon perceived it was necessary, if possible, to ascertain whether the atoms or ultimate particles of the different gases are of the same size or volume in like circumstances of tem- perature and pressure. By the size or volume of an ultimate particle, I mean in this place, the space it occupies in the state of a pure elastic fluid ; in this sense the bulk of the particle signifies the bulk of the supposed impenetrable nucleus, to- gether with that of its surrounding repulsive atmosphere of heat. At the time I formed the theory of mixed gases, I had a confused idea, as many have, I suppose, at this time, that the particles of elastic fluids are all of the same size ; that a given volume of oxygenous gas contains just as many particles as the same volume of hydrogenous." He afterwards became of a different opinion, to which he was led by the following con- siderations : — One atom of nitric oxide consists of one atom of nitrogen and one of oxygen. If, now, there were the same number of atoms in equal volumes, one volume of nitric oxide should be formed by the combination of one volume of nitro- gen with one of oxygen, but, according to Henry's experiments, about two volumes are produced ; hence nitric oxide could only contain half as many atoms in the same space as nitrogen or oxygen.31 Dalton, in his reply, refers to these considerations, and then, with regard to Gay-Lussac's " hypothesis that all elastic fluids combine in equal measures, or in measures that have some simple relation one to another," he proceeds :32 — " In fact, his. notion of measures is analogous to mine of atoms ; and if it could be proved that all elastic fluids have the same number of atoms in the same volume, or numbers that are as i, 2, 3, etc., the two hypotheses would be the same, except that mine is universal, and his applies only to elastic fluids. Gay-Lussac could not but see that a similar hypothesis had been entertained by me, and abandoned as untenable." 30 New System. I, 187-188; A.C.R. 4, 6, 7. 31 Compare New System. I, 70-71 ; A.C.R. 4, 5. ™ New System. 2, 556 ; A.C.R. 4, 25. LECTURE IV.] HISTORY OF CHEMISTRY. 6l Dalton shows, moreover, the poor agreement between the results of Gay-Lussac and those of Henry, and this confirms him in his conclusion that the former had not observed accurately. It cannot be denied that Dalton's contention is well founded. If the atomic theory was chosen as the basis for chemical speculation, the law of gaseous volumes, as Gay- Lussac stated it, could only be brought into harmony with it if the assumption were admissible that equal numbers of the smallest particles are present in the same volumes of all gases. This assumption agreed with the physical properties of gases, but was, as Dalton quite rightly concluded from known facts, impossible. The three rules adopted by Dalton likewise told against the accuracy of this hypothesis. In water, for example, it would have been necessary to assume two atoms of hydrogen for one of oxygen ; and it may be that this was also a reason for Dalton's coming forward so decidedly in opposition to Gay- Lussac. It will be clear from these explanations that there was a real difficulty in bringing Gay-Lussac's law into harmony with the atomic theory. Avogadro was the first to show how this difficulty can be got over.33 The Italian physicist distinguishes molecules integrantes and molecules elementaires, which,, for brevity and simplicity, we shall translate by molecule and atom respectively. The physical properties of the gases (especially the similarity in their behaviour towards changes of pressure and of temperature) lead Avogadro to assume in equal volumes of all gases, the same number of molecules ; and the distances of the latter from one another he considers to be so great in proportion to their masses, that they no longer exercise any attraction upon one another. These molecules are not supposed, however, to constitute the ultimate particles of matter, but are assumed to be capable of further subdivision under the influence of chemical forces. According to Avogadro, therefore, sub- stances (elements and compounds alike) are not converted, in passing into the gaseous state, into indivisible particles, but 33 Journ. de Phys. 73, 58 ; A.C.R. 4, 28, 62 HISTORY OF CHEMISTRY. [LECTURE IV. only into molecules integrantes, which in turn are composed of molecules elementaires. He bases his view upon the following considerations : — If nitric oxide, which aiises without contrac- tion from equal volumes of nitrogen and oxygen, contains as many molecules as the mixed gases do, then the combination cannot consist in the union with one another of previously separate molecules, since this would necessarily involve a diminution in the number of particles ; but it must be brought about by an exchange. The molecules both of nitrogen and of oxygen must split into two parts, and these then combine by mutual exchange. While, therefore, before the combination, the gaseous mixture consists of dissimilar molecules, of which the one half are composed of two atoms of nitrogen, and the other half of two atoms of oxygen, the product of the combination consists of the same number of molecules, but they are all like molecules and each has been formed by the union of one atom of nitrogen with one of oxygen. Consideration of the volume relations in the formation of ammonia, likewise points to a subdivision of the gas particles of elementary substances. All these discus- sions assume the simplest character when the word molecule (molecule integrante) is employed instead of volume, the two conceptions being identical for the gaseous state, according to Avogadro's definition. From Gay-Lussac's numbers, then, it appears that the molecule of ammonia consists of half a mole- cule of nitrogen and one and a half molecules of hydrogen ; that the molecule of water contains half a molecule of oxygen and one molecule of hydrogen, etc. If the simplest hypothesis with respect to the divisibility of the molecule is adopted, so that it is not necessary to introduce fractions of atoms, then the molecules, not only of hydrogen, but also of oxygen and of nitrogen, must consist of two elementary atoms ; and the proportions, by volume, in which the gases combine, then give the number of chemical smallest particles which go to form the molecule. Avogadro finds, for example, that two atoms of hydrogen and one of oxygen are necessary for the formation of water ; that three atoms of the former gas and one of nitrogen LECTURE IV.] HISTORY OF CHEMISTRY. 63 are present in ammonia, etc., and he thus arrives at results quite different from those of Dalton. He points this out explicitly in his paper, and draws attention to the fact that in his determinations, he starts from legitimate physical principles, whereas Dalton's rules contained arbitrary assumptions. He lays stress on the fact that Dalton, in case he wishes to identify the physical with the chemical atoms (molecules int'egr antes and elementaires\ will be forced to assume that in those combinations which take place without contraction, the composite atoms must be further removed from one another than the uncombined ones. Avogadro is able, from their densities, to determine directly the molecular weights of the elements known in the gaseous state. This, however, is not sufficient for him, and he also attempts the determination in the cases of other elements. Here, however, he has recourse to more or less doubtful hypotheses. He finds the atomic weight of carbon to be 11.3, and that of sulphur 31.3, referred to that of hydrogen assumed = i ; that is, he finds numbers which very nearly agree with those adopted at present. I shall not enter into the more minute details of this most interesting paper, and shall only remark further that Avogadro admits the possibility of molecules of elements consisting of 4, 8, etc., atoms, and believes that nature has, in this very way, equalised the difference between simple and compound substances. Starting from similar views, Ampere writes a paper on the same subject three years later (i8i4).34 His conclusions are, however, less simple, as he tries to explain at the same time the crystalline form of substances, by the position of the atoms in the molecule. These speculations met, on the whole, with but little attention in the chemical world. It seems as if a distinction between atom and molecule was not regarded as justifiable, and accordingly neither Avogadro's nor Ampere's ideas exercised any immediate influence upon the science. This may also be a4 Ann. Chim. 90, 43. 64 HISTORY OF CHEMISTRY. [LECTURE IV. explained by the fact that the hypothesis only led to decisive results as to the number of atoms contained in the molecule (and thus to the determination of the atomic weight) in the case of gaseous substances, and was not applicable to solids and liquids. Chemists, therefore, looked for new generalisa- tions, and the next impulse in this direction was given by Wollaston. Wollaston had carried out an investigation on the carbonates, in i8o8,35 which appeared simultaneously with an examination of the oxalates by Thomson.36 It was shown in the papers of these chemists that carbonic acid can form compounds with one and with two parts, and oxalic acid with one, with two, and with three parts of potash. These experiments produced a great impression, because at that time there were few facts of this kind known which had been minutely examined ; on this account, they formed an important support to the law of multiple proportions. But if Wollaston, on the one hand, thus exerted an influence upon the rapid recognition which the atomic theory met with, and consequently came to be regarded even by authorities as an adherent of the theory,37 still, by a later paper,38 he contributed to the abandonment of the atom by a section of chemists, as too indefinite a basis for chemical considerations. In 1814, Wollaston, not without justice, represents to Dalton how uncertain and arbitrary is his estimation of the number of atoms in a compound ; and how, in consequence, the atomic weights are wholly hypothetical numbers, so that, in his opinion, they should not be adopted. He advised, instead of the con- ception of the atom, the introduction of the equivalent, which word he employs for the first time. Wollaston was well acquainted with Richter's works,39 and he derives the concep- tion of the equivalent principally from his investigations. I must at once remark, further, that, with him, not only are those 35 Phil. Trans. 1808, 96 ; A.C.R. 2, 34. 36 Phil. Trans. 1808, 63 ; A.C.R. 2, 41. 37 Kopp, Geschichte. 2, 373. ** Phil. Trans. 1814, i ; Ann. Chim. 90, 138. ^ Even Wollaston remarks (loc. cit.) that Wenzel's analyses do not agree with the law of neutrality. LECTURE IV.] HISTORY OF CHEMISTRY. 65 quantities of two bases equivalent which are neutralised by the same quantity of acid, and those quantities of metal equivalent which mutually precipitate each other (and therefore unite with the same weight of oxygen), but that also, in his determinations, he extends far beyond these limits, without, it would appear, ever clearly perceiving that he falls into exactly the same error that he points out to Dalton. I go further, even, and assert that the uncertainty was increased by him, since he first employed the equivalent in the sense of the atom, and thereby attached to it the vague signification which remained connected with it ; and so it was this very paper which principally led chemists to the fusion of the two conceptions, and thereby to the tacit and inaccurate assumption that the atoms were equivalent, an error which gave rise to great confusion. I shall here give an example of Wollaston's determinations, so that the reader may obtain at least an idea of his method, and may satisfy himself as to the accuracy of my opinion. Wollaston sets out from the equivalent of oxygen, which he assumes = 10; from this he determines the equivalent of hydrogen to be 1.3, clearly because 1.3 parts of hydrogen (according to the estimations of the period) unite with 10 parts of oxygen to form water; thus equivalent quantities are, with Wollaston, the quantities in which substances combine. But how, we may ask, does he proceed in the cases of substances that combine in more than one proportion — in the case of carbon, for example? Does he recognise several equivalents^ here ? The answer is, no, he never appears to think that such a thing can be possible. He adopts the equivalent of carbon as 7.5, determining it from that of carbonic anhydride, which, according to him, is 27.5. He does not give any reason, however, for choosing the latter number, and we are left to find it out for ourselves. We might suppose that Wollaston regarded as the equivalent that quantity of carbonic acid which saturates a quantity of a base containing 10 parts of oxygen, whereby he would necessarily have adhered to the view stated above, that the combining weight is identical with the equiva- lent. He is obliged, however, as a consequence from his own E 66 HISTORY OF CHEMISTRY. [LECTURE IV. figures, to assume in carbonic acid, two equivalents of oxygen for one of carbon. In this case, therefore, the combining weight and the equivalent were no longer identical, but the one was twice as great as the other. That such results were unavoidable in his method, ought to have been clear to him at once, since he was acquainted with the law of multiple propor- tions. Wollaston does not, however, pause in the least on account of this result ; he does not bestow a word upon it ; and he continues his determinations of equivalents uncon- cernedly, but we shall not follow him in these, as they possess no further interest for us. I think I have now shown that Wollaston's equivalents present the same uncertainties as Dalton's atomic weights, and that the views stated by him must be called retrograde because he believed that he was dealing only with real unambiguous conceptions, free from all hypotheses. It may perhaps seem that the opinion stated here is severe and unjust, but when the further development of chemistry is followed up, and it is seen how a more rapid progress during the succeeding decade was prevented by this very confusion of equivalent and. atom (er combining weight), and how a most vigorous struggle was necessary before the separation of the conceptions could be re-introduced, the reader will probably adopt my view. It is true that the blame does not rest with Wollaston alone, because a school arose in Germany also, probably stimulated by him, which represented the same ideas. This school was at first, no doubt, dominated by the great influence of Berzelius, but afterwards, especially at the beginning of the fifth decade of this century, it took the lead itself. I do not intend to withhold the details of these highly interesting developments, but in the next lecture I must direct attention to the electrical phenomena which at this time begin to exercise a great influence upon our science. LECTURE V. DAVY'S ELECTRO-CHEMICAL THEORY— DISCOVERY OF THE ALKALI METALS — DISCUSSION REGARDING THEIR CONSTITUTION — DOES HYDROCHLORIC ACID CONTAIN OXYGEN? — HYDROGEN THEORY OF ACIDS. IF we look back to the period when the alkalies were regarded as simple and undecomposable substances, we can easily understand the enthusiasm with which the chemical world greeted the discovery of potassium and sodium. We are all familiar with the remarkable properties of these elements — with their metallic appearance and low specific gravity, their change- ability in the air, their easy inflammability upon water, etc. It can be understood, therefore, that when substances possessing such properties had once been seen, illusions of all kinds were entertained by some people. The idea was arrived at that the substances hitherto known were only compounds, and that the aim of chemistry was now to discover the true elements, which, it was supposed, would resemble potassium and sodium. It is likewise comprehensible why the agency which had accomplished such results should be admired and overestimated. Everything was supposed to be possible by means of it, and the direction which chemistry would necessarily follow — that of electro- chemical development — was clear to every one. The galvanic current, at that period an entirely new agent, had accomplished this marvel, and it was itself a marvellous thing. By its aid it had become possible to decompose compounds into their true elements ; hence it is not surprising that this agency was re- garded as identical with the one which gave rise to combinations, i.e., with affinity. It was believed that an explanation was thereby furnished for two things, both of which stood in need of it — that is to say, for electrical and for chemical phenomena, 68 HISTORY OF CHEMISTRY. [LECTURE V. and the connection between these two was now shown. The further development of the electro-chemical theories appeared at that time to be the highest aim of our science; at a later date we see these theories abandoned. The extraordinary enthusiasm was succeeded by an indifference just as extra- ordinary. In those cases where it was formerly believed (as regards the structure of the most complicated compounds) that the real secrets of nature were being discovered, the ordinary phenomena of decomposition were afterwards recognised. A secondary importance only, was usually ascribed to the latter in determining the constitution of substances. The electrical properties of substances, which at that time also indicated their position in the system, afterwards became of less consequence in the determination of their chemical character. This is just such an example as is met with in the history of every science. Some great end is achieved, in comparison with which everything else becomes dwarfed into insignificance. Every endeavour is turned towards development in the new direction, and a system is established which has the observed phenomena for its basis. Then facts appear which are in conflict with the views- that have thus arisen. These facts are a sufficient ground for some to abandon the theory ; but for others they act merely as a stimulus to bring the new experi- ments into harmony with the theory, and this compels them to have recourse to further hypotheses. In this way a controversy is developed which only ends when the supporters of the older view have their eyes opened so as to recognise how greatly the originally simple and elegant theory has been disfigured. The whole system now falls to pieces, and it is no longer compre- hensible how any one could have permitted himself to be guided by such views. The greatest folly is now perceived in what had been previously recognised as the highest wisdom. So the times change ! The electro-chemical theories had the same fate. If we consider the astonishing discoveries which were made by means of the galvanic current, we can understand the importance which attached at that time to electrical phenomena ; and I LECTURE V.] HISTORY OF CHEMISTRY. 69 believe I may almost assume that we would do similarly if we were placed in the same position. Listen, and judge ! In 1789 Galvani1 discovered that upon simultaneously touching a muscle and a nerve of a frog, by means of two different metals which were joined together by a conductor, the frog was convulsed. Galvani's explanation of this fact was controverted and replaced by another, in 1792, by Volta.2 We find the question of the cause of the electrical current discussed for a long period. Is the contact of the two metals sufficient to generate electricity, or must they also be separated by a decomposable fluid conductor ? The reader is, no doubt, aware of the answer which we must now give to this question, in accordance with our scientific principles, even although the opposite view is perhaps not wholly refuted.3 We cannot in this place further occupy our attention with these theories, which belong to physics and not to chemistry. Nicholson and Carlisle 4 observed in 1800, that upon discharging the galvanic pile through water, the latter is de- composed into its constituents, hydrogen and oxygen. Many endeavours were made to observe similar phenomena in the cases of other substances, but the first of the more important investigations into the nature of the decomposition of chemical compounds by means of the electric current, comes from Berzelius and Hisinger, and was published in i8o3.5 These two investigators studied the action of dynamic electricity upon salt solutions, and, further, upon ammonia, sulphuric acid, etc. Their apparatus was so arranged that they were able to collect, separately, the constituents liberated at the different poles. In this way they arrived at the highly remarkable result, that substances may be divided into two groups in respect to their behaviour towards the galvanic current; that hydrogen, the metals, the metallic oxides, the alkalies, the earths, etc., 1 De Viribus Electricitatis in Motu Musculari Commentarius. Com mentarii Acad. Bononiae. Vol. 7 (1791). 2 Giornale Fisico-medico di L. Brugnatelli, 1794; Gren's Journal der Physik. [2], 2. 3 Compare Wiede- mann, Galvanismus. I, 25. 4 Nicholson's Journal (quarto), 4 (1801), 183. 5 Ann. Chim. 51, 167. 70 HISTORY OF CHEMISTRY. [LECTURE V. separate at the negative pole of the battery, while oxygen, the acids, etc., separate at the positive pole. Besides this, they believed that they had discovered relations between the quantities of the substances decomposed, their mutual affinities, and the quantities of electricity derived from the battery. With respect to the cause of the decomposition, they only express themselves in a very uncertain manner ; they believe that this decomposition may be explained by the greater or less attraction that the electricity exerts upon the different substances. I now turn to Humphry Davy's investigations, which, as he says himself, were begun in i8oo.c He commenced them with, apparently, a quite unimportant question. Even in the first experiments on the decomposition of water 7 it was believed that the formation of alkaline and of acid substances as products of the electrolysis had been observed. Cruickshank 8 and Brug- natelli 9 confirmed this observation, and a belief was entertained in the conversion of water into alkalies and acids under the influence of the electricity. Simon 10 had already opposed this view, and Davy refuted it by decisive experiments.11 Davy causes the decomposition to take place in vessels made of different materials — glass, agate, gold, etc. — and satis- fies himself that the nature and quantity of the substances liberated are thereby varied. This leads him to the assumption of the decomposition of the vessel. But even when he carries out the decomposition in gold vessels, he observes the forma- tion of the volatile alkali (ammonia) and of nitric acid. These, he now concludes, owe their formation to the air (nitrogen) which was dissolved in the water. In order to satisfy himself of the accuracy of this view, he causes the decomposition to take place in closed vessels, pumping out the air which is in contact with the surface of the water, and substituting for it an atmosphere of hydrogen. In this way he succeeds in proving that pure water is decomposed by the action of the electric 6 Phil. Trans. 1807, 2. 7 Nicholson's Journal (quarto), 4 (1801), 183. 8 Ibid. 188-189. 9 Phil- Mag- 9> 181. 10 Gilb. Ann. 8 (1801), 36. 11 Bakerian Lecture for 1806 (Phil. Trans. 1807, i). LECTURE V.] HISTORY OF CHEMISTRY. 71 current into its constituents, hydrogen and oxygen, but that no other kind of change occurs in the water; and that all observa- tions of the apparent occurrence of such change are to be ex- plained either by some action upon the vessel in which the experiment has been conducted, or by some impurity in the water. This enquiry is comparable, in many respects, with the first of Lavoisier's investigations.12 In both cases an endeavour is made to refute a statement based upon inaccurate observation, and that not simply by speculation, or by the assertion that it is in contradiction to general views. In opposition to the earlier superficial experiments, new ones are adduced which have been carried out with the most minute attention to all the conditions. In both cases the purpose is attained, and the older inaccurate view is replaced by a correct one. Such results as those obtained in these cases by Lavoisier and by Davy are often called negative, but most people will agree with me when I assert that they may be of great positive value. Davy, however, does not stop here. He next investigates the decomposition of salt solutions, and finds confirmation of the statements of Hisinger and Berzelius. But he proceeds with still greater circumspection, and endeavours to follow up the phenomena more exactly. All the means are at his com- mand, and he does not fail to avail himself of them. Direct observation shows Davy that hydrogen, the alkalies, the metals, etc., are separated by means of the current at the negative pole, and oxygen and the acids at the positive pole. From this he concludes that the former substances possess a positive, while oxygen and the acids possess a negative electri- cal energy ; that in this case, as usual, the oppositely electrified bodies attract each other ; and that, in consequence, the posi- tive substances separate at the negative pole, and vice versa. In this assumption Davy had arrived at a conception, or, shall I say, an explanation of the phenomena of decomposition ob- served in the galvanic circuit. But he goes a step further, and 12 See p. 22, 72 HISTORY OF CHEMISTRY. [LECTURE V. tries to refer all chemical combination and decomposition to similar causes. Volta had assumed13 that simple contact of two hetero- geneous bodies was sufficient to place them in opposite electri- cal conditions, and this hypothesis explained the generation of the electric current to himself and to his numerous followers. Davy identified himself with this opinion, and tried to prove its accuracy by direct experiments.14 He brought dry, in- sulated acids into contact with metals, and showed by means of the gold-leaf electroscope that the acids thereby become negatively, and the metals positively electrified. He observed similar phenomena on rubbing sulphur upon copper, when the former became negative and the latter positive, Davy found further that these electrical energies which, in the last case, for instance, are only slight at ordinary temperatures, increase considerably on heating, and are very great at the melting-point of sulphur. On still further elevating the temperature, the two substances unite with evolution of light, and the compound obtained is non-electric. Davy con- cludes from this that the combination consists in a mutual discharge of the opposite electricities, and that heat and light, which appear simultaneously, are consequences of this discharge. According to him, chemical affinity is produced by difference in electrical condition, and the affinity increases or diminishes the greater or less this difference is. In cases where there is considerable difference of energy, equalisation is accompanied by phenomena of fire ; with feebly electrified substances, only small quantities of heat are evolved ; but if combination is to take place at all, the electrical energy must always be able to overcome the cohesion of the substances. Davy tries to prove directly the dependence of chemical affinity upon electrical condition, since he says : — 15 " As the chemical attraction between two bodies seems to be destroyed by giving one of them an electrical state different 13 Brugnatelli, Ann. di. Chim. 13 and 14 ; compare also Ann. Chim. 40, 225. 14 Phil. Trans. 1807, 32. 15 Ibid. 1807, 39. LECTURE V.] HISTORY OF CHEMISTRY. 73 from that which it naturally possesses ; that is, by bringing it artificially into a state similar to the other, so it may be in- creased by exalting its natural energy. Thus, whilst zinc, one of the most oxidable of the metals, is incapable of combining with oxygen when negatively electrified in the circuit, even by a feeble power ; silver, one of the least oxidable, easily unites to it when positively electrified ; and the same thing might be said of other metals." He observes in another place that if there were no cohesion, the chemical attraction would necessarily be proportional to the electrical forces. Both are, in his view, effects of the same power, which, if it extends to the mass of a substance, gives rise to electricity, while, if it excites the smallest particles, it produces affinity.16 By the action of the electric current, the electricity which was liberated from the atoms upon their com- bination, is restored to them again, and decomposition is there- by effected. In this action the positive substance goes to the negative pole and vice versa. It must be admitted that these views start from a simple and a clear idea, and that by the application of this idea they explain the observed facts in a way that is easily understood. They therefore fulfil the conditions which are required in a scientific hypothesis, and secure that their founder, Davy, shall always be regarded as an investigator of undoubted originality. Davy's fame spread very rapidly, and, when he succeeded, a year afterwards, in isolating the alkali metals, it appeared as if he had, for the first time, pointed out the proper line of ad- vance in chemistry. It is true that Davy's theory was illumi- nated by but short glimpses of sunshine, for ten years later we find that it is abandoned. It was certain to fall as soon as the contact with one another of heterogeneous substances was no longer regarded as a source of the excitation of electricity ; and the affirmative view with respect to this matter was soon vigor- ously contested. Ritter, in particular, strove to show17 that 16 Davy, Elements of Chemical Philosophy, 165. 17 Ritter, Elek- trisches System, 49 ; see also Gehlert's Physikalisches Worterbuch. 4, 795. 74 HISTORY OF CHEMISTRY. [LECTURE V. galvanic currents are only produced simultaneously with the occurrence of chemical decompositions. He assumed that electrical phenomena are the consequences of chemical pro- cesses, but that mere contact is not sufficient to produce different electrical conditions. Davy's theory could not be brought into harmony with these views, and accordingly it was given up. This was not, however, on chemical but on physical grounds. A new system had in fact already arisen, which was able effectively to take the place of the old one. This was the electro-chemical theory of Berzelius. I postpone its discussion until the next lecture, because otherwise I should require to enter fully into the re- searches of Berzelius ; whereas I desire at this place to explain more clearly the influence of Davy upon chemistry, by giving some account of the discovery of potassium and sodium, and of the discussion regarding the nature of these substances. In the course of his investigations respecting the conversion of water into acid and basic substances, Davy had had oppor- tunity of gaining a knowledge of the decomposing power of the electrical current, since neither glass, agate, nor felspar had proved able to resist its effects. He thus hit upon the idea of exposing the alkalies also to this action, in order to separate them into their constituents if any such were present in them.18 For these experiments he first employs concentrated aqueous solutions of potassium and of sodium hydroxides ; and as he does not succeed in obtaining products of their decomposition in this way, he next passes the current through the fused alkalies. He now observes the formation of small metallic globules, which burn, however, with great brilliancy, as soon as they come into contact with the air. Yet, by suitable arrange- ments, he succeeds in isolating small quantities of potassium and of sodium, and in studying their most important properties. I remark, in passing, that he only obtained potassium in the fused state. Gay-Lussac and Thenard, who showed how to effect the reduction of the alkali metals by means of iron, in 18 Phil. Trans, 1808, i ; A.C.R. 6, 5. LECTURE V.] HISTORY OF CHEMISTRY. 75 i8o8,19 first proved that potassium is solid at ordinary tempera- tures. They used purer materials, and, by their method, they had much larger quantities of the new substances at their disposal. I cannot enter here into the complete history of potassium and sodium, although the subject becomes highly interesting from the facts that all Davy's experiments are at once checked by the French chemists ; that the latter then bring forward independent results, which Davy doubts ; and so forth. But one point in this somewhat vigorously sustained discussion seems to me to be of sufficient importance to deserve attention, namely, the views concerning the constitution of potassium and of sodium, and those concerning their relation to the alkalies. In the decomposition of the alkalies, Davy had observed that the potassium and sodium appear at the negative pole, whilst oxygen is simultaneously evolved at the positive pole.20 He had further found that the new substances possessed the property of reducing metallic oxides ; 21 and he believed that the alkalies were reproduced when the metals were burned in oxygen.22 From these results he draws the conclusions that the alkalies are the oxides of metals, and that the substances he has discovered are the metals themselves.23 The physical properties of the substances, especially their metallic lustre, supported this view, although their low specific gravity seemed to be an argument, but not a sufficient one, against his con- clusions. Davy accordingly proposes, for the substances, the names Potassium and Sodium, in which the termination is intended to indicate their metallic nature. Davy now holds so firmly to this hypothesis regarding the constitution of the alkalies, that he also regards as oxides many other substances about whose composition the evidence is not yet by any means so clear. Thus, like his contemporaries, he 19 Ann. Chim. 65, 325. 20 Phil. Trans. 1808, 6 ; A.C.R. 6, 10. 21 Phil. Trans. 1808, 19; A.C.R. 6, 22. ** Phil. Trans. 1808, 8; A.C.R. 6, ii. 23 Phil. Trans. 1808, 32; A.C.R. 6, 34. 76 HISTORY OF CHEMISTRY. [LECTURE V. assumes the existence of oxygen in hydrochloric acid, and, in \ opposition to the experiments of Berthollet,24 advances the opinion that ammonia is also an oxygen compound.25 He suspects the presence of oxygen in silicic acid, which he tries to reduce ; 26 in the earths, which, as we know, he succeeded in reducing ; 2T finally also in phosphorus and sulphur,28 a view which was refuted by Gay-Lussac and Thenard.29 Some time afterwards, when engaged upon the investigation of ammonium amalgam 30 (which had been discovered shortly before by Seebeck 31 and more minutely studied by Berzelius and Pontin 32), he finds its behaviour to be analogous to that of the other amalgams, and assumes that it is produced by the combination of mercury with a hypothetical substance, am- monium, resembling the metals, and itself containing hydrogen and ammonia. In comparing now this ammonium with the metals, he is led to ascribe a similarity of constitution to both ; that is, he assumes the existence of hydrogen in the metals also, whereby their combustibility could be explained. Davy states this as a possibility, which he most correctly recognises as identical with Cavendish's phlogiston theory, and which he naturally extends to potassium and sodium. Gay-Lussac and Thenard arrived almost simultaneously at the same view.33 They had investigated the action of potas- sium on ammonia gas, and, in doing so, had observed the formation of a green substance (potassamide), with simultaneous evolution of hydrogen. In these experiments, which they carried out quantitatively, they found the quantity of hydrogen evolved to be identical with that which the quantity of potas- sium employed would have liberated from water. They further showed that, in addition to the formation of potash, the whole 24 Mem. de 1'Acad. 1785, 324 ; Stat. Chim. 2, 280; E. 2, 238. w Phil. Trans. 1808, 35; A.C.R. 6, 37. 26 Phil. Trans. 1810, 59. 27 Ibid. 1810, 16, 62, etc. 28 Ibid. 1809, 67, etc ; Ann. Chim. 76, 145. M Recherches physico-chimiques. I, 187. 30 Phil. Trans. 1810, 37. 31 Gehlen's Journal fiir die Chemie, etc., 5, 482. 32 Bibliotheque Britannique (1809), 122, Nos. 323 and 324; Gilb. Ann. 36 (1810), 261. 33 Ann. Chim. 66, 205. LECTURE V.] HISTORY OF CHEMISTRY. 77 of the ammonia employed was generated anew on the de- composition of the green substance by the action of water upon it. They explained these observations by the assumption that potassium consisted of potash and hydrogen ; and that the latter was liberated by treatment with ammonia, as well as with water, the alkali combining simultaneously with the ammonia or with water. Thus, in accordance with their view, potassarnide was composed of ammonia and potash, and it was decomposed into these constituents by the action of water upon it. Davy, meanwhile, had reverted to his first explanation, and he now attacks these latest results obtained by Gay-Lussac and Thenard.34 In so far as the accuracy of his experiments is concerned, however, he falls short of his opponents, although he is more fortunate and more masterly in the interpretation of them. In his view, the evolution of hydrogen during the action of potassium on ammonia, arises from the decomposition of the latter. According to him the green substance is com- posed of potassium and the residue from the ammonia gas. On treating the green substance with water, the latter is decomposed into its constituents, regenerating ammonia, and giving its oxygen to the potassium which is thus converted into potash.35 In addition to this, since he considers fused caustic potash to be free from water, he finds in the conditions necessary for the preparation of potassium, a further argument against its containing hydrogen. Gay-Lussac and Thenard persist, at first, in their previous opinion that potassium and sodium are hydrogen compounds,36 and in this they are supported by the contemporaneous state- ments made by Berthollet 3T and by D'Arcet,38 who regard fused potash as containing water. In 1811, however, they adopt Davy's view.39 The reason for their change of opinion is to be found in the following facts. They had observed that the 34 Phil. Trans. 1808, 365 (Note). 33 Ann. Chim. 75, 168 et seq., 264. 36 Ibid. 75, 299. 37 Mem. d'Arcueil. 2, 53. M Ann. Chim. 68, 175. 0 Rech. phys. chim. 2, 250, etc. 78 HISTORY OF CHEMISTRY. [LECTURE V. substance obtained upon burning potassium is different from potash, in so far that it contains more oxygen than the latter ; 40 and they point out that the fact of potassium containing hydrogen would necessitate that the new oxide should contain water, since the production of uncombined water is not observed during the combustion. Further, as the oxide is decomposed by dry carbonic anhydride with the formation of oxygen and potassium carbonate, their hypothesis as to the composition of potassium leads to the assumption of the presence of water in salts, in cases where analysis does not reveal any. From this period onwards the non-decomposability of the metals was no longer seriously doubted. At the same time the elementary nature of phosphorus and of sulphur was estab- lished anew by decisive experiments of Gay-Lussac and Thenard,41 and the view that ammonia contained oxygen was recognised as an error by the younger Berthollet.42 To the foregoing, I may add an account of another scien- tific discussion, between Davy on the one hand, and Gay- Lussac and Thenard on the other, which is also of importance inasmuch as it led to the overthrow of Lavoisier's theory of acids. The discussion ii\ question touches the composition of hydrochloric acid. In this acid, as in all others, Lavoisier had assumed the existence of oxygen. Its presence there had never been established, it is true, but the general theory required it ; and as this theory was almost universally adopted, the existence of oxygen in hydrochloric acid seemed to be unquestionable. I say it was almost universally adopted, -because Berthollet, for instance, was of a different opinion. In 1787 the latter had examined hydrocyanic acid, discovered some time previously by Scheele,43 and had found in it carbon, hydrogen, and nitrogen only. It was likewise known, from other investiga- tions by Scheele, that sulphuretted hydrogen contains hydrogen and sulphur only ; and consequently Berthollet felt justified in regarding other elements as acid-producing, as well as oxygen.44 40 Rech. phys. chim. I, 125. 41 Ann. Chim. 73, 229. 42 Mem. d'Arcueil. 2, 268. 43 Ann. Chim. I, 30. 44 Stat. Chim. 2, 8 ; E. 2, 8. LECTURE V.] HISTORY OF CHEMISTRY. 79 He does not appear, however, to have had many adherents in this view, and, so far as hydrochloric acid is concerned, he also assumed the existence of oxygen in it. Chlorine was looked upon as oxygenated muriatic acid, and was supposed to be pro- duced from muriatic (hydrochloric) acid by its taking up oxygen. These latter views were strengthened by the experiments of Henry, and by the interpretation of them which he gave.45 He passed electric sparks through gaseous hydrochloric acid, which was confined over mercury, and obtained hydrogen, while the metal was simultaneously attacked by what he believed to be free oxygen. This led him to the assumption of the presence of water in hydrochloric acid, a view which found general approval since the investigations of others appeared to be in agreement with it. In 1808, Davy had decomposed hydrochloric acid by means of potassium,46 and in this way had obtained hydrogen and potassium chloride, the latter of which he had also pre- pared by burning potassium in chlorine. He showed, in 1809, that the chlorides (muriates) of the metals are not decomposed by heating them with phosphoric glass,47 or with silicic anhy- dride, but that decomposition at once begins when aqueous vapour is passed over the mixture.48 Davy was of opinion that Henry's hypothesis furnished the explanation of these experi- ments, and that hydrochloric acid could only be separated as soon as the quantity of water necessary for its existence was supplied. Gay-Lussac and Thenard further showed, about the same time, that water is produced as well as silver chloride by the action of this acid upon silver oxide ; and, as formerly, they assumed that this water was already present in the hydrochloric acid.49 They then effected the synthesis of the acid, by expos- ing a mixture of chlorine and hydrogen to sunlight.50 On this occasion, they advance a complete theory regarding hydro- 45 Phil. Trans. 1800, 191. 46 Ibid. 1809, 91; A.C.R. .9, 7. 47 Phosphoric glass is calcium metaphosphate ; obtained by sufficiently strongly heating monocalcium phosphate. 48 Phil. Trans. 1809, 93 ; A.C.R. 9, 9. 49 Rech. phys. chim. 2, 118. 50 Ibid. 2, 159. 8o HISTORY OF CHEMISTRY. [LECTURE V. chloric acid and chlorine, by means of which they are able to explain all their experiments.51 According to them, hydro- chloric acid is a compound of an unknown radical, muriaticum, with oxygen and water; chlorine, on the other hand, is anhydrous hydrochloric acid combined with more oxygen, or, what amounts to the same thing, it is ordinary hydrochloric acid minus hydrogen. On this hypothesis the above-men- tioned experiment of the synthetic formation of hydrochloric acid is easily explained. The other facts connected with the matter can be explained on the same hypothesis in just as logical a manner. It is true that the two French philosophers exerted themselves fruitlessly in trying to give direct proof of the presence of the supposed oxygen. It was in vain that they passed hydrochloric acid gas over red-hot charcoal : no change was observable, and this negative result might well have led them to another explanation.52 They point out that the hypothesis that chlorine (oxygenated muriatic acid) is a simple substance, and hydrochloric acid its hydrogen compound could also serve as the basis for the explanation of the observed facts; they prefer, however, to adhere to the old view. Davy, who arrived, independently it would appear, at the latter assumption, declares himself a decided adherent of it.53 He lays great stress on the fact that it agrees with Scheele's original idea, in accordance with which chlorine was dephlogis- ticated hydrochloric acid ; and he tries to support this view by means of new arguments and new experiments. He draws attention to the fact that chlorine does not become converted into hydrochloric acid by the removal of oxygen, but only by treatment with substances containing hydrogen ; and, further, that chlorine is a neutral substance, which, adopting the old hypothesis, is not in agreement with Lavoisier's theory, since it would then be necessary to assume that a substance indifferent to litmus had been obtained from an acid by the addition of 51 Mem. d'Arcueil. 2, 339 ; Bulletin de la Soc. Philomatique, No. 18, May, 1809. 5- Mem. d'Arcueil. 2, 357-358. 53 Ann. Chim. 76, 1 12 and 129. LECTURE V.] HISTORY OF CHEMISTRY. 8 1 oxygen to the latter. Finally, he points out to chemists the many hypothetical substances to which it is necessary to have recourse in order to preserve the older view, whereas the new one explains the facts in the simplest manner. Gay-Lussac and Thenard do not admit this. In 1811, when they publish their investigations in full, under the title : Recherches physico-chimiques, they place the two theories side by side, and show how both suffice to explain the facts.54 Never- theless, they declare against the new system, and the reason they give for doing so, deserves to be mentioned. If chlorine were a simple substance, dry sodium chloride would necessarily decompose water when it dissolved in it, in order that muriate of soda might be produced ; and this they consider more than unlikely. Moreover they held Lavoisier in too high esteem to abandon lightly any such doctrine as the one advanced by him that all acids contain oxygen. But their resistance did not avail them long ; the force of facts was stronger than they were, and Gay-Lussac was far too clear-headed a thinker to be blind to them. The facts thaf made him an adherent of Davy's view, in 1813, were especially his own experiments upon iodine,55 which had been discovered by Courtois and described by Clement,56 and whose analogy with chlorine he recognised and emphasised ; and further, the discovery of hydriodic acid. From this time onwards, the new theory gains ground, and even Berzelius is unable to alter the current of opinion, although he puts himself to every conceiv- able pains, in a paper published in 1815, to deter chemists from the threatened step.57 After pointing out that the hypothesis of muriaticum is still in a position to explain the facts, he draws attention to the fact that it, alone, is in agreement with the general theory of acids ; whereas, on the assumption that there is no oxygen present in hydrochloric acid, a distinction is drawn between this compound and the other acids, to which, nevertheless, it shows the greatest resemblances. The salts 54 Rech. phys. chim. 2, 94. 55 Ann. Chim. 91, 5. 56 Moniteur 1813, Nos. 336 and 346. 57 Ann. Phil. 2, 254; Schweigger's Journal, 14, 66, F 82 HISTORY OF CHEMISTRY. [LECTURE V. also, would then necessarily fall into two classes ; or in other words, it would be necessary to assume differences in con- stitution in a series of substances whose behaviour is similar in ever-y respect. He thinks, further, that he is justified in con- cluding from the laws of combination, that chlorine is not an element. I do not enter more minutely into the matter since his arguments had little effect. They came too late. Gay- Lussac's investigation of hydrocyanic acid in the same year,58 indisputably proves the acid nature of this compound, and the fact that it does not contain oxygen ; and hence even Ber- zelius cannot maintain Lavoisier's definition of the acids and of the acidifying principle. Some other cause which might furnish the acid character observed in certain substances was now sought for. The conception of an acid seemed so definite at that time, and the substances included in this class were so distinctly separated from all other substances, that it was necessary to enquire into the cause which occasioned this difference. Besides, it cannot be denied that Lavoisier, and even the chemists of the beginning of this century were still influenced, in a certain respect, by the ideas of the Greek philosophers. Just as the latter ascribed general properties to the" presence of a common constituent and identified the particular property to a certain extent with a particular constituent — just as they explained combustibility, for example, by the presence of a fire-material — so Lavoisier and his adherents believed that in oxygen they had discovered the acidifying principle. In a similar manner we find Davy, after he was satisfied that hydrochloric acid contains hydrogen and chlorine only, stating the view that the chlorine is the acidifying principle in it, and the hydrogen its basis or radical.59 At a later date Gay-Lussac 60 introduces the name " hydracids " for the acids free from oxygen, and places hydrochloric acid, hydrocyanic acid, sulphuretted hydrogen, and hydriodic acid in this class. 53 Ann. Chim. 95, 136. 59 Phil. Trans. 1810, 231 ; A.C.R. 9, 21. *° Ann. Chim. 91, 148 ; 95, 162. LECTURE V.] HISTORY OF CHEMISTRY. 83 Even if the cause of the acid nature was to be found in the chlorine, iodine, etc., rather than in the hydrogen, still the latter was the common constituent of them all, and hence better adapted to the formation of a name. Davy's investigations upon chloric and iodic acids lead him to much more general views. " Acidity does not depend upon any peculiar elementary sub- stance, but upon peculiar combinations of various substances." 61 At this time he seeks to prove that it is not oxygen which determines the peculiar character of an acid. Thus, for example, when oxygen is united to common salt, the neutrality of the substance is not disturbed ; whereas, on the other hand, the saturating capacity of chloric acid is not altered when all the oxygen is removed from it. This obliges Davy no longer to regard chloric acid (in accordance with Lavoisier's view) as an oxide of the radical chlorine, which, combined with water, forms the hydrated acid. He finds that, without water, chloric acid cannot exist ; and for this reason he regards it as a ternary compound of hydrogen, chlorine, and oxygen. Again, the existence of euchlorine, which he obtains from chloric acid and hydrochloric acid,62 provides him with a reason against Lavoisier's hypothesis regarding acids. The principles of a new theory of acids are included in Davy's discussions ; but he did not follow them up sufficiently, otherwise, they might have prevented the distinction which now began to be drawn between acids which did and those which did not contain oxygen. The same remark holds likewise with respect to Dulong, who had read a paper before the French Academy in 1815, in which he had stated his view regarding acids. This paper, unfortunately, does not appear to have been printed in extenso, and therefore I am able to give but little account of it.63 Dulong, on this occasion, examined oxalic acid. The behaviour of some of its salts, which give off water when heated, led him to the opinion that the acid might be regarded as a hydrogen compound of 61 Phil. Trans. 1815, 219. 62 Ibid. 1811, 155 ; A.C.R. 9, 63. 63 Mem. de 1'Acad. 1813-1815, Histoire, p. cxcviii. ; see also Schweigger's Journal. 17, 229; Ann. Phil. 7, 231. 84 HISTORY OF CHEMISTRY. [LECTURE V. carbonic acid, — hydro-carbonic acid. On saturation with a metallic oxide, the oxygen of the oxide combines with the hydrogen of the oxalic acid to form water which can then be driven off, and a compound of carbonic acid with metal remains behind. Here we find the view for the first time represented (by Dulong and Davy) that the water produced during salt formation was not already contained in the (oxygenised) acid, and that, in a salt, it is not a metallic oxide but the metal, as such, that exists. It is stated that Dulong made a similar assumption for the other acids, but unfortunately the development of his ideas has not been handed down to us. Besides, hypotheses of this kind met with little approval at that time ; voices were loud in their condemnation from the most different sides. Gay-Lussac declared himself emphatically against them ; w and Berzelius (who was then beginning to exercise a predominant influence) seeing that he was obliged to admit the existence of acids free from oxygen, introduced a strict distinction between the latter and the acids containing oxygen, and so between the haloid and the amphid salts. This single point irj the system of Berzelius must not, however, be treated of separately, but the system must be dealt with as a whole. His views are of the utmost importance, since they dominated theoretical chemistry for twenty years. I shall devote the next lecture to their consideration. 64 Ann. Chim. [2] I (1816), 157. LECTURE VI. BERZELIUS AND HIS CHEMICAL SYSTEM — DULONG AND PETIT'S LAW — IsoMORrHisM — PROUT'S HYPOTHESIS — DUMAS' VAPOUR-DENSITY DETERMINATIONS — GMELIN AND HIS SCHOOL. BERZELIUS adopts dualism as the basis of his system. Even before his time, the majority of compounds had been looked upon as consisting of two parts. A uniform mode of regarding them in this aspect became possible to a still greater degree in the light of the electro-chemical phenomena, and it is the great merit of Berzelius to have introduced this into the science and to have established it. In his view, compound substances are produced by the arrangement of the atoms side by side.1 Compounds of the first order are formed in this way from the smallest particles of the elements ; these compounds give rise in turn to the forma- tion of compounds of the second order ; and so on. Berzelius, like his predecessors, seeks, in affinity, the reason for the com- bination of two atoms, but this again is for him, as it was foi Davy, a consequence of the electrical properties of the smallest particles. He differs from Davy very essentially, however, in the manner in which he assumes the electrical distribution. But, quite apart from this, the two theories are not to be compared as regards their importance for our science. Davy, no doubt, advanced ingenious ideas as to the mode in which he considered the chemical and electrical phenomena to Le inter-related ; but from these hypotheses, by means of which a 1 Essai sur la theorie des proportions chimiques et sur Pinfluence chimique de 1'Electricite, 26. See also Berzelius, Lehrbuch der Chemie, First Edition, Vol. 3, part I. Compare also Schweigger's Journal. 6 (1812), 119. 86 HISTORY OF CHEMISTRY. [LECTURE VI. number of facts could be most excellently explained, he never succeeded in producing a theory which might serve as the foundation of a chemical system. Berzelius was the first to do this. He made it his life's task to establish in chemistry a uniform system which should be applicable to all the known facts • and he accomplished it. Hence his views are of far greater importance in the development of chemistry than those of Davy are. According to Berzelius, it is not only when two substances are brought into contact that electricity is generated, but it is a property of matter; and in every atom, two oppositely electrical poles are assumed.2 These poles do not, however, contain equal quantities of electricity. The atoms are unipolar, the electricity of the one pole predominating over that of the other; and thus every atom (and therefore every element) appears to be either positively or negatively electrical. In this respect it is possible to arrange the elementary substances into a series, so that each member is always more electro-negative than the next succeeding one. Oxygen stands at the top, and is absolutely electro-negative,3 while the other substances are only relatively positive or negative according as they are com- pared with elements which come before them or after them in the electrical series. This series does not constitute a table of affinities in the Geoffrey- Bergman sense ; and it does not express the affinity of the individual substances for oxygen, for example. Berzelius has not forgotten Berthollet's teaching, that affinity is not of a constant character and independent of the physical conditions; as he supposes this unipolarity to be ; and he is also well aware that oxygen can be removed from metallic oxides by carbon or sulphur, that is to say, by other electro-negative substances. With him, affinity depends princi- pally upon the intensity of the polarity, i.e., upon the quantity of electricity which is contained in the two poles. This is variable, however, especially with changes of temperature. - Essai etc. 85. 3 In Schweigger's Journal. 6, 129, where he states his electro-chemical theory in detail, Berzelius calls oxygen electro-positive. LECTURE VI.] HISTORY OF CHEMISTRY. $7 Generally speaking, it is increased by supplying more heat, and this explains why certain combinations only take place at a high temperature.4 During the combination of two elements, the atoms arrange themselves with their opposite poles towards each other, and mutually discharge their free electricities, whereby the phenomena of heat and of light are produced. The old doctrine is explained at the same time — Corpora non agunt nisi soluta (substances do not interact unless dissolved), since free motion of the smallest particles is only possible in the liquid state. When a substance is subjected to the action of the electric current, the latter restores to the atoms their original polarity, whereby the substance breaks up into its constituents. A compound of the first order is not electrically (nor yet chemically) inactive, since, during the combination, only one pole of each atom is neutralised ; it is still unipolar, and it can enter into further combinations (of the second order) which are likewise endowed with electrical forces ; but the intensities of these forces diminish, the higher the order of the compound becomes, since the stronger poles are, in general, neutralised first. According to Berzelius, the specific unipolarity of the oxides depends solely upon the radical or element combined with the oxygen. The latter gives rise to the most powerfully electro-positive and electro-negative substances (alkalies and acids) ; and as it cannot, therefore, be itself the cause in both cases, it cannot be the cause in either.5 All chemical reactions, and, consequently, the phenomena of heat and light that accompany them, are, according to Berzelius, produced by electricity, which " thus seems to be the first cause of the activity all around us in nature." 6 If a substance C is to decompose the compound AB^ so that B may become free, then C must be able to neutralise a greater amount of the electrical polarity of A than B can. Further, a mutual exchange between AB and CD only occurs 4 Lehrbuch. Second Edition, Vol. 3, Dresden (1827), part I, 73-74. 5 Ibid. 76. 6 Ibid. 77. 88 HISTORY OF CHEMISTRY. [LECTURE vi. if the electrical polarities are better equalised in AC and BD than they were previously. In reactions of this kind, Ber- zelius, like Berthollet, assumes an influence upon the resulting phenomena, of the quantities of the substances present and of cohesion ; but he differs from Berthollet in regarding the affinity as a function of the electrical polarity, and as independent of the saturating capacity of the substance. This theory constitutes the basis of the dualistic theory of chemical composition. Berzelius establishes it as follows : — 7 " If the electro-chemical views are accurate, it follows that every chemical combination depends wholly and only upon two opposite forces, namely, the positive and the negative electricities, and that every compound must be composed of two parts, united by the effects of their electro-chemical reac- tions, since there is not any third force. From this it follows that every compound substance, whatever the number of its constituents may be, can be divided into two parts, of which the one is positively and the other is negatively electrical. Thus, for example, sulphate of soda is not composed of sulphur, oxygen, and sodium, but of sulphuric acid and soda, each of 'which can, in turn, be. separately divided into an electro- positive and an electro-negative constituent. In the same way, also, alum cannot be regarded as immediately composed of its elementary constituents, but is to be looked upon as the pro- duct of the reaction of sulphate of alumina, as negative element, with sulphate of potash, as positive element; and thus the electro-chemical view justifies what I have said with respect to compound atoms of the first, second, third, etc., orders." As may be perceived from this, Berzelius had formed a definite opinion as to the composition of compounds, and he went so far with this opinion, that he regarded as inadmissible the assumption, which lies readiest to hand, that the substance is composed of its elementary constituents. He thought he knew the arrangement of the atoms in compounds so minutely (principally from the decompositions which they underwent 7 Lehrbuch. 3, part I, 79-80. LECTURE VI.] HISTORY OF CHEMISTRY. 89 when subjected to the influence of the electric current), that he only regarded one particular view as possible. As an appendix to the compounds, Berzelius takes solu- tions into consideration. He does not place these in the same class with the compounds, because a disappearance of heat is observed during their formation, and therefore no electrical discharge can take place in the operation.8 Before proceeding to the further statement of the system of Berzelius (especially before turning to his interesting and extremely important method of atomic weight determinations) I wish to say something about the nomenclature 9 and the system of notation,10 which he had proposed some years before. In doing so, I can be all the more concise, as the former is merely the perfecting of the system introduced by Guyton, Lavoisier, Berthollet, and Fourcroy,11 and both are, no doubt, well known. I can, therefore, confine myself to what is of essential importance, or is characteristic of the point of view of Berzelius. Substances are divided into ponderables and imponderables. Amongst the latter we find electricity, magnetism, heat, and light. The former are divided into elements and compounds, solutions and mixtures. Amongst the simple substances, Berzelius places the metals and the metalloids. He uses a word here which Erman had employed before him for desig- nating the metals of the alkalies and of the earths;12 but Berzelius is the first to give it the meaning that we still attach to it. The oxygen compounds are either called oxides or acids. The substances of this class which possess neither basic nor acid properties, and contain relatively little of the negative element, are called sub-oxides. The basic salt-forming oxygen compounds are designated oxides ; when an element or a radical forms two substances of this kind, these are distinguished by the terminations of the specific name. This is very easy 8 Lehrbuch. 3, part I, 80. 9 Journ. de Phys. 73, 253. 10 Essai etc. ill. u Compare p. 32. 12 Gilb. Ann. 42, 45. 90 HISTORY OF CHEMISTRY. [LECTURE VI. in the Latin nomenclature (which Berzelius proposes to employ) ; for example, oxidum ferrosum (with the smaller ratio of oxygen) and oxidum ferricum. Finally, superoxides are also distinguished. These contain a relatively large proportion of oxygen, and must be reduced before they form salts. What Berzelius says respecting compounds with water, is of interest. According to him, water may occur in compounds, combined in three different ways. It plays the part, either of an acid, as in the caustic alkalies, or of a base, when it unites with acids. In both of these cases it is called water of hydra- tion, and is distinguished from water of crystallisation which unites with salts, and can be separated from these again, without their being, thereby, essentially changed in their nature. The system of notation of Berzelius is original with himself, and, up to the present, it has always proved so thoroughly practical that we have retained his proposals almost without alteration. In his system, the atom of an element is represented by the initial letter of the Latin name of the element ; and by placing the symbols side by side, the atoms of compounds are obtained. When several atoms of one element occur in a compound, a figure giving the number of these is placed after the letter, and above (or below) the line. The so-called double atoms (i.e., two atoms of an element, which occur together) constitute an exception to this; in these the symbol for the atom is " barred." 13 Thus, for example, H = H2, or two atoms of hydrogen ; HO = H2O, or one atom of water, consisting of two atoms of hydrogen and one of oxygen, etc. In the cases of more complicated compounds, several letters are separated from others by the sign + ; and the mode of division is dependent on the dualistic view. For the sake of shortness, the atom of oxygen is often represented by a point, and that of sulphur by a vertical stroke ; and in compounds, these marks are placed above the symbol of the element with which the oxygen or sulphur is combined — a method of writing formulae that has now, however, been abandoned. 13 Lehrbuch. 3, part I, 108. LECTURE VI.] HISTORY OF CHEMISTRY. QI With these indications, we shall now leave this matter, and pass on to a much more important subject in the system of Berzelius, that is, to the mode by which he determined the number of atoms in a compound. He was the first who took account of purely chemical facts in these determinations. He rejects Dalton's rules entirely, pointing out, with justice, their groundlessness : — " When only one compound is known, there is surely something arbitrary in assuming that it consists of one atom of each element, altogether regardless of the other relations of the compound." 14 Berzelius tries to establish the view, however, that regu- larities must exist, which determine the number of the atoms that mutually combine with one another.15 He argues that if an unlimited number of atoms of one element could combine with an unlimited number of atoms of another, there would, in this way, be produced an infinite number of compounds which would differ so little in composition that even our best analyses would fail to show any difference. The view that substances consist of indivisible atoms, and that chemical compounds are produced by the arrangement of these atoms side by side, is not sufficient to explain multiple proportions. In the combina- tion of atoms, special laws must prevail which limit the number of the compounds ; and it is upon these laws, in particular, that chemical proportions depend. He finds his first basis in Gay-Lussac's law of gaseous volumes. This law appears to him to permit of an unequivocal decision of the question, since, with him, atom and volume are identical in the case of simple gases. " We know, for example, with certainty the relative number of atoms of nitrogen and of oxygen in the different stages of oxidation of nitrogen ; that of nitrogen and of hydrogen in ammonia; that of chlorine and oxygen in the different stages of oxidation of chlorine ; " 16 and so on. Amongst the gases, the law of multiple proportions 14 Lehrbuch. 3, part I, 88. 15 Essai etc. 28. 16 Lehrbuch. 3, part I, 89. 92 HISTORY OF CHEMISTRY. [LECTURE VI, finds confirmation in Gay-Lussac's rule ; and, by measuring the volumes, Berzelius can here determine the number of atoms which mutually combine with one another. For example, since two volumes of hydrogen unite with one volume of oxygen, water consists of two atoms of hydrogen and one atom of oxygen. He cannot conceive how anyone can be of a different opinion, and he engages in controversy with Thomson, who only assumes half as many atoms in one volume of hydrogen as in one volume of oxygen. " It has been assumed that water is composed of an atom of oxygen and an atom of hydrogen ; but since it contains two volumes of the latter gas for one of the former, it was concluded that in hydrogen and in inflammable substances generally, the volume weighs only half as much as the atom, while in oxygen, volume and atom have the same weight. As this is only an arbitrary assumption, the accuracy of which cannot even be tested, it appears to me much simpler and more conformable with probability to assume the same relation of weight between the volume and the atom in the combustible substances as in oxygen ; because there is nothing which should make us suppose a difference between them. If water is regarded as composed of two atoms of radical and one atom of oxygen, then the corpuscular (atomic) and the volume theories coincide, so that their difference only consists in the state of aggregation in which they present the substances to us." 17 It must be mentioned that Berzelius does not extend to the compound gases, his view as to the identity of volume and atom, but considers that the atoms of these neither occupy the same space as the atoms of the elements nor show uniformity of volume amongst themselves. That this is so, follows from his atomic weight estimations. With him, H = i = i volume or i atom of hydrogen ; H.2O =18 = 2 volumes or i atom of water ; 73 = 4 volumes or i atom of hydrochloric acid, etc.18 17 Lehrbuch. 3, part I, 44-45. 18 It is true that he calls HC1 = 36.5 an atom of hydrochloric acid and says HG4 is the double atom (see Lehrbuch), but for the most part he actually employs the formula H€l. I return to this again, however, farther on. LECTURE VI.] HISTORY OF CHEMISTRY. 93 Berzelius does not adopt the distinction between physical and chemical atoms, introduced by Avogadro and Ampere ; and he endeavours to surmount the difficulty which had led Dalton to regard Gay-Lussac's law as inaccurate, by completely separating from one another the elementary and the compound gases. It is clear that the law of gaseous volumes, together with the conclusions that Berzelius draws from it, is insufficient. It can only be used to determine the relative number of atoms in a very few compounds, and the founder of the first chemical system is, therefore, obliged to seek for other generalisations of more universal validity. He advances the following rules,19 which are intended, however, to apply to inorganic compounds only. I. One atom of an element combines with i, 2, 3, etc., atoms of another element. He does not state the limit. In 1819, he thinks that more than four atoms of one element seldom combine with one atom of another; afterwards (1828) he drops this limita- tion. II. Two atoms of an element combine with 3 or with 5 atoms of another element. This rule leads him to a discussion of the question as to whether a compound of 2 atoms of one element with 4 or with 6 of another element is identical or not with the combina- tion of i atom of the first element with 2 or with 3 of the second. In his Text-book (i 828) he leans to the latter opinion ; by this time, isomeric compounds were known. The laws of combination of compound atoms of the first, second, and third orders are quite similar, but certain limita- tions here come into play, arising from the fact that when compound atoms combine, they have either the electro- negative, or else, less frequently, the electro-positive constituent common to both, and the proportions in which these atoms then combine are determined by the common element in such 19 Essai etc. 29-30. 94 HISTORY OF CHEMISTRY. [LECTURE VI. a way that the quantities of the latter in the one constituent stand to those in the other as i to i, 2, 3, 4, 5, 6, etc. ; as 3 to 2 or 4 ; or finally, as 5 to 2, 3, 4, 4^, and 6.20 It is interesting to see how, by the aid of these rules, Berzelius determines the number of atoms contained in a compound. As an example, I choose the oxygen compounds, which are unquestionably of the greatest importance. Berzelius believes he has discovered (especially from the consideration of the proportions by volume in the case of gases) that it is usually the electro-negative constituent of which several atoms occur, and, therefore, in the case to be examined here, the above mentioned rules take the following form.21 I. When an element or radical forms several oxides, and the quantities of oxygen in these, as compared with a given quantity of the other element, stand to each other as i to 2, it must be assumed that the first compound consists of i atom of radical and i atom of oxygen ; the second, of i atom of radical and 2 atoms of oxygen (or 2 atoms of radical and 4 atoms of oxygen). When the proportions are as 2 to 3, the first compound consists of i atom of radical and 2 atoms of oxygen ; the second, of i atom of radical and 3 atoms of oxygen, and so on. In conformity with this rule, Berzelius, in 1819, writes soda NaO2, and the peroxide NaO3 ; and his formulae for the other oxides are similar. Thus it comes about that the atomic weights which he proposes for the metals at this period, are double those which he definitely adopts afterwards (i828).2'2 Influenced by reasons which we shall immediately learn, he makes, in his Text-book, the following addition to Rule I : — 23 When the proportion of the quantities of oxygen in two compounds is as 2 to 3, then in the first, i atom of radical can also be combined with i atom of oxygen, and, in the second, 2 atoms of radical can be combined with 3 atoms of oxygen. 20 Lehrbuch. 3, part I, 40. 21 Essai etc. 118. -2 Berzelius, Jahres- bericht 1828, 73. -3 Lehrbuch. 3, part I, 90. LECTURE VI.] HISTORY OF CHEMISTRY. 95 II. When a positive oxide combines with a negative one (a base with an acid, for example), the oxygen in the latter is a multiple, by a whole number, of that in the former, and this number usually is, at the same time, the number of the oxygen atoms in the negative oxide. These are the only two rules which Berzelius advances in 1 8 1 9 in his theory of chemical proportions. In the translation of the second edition of his Text-book (the first German edition) new rules are added, called forth by Mitscherlich's discovery of isomorphism, and by the relations which Dulong and Petit had found (1819) between the atomic weights and the specific heats of solid elements. Since both of these investigations are of the greatest importance with respect to the views of Berzelius upon the question now under discussion, I shall here introduce the results of these investigations, and shall then proceed with the consideration of the atomic weight determinations. Dulong and Petit proved, by exact experiments,24 that the products obtained by multiplying the specific heats of bismuth, lead, gold, platinum, tin, silver, zinc, tellurium, copper, nickel, iron, cobalt, and sulphur, by the respective atomic weights of these elements, are almost identical ; and from this fact they drew the conclusion that the same regularity would hold with respect to all the elements and would lead to the exact determination of their atomic weights. In order to establish the law, Dulong and Petit had assumed the atomic weights of most of the metals, with reference to that of sulphur, to be only half as great as Berzelius had stated them in 1819. They assumed 201 for the atomic weight of sulphur (O= 100), as Berzelius had done, and then made Fe = 339, whereas Berzelius had adopted 693. According to their law, the atomic weight of silver was only one-fourth of Berzelius' number. In the cases of tellurium and of cobalt, they arrive at results which are still further at variance 24 Ann. Chim. [2] 10, 395. 96 HISTORY OF CHEMISTRY. [LECTURE VI. with his, but these do not merit any confidence, as later observers (Regnault 25 and Kopp 26) have found other numbers which agree better. I may take this opportunity to state that Neumann showed, in 1 83 1,27 that Dulong and Petit's law may also be extended to compounds of analogous composition ; that is to say, the specific heats of these compounds multiplied by their equiva- lent weights (as Neumann calls them) give equal products. The law was proved, in particular, for the carbonates and the sulphates. Before I pass on to Mitscherlich's interesting results, I wish to make some historical remarks by way of preface. With Hauy, the crystalline form (primative form) was an important characteristic for the determination of the nature of a substance; difference of form being, in his view, a ground for assuming a different composition,28 although Berthollet contested this.29 Gay-Lussac observed, in 1816, that crystals of potash alum increase in volume in a solution of ammonia alum, without altering in shape.30 Beudant 31 also made very interesting statements in this connection; and, as early as 1817, J. N. v. Fuchs 32 drew attention to the similarity of the crystalline forms of arragonite, strontiamte, and cerussite. Gehlen stated that he had succeeded in preparing crystals of alum with soda.33 These were isolated observations, which were insufficient to overthrow Hauy's doctrine, and which only attained to any importance through Mitscherlich's discovery of isomorphism.34 In 1820, Mitscherlich established the fact that the corre- sponding phosphates and arseniates, with the same number of atoms of water, possess the same crystalline form, so that even the secondary forms coincide. Even at that time, the same number of atoms was assumed to be present in both acids, and 25 Ann. Chim. [2] 73, 5; ibid. [3] i, 129; 9, 322 etc. -6 Annalen. Supplementary Vol. 3, 291. 27 Pogg. Ann. 23, I. w Traite de mineralogie. 29 Stat. Chim. I, 433 ; E. I, 432. 30 Kopp, Geschichte. 2, 406. 31 Ann. Chim. [2] 4, 72 ; 7, 399 : 8, 5 ; 14, 326. 32 Schweigger's Journal. 19, 133. 33 Ibid. 15, 383, Note. M Ann. Chim. [2] 14, 172; 19, 350; 24, 264 and 355. LECTURE VI.] HISTORY OF CHEMISTRY. 97 thus Mitscherlich arrived at the idea that it was the similarity of atomic constitution which gave rise to the identity of form. And he really succeeded in confirming this opinion by a series of facts. He called those substances isomorphous, which exhibit the same crystalline form in corresponding compounds, and which, since they can crystallise together, replace one another in indefinite proportions. He pointed out the isomor- phism of selenic acid and sulphuric acid ; that of magnesia, zinc oxide, nickelous oxide, ferrous oxide, etc., in their neutral sulphates, etc. ; that of alumina, ferric oxide, and manganic oxide. He also showed that Beudant's observation, in ac- cordance with which iron vitriol and zinc vitriol (two salts of different crystalline form, and containing different proportions of water) crystallise together, depends upon the fact that the proportion of water in the one compound is changed, and becomes the same as that in the other. Other observers confirmed Mitscherlich's view by means of many observations,35 so that, at that time, much stress was laid upon the crystalline form of substances ; and chemists supposed that they possessed, in this character, an excellent means of obtaining information regarding their atomic constitution. It was Berzelius, in particular, who, instantly recognising the bearing of the great discovery, applied it to the extension of his system. Isomorphism led him to the following rule : — 36 III. When one substance is isomorphous with another in which the number of atoms is known, then the number of atoms in both is known, because isomorphism is a mechanical consequence of similarity of atomic construction. Guided by these rules, Berzelius endeavours to determine the number of atoms contained in a compound, and from this, he can then deduce the atomic weights. He is quite conscious 35 Literature in the article " Isomorphismus " in the Handworterbuch der Chemie, edited by Liebig, Poggendorff, and Wohler. In the article " Isomorphie" in the second edition of this Dictionary, Arzruni enters par- ticularly into the more recent development of the doctrine of isomorphism. 36 Lehrbuch. 3, part I, 91. G 98 HISTORY OF CHEMISTRY. [LECTURE VI. that his rules, in many cases, cannot lead him to a decisive determination, and that it is really in the cases of gaseous elements only that they can give unequivocal results. But just because he knows upon what shaky grounds he is proceeding, he acts with the greatest caution ; and it is marvellous how often, guided by an acute judgment, he hits upon the correct number, where almost every criterion is wanting. In the case of the oxides, Berzelius constructs for himself a series which furnishes him with the relative quantities of oxygen with which certain weights of the metals combine. In doing so, he does not require to construct a series of this kind for every metal. By calling Mitscherlich's law to his assistance, he is able to supply the places of any stages of oxidation that are wanting in the case of a given element, by those of an isomorphous element. The series is : — 3r Relative Proportion of Oxygen. Cuprous oxide - i Cupric oxide, ferrous oxide, etc. - 2 Ferric oxide, manganic oxide, minium - 3 Lead peroxide, manganese peroxide - 4 Manganic acid - 5 I also give, below, a similar but more accurate tabulation, from the year i835.38 Cuprous oxide - i Cupric oxide, ferrous oxide, etc. - 2 Ferric oxide, manganic oxide, etc. 3 Lead peroxide, manganese peroxide, etc. - 4 Nitric acid, chloric acid, etc. 5 Perchloric acid, permanganic acid, etc. 7 In the compounds noted here, Berzelius assumes i, 2, 3, 4, 5 (and 7) atoms of oxygen ; thus making the simplest assump- tion possible. It is then only a matter of determining, in addition, the number of atoms of the radical or element that is united with the oxygen. His series does not furnish any 37 Lehrl-uch. 3, part I, 97. w Lehrbuch. Third Edition, 5, 89. LECTURE VI.] HISTORY OF CHEMISTRY. 99 information about this, and therefore he seeks for other generalisations. He now rejects the apparently most natural assumption of i atom of radical, which he had made in 1819, since it leads him to atomic weights that are not in harmony with Dulong and Petit's law. He finds a new starting-point (except for silver, tellurium, and cobalt) by assuming the pre- sence in compounds of two atoms of the element concerned, and thus he obtains the following series for the stages of oxida- tion of the most of the metals : — R2O, RO, R2O:?, RO2, R2O5 (R2O7), or RO, RO, RO3, RO2, RO5 (RO7), where he writes RO instead of R2O2, and RO2 instead of R2O4. Berzelius advances several grounds which appear to tell in favour of the accuracy of his choice. The most commonly occurring oxides, such as cupric oxide, magnesia, lime, etc., receive the simplest formula, RO ; further, the oxygen com- pounds of nitrogen and of chlorine, in which he knows, from the volumes, the number of atoms, can be made to fit in with his arrangement. On this account, he regards this series as one that occurs generally distributed, calls it the nitrogen series, and contrasts the sulphur series with it. Berzelius finds the relative quantities of oxygen that combine with sulphur to be i, 2, 2^, and 3. Hence, he writes the stages of oxidation of this element : SO, SO2, S2O5, and SO3. He endeavours to arrange all the oxygen compounds, as far as possible, in the' sulphur and nitrogen series, and assumes, for example, SiO3 as the formula for silicic acid, corresponding with sulphuric acid, an assumption which afterwards gave occasion for many discussions. ".- The sulphur compounds (sulphides) are regarded as con- stituted in a manner analogous to the oxygen compounds. He writes sulphuretted hydrogen HS, because water is HO. In calculating the atomic weights which he deduces from these considerations, Berzelius starts from O= 100, but, in order to permit of comparison with earlier and with subsequent statements, and since it is merely a question of the relative 100 HISTORY OF CHEMISTRY. [LECTURE vi. magnitudes of the numbers, I shall give his values, calculated with reference to oxygen as standard with atomic weight = i6.39 j Atomic Weight Element and Symbol. (O = i6) (Berzelius). Atomic Weight (0 = i6) (German Com- mission, 1898). Arsenic - - As 75.33 75-o Calcium - - Ca 41.03 4O.O Carbon - C 12.25 12.0 Chlorine - - Cl 35.47 35-45 Iodine - I 123.20 126.85 Iron - Fe 54.36 56.0 Manganese - Mn 57-02 55-o Mercury - - Hg 202.86 200.3 Nitrogen - - N 14.18 14.04 Oxygen - - O 16.00 16.00 Phosphorus - P 31-43 31.0 Silicon - Si 44.47 28.4 Silver - Ag 216.61 107.93 Sodium - - Na 46.62 23-o5 Sulphur - - S 32.24 32.06 Before I conclude the consideration of Berzelius' system, I shall add a few words with respect to the formulae of hydro- chloric acid and of ammonia. The atoms of these substances are represented by HC1 and NH3,40 showing that Berzelius did not identify the conceptions of atom and equivalent in all cases, although he employs the names indiscriminately. This might, of course, be regarded as no real exception, since, generally speaking, the double atoms HC1 and ^PK3 are alone employed. Naturally, it is difficult to give an exact account of the views of one who is no longer alive ; but it is in every case necessary to take the different periods into account. I believe then that Berzelius, at first, and till about 1830, tried to extend the law of volumes as far as possible (even to compounds as com- pared with one another), and that this was a reason for assum- 39 Berzelius, Jahresbericht 1828, 73 ; the values are also to be found there calculated for H = i. ** Lehrbuch. Third Edition, 2, 187 and 344. LECTURE VI.] HISTORY OF CHEMISTRY. IOI ing the formulae HC1 and NH:? for hydrochloric acid and ammonia ; but that afterwards, influenced especially by Dumas' investigations,41 he placed much less reliance upon this law, and applied it to the permanent (and elementary) gases alone.42 He was then no longer prevented from believing in an agree- ment between equivalent and atom, even in these substances, and he employed only the formulae H€i and NH3. It follows,. from the foregoing, that Berzelius did not admit the distinction between the physical and the chemical atom, and he thereby establishes an essential difference between ele- ments and compounds. According to him, the atoms of the elementary gases occupy, in general, one half (or one quarter) the space occupied by the atoms of the compound gases. Whilst similarity in behaviour with respect to changes of pres- sure and of temperature was a sufficient reason for assuming the same number of atoms in equal volumes of hydrogen and of oxygen, the same reason was insufficient to justify the same conclusion with regard to chlorine and hydrochloric acid. There was an inconsequence in this, but it was of no material importance, since the experiments bearing most closely upon the matter appeared to negative any general applicability to it of the law of gaseous volumes. The chemical edifice which Berzelius erected was a wonder- ful one, as it stood completed (for inorganic substances) at the end of the third decade of the century. Even if it cannot be said that the fundamental ideas of the system proceed exclu- sively from himself, and if he is indebted to Lavoisier, Dalton, Davy, and Gay-Lussac for a great deal, still it was he who moulded these ideas and theories into a connected whole, adding also much that was original. His electro-chemical hypothesis no doubt had points of similarity with that of Davy, but, in spite of that, it was essentially different from it. Besides, the first method of atomic weight determination, of moderately general applicability, proceeded from Berzelius; and this method 41 Ann. Chim. [2] 49, 210; 50, 170. 42 Lehrbuch. Third Edition, 5,8*, 102 HISTORY OF CHEMISTRY. [LECTURE VI. was so extraordinarily serviceable that it rendered possible the fixing of these most important numbers, so that alteration was necessary in only a few cases. It will thus be understood how the system of Berzelius became the prevailing one, and why his judgment was authori- tative. The publication of his yearly reports (Jahresberi elite), which began to appear in 1821 and were employed not merely for reporting but also for criticising, contributed to increase his influence. Hence the ideas of "others possess only a subordi- nate interest, but still I wish to state the views of some of his contemporaries, so that I may the better characterise the period under review. British chemists had not yet come to a decision with re- spect to Dalton's conception of the atom and Wollaston's of the equivalent. Very little of much importance had mean- while been accomplished in Great Britain. The only thing to which I wish to refer is the hypothesis of Prout, which was the occasion of much discussion. Prout, in 1815, thought it was possible to show that the atomic weights of the gaseous elements are multiples, by whole numbers, of that of hydrogen.43 Stated in this way, the matter seems to be of small importance ; but it gains interest from the fact that, if it is admitted to be generally applicable, it almost necessarily leads to the assumption of a primordial form of matter, and to the view that the manifold peculiarities of sub- stances are explicable by the varying distribution of this matter in space. Thomson44 set himself the task of extending the statement of Prout to all the elements, and, for this purpose, he carried out a large number of atomic weight determinations. His results are worthless, however, as Berzelius somewhat bluntly points out to him.45 At a later period, Prout's hypothesis was taken up again by Dumas,46 after it had been shown that a more accurate deter- 43 Ann. Phil. 6, 321. ** Thomson, An Attempt to establish the First Principles of Chemistry by Experiment, 2 vols. London (1825). 45 Berzelius, Jahresbericht 1823, 40. 46 Ann. Chim. [3] 55, 129. LECTURE VI.] HISTORY OF CHEMISTRY. 103 initiation of the numbers told in its favour. In particular, the atomic weights of the best known elements, such as, oxygen, hydrogen, nitrogen, carbon, (chlorine?), bromine, iodine, etc., appeared to be in harmony with it. Stas has proved, however, by means of experiments in which the highest degree of accu- racy and completeness was attained,47 that even in the case of those elements which appeared in accord with it, the hypothesis does not in any instance hold rigidly, and can only be looked upon as an approximation. The highly important theory more recently established by Newlands, Lothar Meyer, and, especially, Mendelejeff, dealing with the periodic relation of the properties of the elements to the magnitudes of their atomic weights, can only be entered upon in a subsequent lecture. In France, the law of volumes in its most extended sense became the basis of the atomic considerations. It was Dumas, especially, who took up a very decided position in this connec- tion. . He shows that the conception of the equivalent cannot be employed as the basis of a system, because it loses its signifi- cance when it is extended further than to acids, to bases, and to other substances which closely resemble each other (oxides and sulphides) ; and, especially, that it becomes quite vague when the attempt is made to identify the equivalent with the com- bining weight,48 since very many substances can combine in several proportions. Thus, for example, 8 parts of copper are combined with i part of oxygen in cuprous oxide, while for 8 parts of copper, 2 parts of oxygen are contained in cupric oxide. Calculated from these numbers, the equivalent (com- bining weight) of copper, referred to that of oxygen as unity, is 8 or 4. Dumas believes that, by adopting Avogadro's hypothesis as a basis, he has obtained a sure guide in considerations regard- ing atomic weights. He assumes that in equal volumes of all 47 Recherches sur les lois des proportions chimiquesetc., Bruxelles 1865, and Recherches sur les rapports reciproques des poids atomiques, 1860. 48 Dumas, Traite de Chimie appliquee aux arts. Paris (1828-46). 104 HISTORY OF CHEMISTRY. [LECTURE VI. gases (at the same temperature and pressure) there is the same number of (physical) atoms, but that these are still divisible by , chemical means. "We call atoms, the groups of chemical molecules that exist isolated in the gases. The atoms of the elementary gases always contain a certain number of molecules which is unknown to us." 49 The ratio of the densities of the gases gives Dumas the ratio of their atomic weights. In fixing the atomic weights of the solid elements he makes use of the law of Dulong and Petit, which he regards, accordingly, as hold- ing for groups of chemically smallest particles — molecules, as we should now say. Further, he employs for the same pur- pose, the relative densities of volatile compounds, making assumptions, from analogy, as to the volume relations of the unknown elementary gases contained in them. Thus he finds the atomic weight of sulphur from the density of sulphuretted hydrogen, which he assumes to be constituted like water and to consist of 2 volumes of hydrogen and i of sulphur vapour ; and that of phosphorus from phosphuretted hydrogen, which he supposes to be constituted like ammonia. His determina- tion of the atomic weight of carbon is noteworthy. He deduces it from the relative densities of ethylene and of marsh gas. He assumes in the latter (as Gay-Lussac had also done previously) 2 volumes of hydrogen for i of carbon vapour, and in the former, equal volumes of the two. He thus finds the atomic weight of carbon to be one half of what Berzelius had estimated it to be, that is 6, if that of hydrogen is assumed equal to i. In general, however, the values which he assigns to the atomic weights of the better known elements are the same as those of Berzelius. Mercury, silicon, etc, form exceptions. Dumas does not state the weights of the chemically smallest particles. Berzelius contested the principles of the system just con- sidered, although they were so closely related to his own.50 He thinks that it is absurd to assume fractions of atoms, and says it was formerly the custom to abandon hypotheses as soon 49 Dumas, Traite. I, 41. 50 Berzelius, Jahresbericht 1828, 80. LECTURE VI.] HISTORY OF CHEMISTRY. [05 as they led to absurdity. Dumas stands entirely isolated in his views, but, in spite of this, he would probably have adhered to them, had he not himself discovered facts which caused him to doubt the accuracy of Avogadro's hypothesis. Dumas was not only an ingenious thinker, but he was also an excellent experimenter ; and, since he had chosen the den- sities of gases and vapours as the basis of his atomic theory, he thought it necessary to increase our knowledge respecting these densities. He succeeds in elaborating a method for carrying out determinations of this kind at high temperatures, and em- ploys it for ascertaining the relative densities of the vapours of iodine, phosphorus, sulphur, mercury, etc.51 His results, from which he anticipated confirmation of his views, lead him to abandon them. He finds the density of phosphorus vapour to be twice as great, and that of sulphur vapour to be three times as great as he had previously assumed, whilst that of mercury vapour is only one half of what he had supposed. In view of these facts he begins to doubt ; in fact he declares that even the simple gases do not contain, in the same volume, the same number of chemical atoms. According to him, the assumption may still be made that there are the same number of molecular or atomic groups present in equal volumes of all gases ; but that this is only a hypothesis, which cannot be of any service.52 Dumas is obliged to admit that Gay-Lussac's law, when applied in the manner he had done to the determination of atomic weights, furnishes erroneous results. Hence he believes that it cannot be employed for this purpose, and he now abandons Avogadro's hypothesis. Berzelius, too, can no longer maintain the identity of volume and atom in the cases of the elementary gases, and has to confine his proposition to the incondensable elastic fluids.53 It must be admitted that the law, when so stated, was not capable of any extended application, and was more than insufficient for the determination of the atomic weights of the 51 Ann. Chim. [2] 33, 337; 44, 288; 49, 210; 50, 170. 52 Le9ons. 268 and 270. 53 See p. 92. 106 HISTORY OF CHEMISTRY. [LECTURE VI. majority of ihe elements. And how did it fare with the other generalisations upon which his system rested ? The hypothesis of Dulong and Petit was not without exceptions in its appli- cability, as I have already stated. The numbers deduced from it for silver, cobalt, and tellurium were not in harmony with Berzelius' determinations, that is, with the atomic weights required by the chemical analogies, and by isomorphism ; so that even this hypothesis was not tenable when rigidly con- sidered. Mitscherlich's law still remained, and the majority of chemists believed that it permitted of an unerring conclusion as to the atomic constitution. Other voices were heard, how- ever, which indicated doubts, especially after Mitscherlich had shown that there are dimorphous substances, i.e. substances which can occur in two crystalline forms.54 Attention was drawn to the fact that, as the occurrence of dimorphism proved, the crystalline form of a substance was not determined solely by the number of its atoms.55 Of all the physical laws that had been applied to the determination of atomic weights, there thus remained not one upon which full reliance was placed. The conception of the atom was looked upon, in consequence, as uncertain and hypothetical. Chemists believed they would have to be contented with the combining weight or the equivalent, the latter of which had gained new support from Faraday's electrolytic law.56 At the end of the fourth decade of the century, we thus find the atomic theory— the most brilliant theoretical achievement of chemistry — abandoned and dis- credited by the majority of chemists, as a generalisation of too hypothetical a character. A new school had arisen, which had adopted Wollaston's equivalents, and which sought, successfully, to supplant the system of Berzelius. At the head of this movement there stands L. Gmelin. The views of this chemist are of all the more importance from the fact that he expounded them in his excellent Hand-book • 54 Ann. Chim. [2] 24, 264. 55 Ibid. [2] 50, 171. ** Experimental Researches in Electricity, Series 3, § 377, Series 7, § 783 et scq. 1833-34. LECTURE VI.] HISTORY OF CHEMISTRY. 107 f which latter, on account of its completeness, had, at that time, already become widely distributed. With Gmelin, there is no strict distinction between mixtures and compounds, and this proves that he does not believe in the real existence of atoms. Two substances, especially when they possess only a weak affinity for each other, can combine, according to him, in an infinite number of proportions ; but the greater the affinity, the greater is their tendency to combine in few proportions only.57 These proportions then stand to each other in simple relations. "There can therefore be assigned to every substance a certain weight in which it combines with definite weights of other elements. This weight is the stochiometric number, the chemical equivalent, the mixture-weight or atomic weight, and so on. Compounds are composed in such proportions that one mixture-weight of one substance is united to J, \, \, f, }, i, i£, 2, 2 J, 3, 4, 5, 6, 7, or more mixture-weights of the other." According to Gmelin, Gay-Lussac's law runs : — One measure of an elastic fluid substance combines with i, ij, 2, 2^, 3, 3^, and 4 measures of the other. His table of equivalents is well known. It ran: — H=i, O = 8, S=i6, C = 6, etc. Water was written HO, and in formulae generally, the endeavour was made to replace by simplicity what they had lost in conception and in purpose. Chemistry was to become a science confined to observation — indeed almost to description alone. Skill in manipulation was all that was required ; speculation was banished as dangerous. It had come to this then :— Inorganic chemistry, in connec- tion with physics, had not been able to maintain the conception of the atom. It is my business to show, in the next lectures, how it was reintroduced into the science by means of organic chemistry. 57 Handbuch der theoretischen Chemie. Second Edition, 1821. LECTURE VII. ORGANIC CHEMISTRY AT THE COMMENCEMENT OF ITS DEVELOPMENT — ATTEMPTS TO DETERMINE THE ELEMENTARY CONSTITUENTS OF ORGANIC COMPOUNDS — ISOMERISM AND POLYMERISM — VIEWS REGARDING CONSTITUTION — RADICAL THEORY. IN this lecture I shall endeavour to give an account of the development of organic chemistry. I have intentionally post- poned this subject until now because I wished to consider it in a connected manner, because it had almost no influence during the first three decades of the present century upon the perfect- ing of general theories, and also because the views which constitute the basis of inorganic chemistry did not, at first, seem to be capable of any application to the organic branch of the science. Thus we £nd Berzelius, in 1828, treating the subject of the organic compounds separately. The electro- chemical theory, the law of multiple proportions, and the law of volumes did not appear to dominate substances derived from the animal and vegetable kingdoms ; these substances were subject to the so-called vital force, the nature of which was wholly unknown and obscure. It only became possible to extend to organic chemistry also, the laws which held for inorganic substances, after the study of this part of the subject had further attracted thinkers to itself. The opinions and hypotheses to which the examination of the better known substances had led, had now to be turned to account in the younger branch of the science. It was dualism, in particular, which was now introduced into organic chemistry also. Lavoisier had already assumed, as has been stated previously, that the acids consist of oxygen and a basis ; and that, in inorganic compounds, the latter is an element, while in organic LECTURE VII. J HISTORY OF CHEMISTRY. lOQ compounds it is a compound radical. The latter name had not been lost. The chemical nomenclature had been based upon it, and the dualism of Berzelius was a happy extension of it. All observations seemed to be in agreement with it. Thus the salts (at that time the best known class of substances) are formed from acid and base, and they can be decomposed again into these constituents. Why should we not endeavour to look upon organic compounds also as formed in a similar way ? Since organic compounds (according to the view held at that period) consisted of at least three elements, any simple de- composition must still leave one part of a composite nature. Organic chemistry thus became the Chemistry of the Compound Radicals. In thus employing the term radical, its original definition was retained. After the removal of the oxygen from a substance, the residue which remained, and which, moreover, played the part of an element, was called a radical. Wohler and Liebig remodelled this conception. In their admirable research on bitter almond oil and the allied compounds, they showed that we may assume the existence, in these substances, of an oxygenated group which remains unchanged in the majority of the reactions, and therefore behaves like an elementary substance. On this account, they called it the radical of bitter almond oil. By this departure the first great step had been taken ; organic chemistry had become independent ; it had freed itself from the fetters which had been placed upon it ; if, from within its own limits, it had not produced a new idea, it had at least given new and increased importance to one which was already held. From this time forward, it proceeds on its own way, and pays no heed to the limitations which some desired to place upon it. The very harmonious edifice of chemistry suffers in consequence. Every endeavour is made to adapt it to the new ideas. But it is in vain — the breach is unavoidable. The young science, quite conscious of its own strength, dares to make an assault upon the foundations, and, in spite of stays and props, the structure begins to totter. The attack upon the electro-chemical theory led to an embittered controversy 110 HISTORY OF CHEMISTRY. [LECTURE VII. between its supporters, with Berzelius at their head, and the adherents of the substitution theory or the theory of types. This controversy was triumphantly passed through by the latter, and led to the complete separation of organic and inorganic chemistry. At any rate the endeavour was still made to retain in the latter, as before, the dependence of the chemical upon the electrical forces, whilst the newest facts in the domain of organic chemistry appeared to be incompatible with this. Our science thus fell anew into two schools, and the principles which guided the one school were rejected by the other. Simultaneously with the abandonment of the electro- chemical hypothesis, the radical theory was also given up ; there was now no longer any actual need for it, and, in the form in which it had been advanced, it was insufficient. A great deal had been discarded as useless, and therefore it is by no means inadvisable to inquire into the principles which still remained with the representatives of the new school. The views as to the preservation of the type, and as to substitution, although, of course, most valuable for the comprehension of many reactions, could scarcely be employed as the basis of a complete system. But amongst the ruins left upon the battle- field, and found there when it was cleared up, there was a jewel, which, although little heeded during the controversy, was now capable of becoming of great significance when the question was no longer one of getting rid of old views, but one of setting up new views in their stead. The atomic theory, despised by many, forgotten by some, was now destined to arise again in its original brilliancy, although a hard struggle was necessary. New foundations for the determination of the relative masses of the atoms had to be obtained. It was Gerhardt, in particular, who insisted on the necessity of fixing upon comparable quantities for these determinations. But whence was the standard to be derived? Liebig's polybasic acids, and Dumas' substitution, had at length taught chemists the difference between atom and equivalent, so that there could no longer be any question with respect to the latter. Recourse was again taken to Avogadro's hypothesis; but this LECTURE VII.] HISTORY OF CHEMISTRY. Ill still proved insufficient, and chemical reasons were required in order to convince chemists. Gerhardt, who received very substantial support from Laurent, exerted himself in vain to adduce decisive proofs of the accuracy of his ideas. At this juncture, Williamson's investigations appeared, and they gave a real foundation to the thoughts that had flitted before Gerhardt's mind. This gifted chemist had shown the way; and, in imitation of him, it was extensively followed, since it permitted of a direct comparison of the quantities entering into reaction. Thus there arose the conception of the chemical molecule. In Gerhardt's system, which was now rapidly gain- ing recognition, this conception found its formal expression in the theory of types. I shall here conclude this sketch, in which I have indicated, in general outline, the different phases of the historical develop- ment ; and I shall now proceed to a detailed account. As early as the second half of the seventeenth century, Lemery separated organic from inorganic chemistry. He divided substances, according to their origin, into three classes, viz., mineral, animal, and vegetable.1 The phlogistians occupied themselves chiefly with the first class. Scheele deserves to be mentioned as the discoverer of an extensive series of organic substances.2 Lavoisier believed that compounds belonging to this class consisted of carbon, hydrogen, and oxygen ; Berthollet proved the presence of nitrogen in substances of animal origin ; 3 at a later date, it was recognised that all the elements can enter into organic combinations, but that carbon must never be absent.4 It is difficult to say what substances were regarded as organic compounds at the beginning of the present century. This class naturally included all substances occurring in the 1 Kopp, Geschichte. 4, 241. 2 See pp. 10-11. 3 Journ. de Phys. 28, 272. 4 Gmelin, Handbuch der Chemie. Fourth Edition, 4, 3 ; E. 7> 4- 112 HISTORY OF CHEMISTRY. [LECTURE VII. plant or animal organism, but from these there were prepared a large number of other compounds, whose position in the system had also to be determined ; and a decision as to which class they were to be enumerated in was often an arbitrary matter. Simplicity of composition was frequently a reason for placing substances in the inorganic class. With regard to many substances, the views as to their nature had changed in course of time ; for example, in the case of the cyanogen compounds, which were first classed as organic and afterwards as inorganic substances. Wherever it was possible, Lavoisier's idea was upheld that, in organic substances, the basis com- bined with oxygen, or the radical, consists of several elements. This gave rise, at a later date, to Liebig's definition of organic chemistry as the chemistry of the compound radicals. The study of the compounds belonging to this class lagged considerably behind that of the others. The reason lay partly in the easy alterability of these substances, and, thus, in the greater difficulty which their isolation presented ; partly also in the scarcity of methods for th^ir analysis. At the beginning of this century, when qualitative analysis had already attained a high degree of accuracy, and even the quantitative method had found excellent exponents in Proust, Klaproth, and Vauquelin, Lavoisier's experiments with alcohol, oil, and wax, were the only ones in existence, designed to ascertain the composition of organic compounds ; and these, as may easily be understood, were not very accurate. It is thus explicable that Berzelius should still doubt, in 1819, whether the law of multiple proportions held in organic chemistry also.5 He was well aware that when organic com- pounds unite with inorganic ones — organic acids with metallic oxides, for example — -the same regularities are observed as in inorganic chemistry ; but he believed the proportions in which carbon, hydrogen, oxygen, and nitrogen unite, to be so varied that Dalton's law lost its significance, simply because i, 2, . . n atoms of one element could unite with i, 2, . . m atoms of 5 Essai etc., 96; compare also Lehrbuch. 3, part I, 151. LECTURE VII.] HISTORY OF CHEMISTRY. 113 another. Nevertheless Berzelius himself afterwards assisted more than anyone else in extending the laws of stochiometry to organic chemistry, inasmuch as he materially improved the method of elementary analysis employed at that time, and thereby provided himself and others with the means of ascer- taining the composition of organic substances. It may not be out of place to give some account here of the history of elementary analysis, just because the views respecting organic compounds were essentially changed in consequence of its development. I shall not again revert to Lavoisier's method, which I have already indicated in an earlier lecture.6 There are almost thirty years between his experiments and those of his nearest successors. I pass over the experiments of Saussure,7 Ber- thollet,8 etc., as well as the first labours of Berzelius 9 upon this subject. These furnished analytical processes that were sufficient, perhaps, in special cases, but cannot by any means be regarded as general methods. On the other hand, the investigation of Gay-Lussac and Thenard, in 1811, deserves our attention.10 They burned the organic substance with potassium chlorate, by forming small pellets of the mixture and allowing these to fall into a perpendicularly placed tube, the lower end of which was heated red-hot. The tube was closed above by means of a stop-cock furnished with a recess designed for the reception of the pellets. The gases from the combustion had to make their escape through a side tube into a eudiometer, and they were there measured. Gay-Lussac and Thenard then absorbed the carbonic anhydride formed, and determined the oxygen left behind. They knew, further, the quantity of sub- stance burned and the quantity of potassium chlorate mixed with it. By the help of Lavoisier's equation : — Substance + Oxygen employed = Carbonic anhydride + Water 6 See pp. 28-29. 7 Journ. de Phys. 64, 316 ; Bibliotheque britannique. 54, No. 4 ; 56, 344 ; Ann. Phil. 4, 34. 8 Mem. d'Arcueil. 3, 64 : Mem. de 1'Acad. 1810, 121. 9 Gilb. Ann. 40 (1812), 246. 10 Rech. phys. chim. 2, 265. H 114 HISTORY OF CHEMISTRY. [LECTURE VII. they were then able to calculate the quantity of water pro- duced by the combustion, and, therefore, the composition of the compound. Gay-Lussac and Thenard carried out the analysis of twenty substances in this way. Their results are moderately accurate, but the method still left much to be desired. The combustion was very violent ; it was accompanied by explosion, and was, therefore, frequently incomplete. The next great step in the development of elementary analysis was made by Berzelius in i8i4.n By carrying out the combustion with a mixture of potassium chlorate and sodium chloride, he secured the much more moderate progress of the analysis. His method, also, differs very essentially and advan- tageously from the earlier one, in so far that he did not gradually introduce the substance intended for combustion into a red-hot tube, but instead, put the whole quantity of the substance, along with the oxidising material, into a tube which he gradually heated to redness in a horizontal position. Further, he was the first to weigh the water directly, which he did after having absorbed it by means of calcium chloride ; whereas he deter- mined the carbonic anhydride either by volume or by weight. This mode of carrying out the analysis already approximates closely to the present method ; it was still further improved by employing cupric oxide instead of potassium chlorate. This was first used by Gay-Lussac for nitrogenous substances,12 but a year afterwards it was employed by Dobereiner in the com- bustion of substances free from nitrogen.13 Analyses were carried out by this process for more than ten years, until the method was modified by Liebig14 in 1830, and brought into the form now employed. As a consequence of Liebig's investigations, elementary analysis became an easily 11 Ann. Phil. 4, 330 and 401. 12 Ann. Chim. 95, 154; Ann. Phil. 7, 357- 13 Schweigger's Journal. 17, 369 ; compare also Chevreul, Re- searches chimiques sur les corps gras d'origine animale (1823). 14 Pogg. Ann. 21, i ; more fully in his " Anleitung zur Analyse organischer Korper." Braunschweig (1837); English Translation, by Gregory: "Instructions for the Chemical Analysis of Organic Bodies," Glasgow (1839). LECTURE VII.] HISTORY OF CHEMISTRY. 115 accomplished operation, which, in so far as accuracy was con- cerned, might be placed alongside of other analyses. A rapid advancement of organic chemistry dates from this period. Since a simple and sure means of determining the composition of the substances was now available, investigations which had not previously been attempted on account of the endless difficulties associated with them, now became possible and were actually carried out. No doubt many analyses had already been carried out by the method of Berzelius, and the conviction became more and more settled that the law of multiple proportions was applicable to organic compounds also, and that formulae similar to those assigned to mineral substances could be assigned to them. But an important distinction was still drawn, in the third decade of this century, between these two classes of substances. It was supposed that the latter alone were producible artificially; while the synthesis of the former was wholly beyond our power and was reserved for the living organism, in which it was per- formed under the influence of the Vital Force. From such naturally occurring substances chemists had, it is true, learned to prepare, by dry distillation, by treatment with nitric acid, with alkalies, etc., other substances which were likewise classed amongst organic compounds, but these were, for the most part, simpler in composition, and the material existing in nature always remained the starting point. In this connection, an excellent investigation for that period, by Chevreul, deserves to be mentioned,15 in which the author showed that the fats consist of an acid and of glycerine (a substance discovered by Scheele), and that they should, accordingly, be placed in the series of ethers, where all those substances were classed which could be separated by means of alkalies into an acid and an indifferent substance (an alcohol). This and similar investigations could not, however, shake the belief in a vital force under whose influence all organic compounds originated. As yet no one had succeeded in arti- 15 Rech. Chim. Il6 HISTORY OF CHEMISTRY. [LECTURE VII. ficially preparing any substance occurring in the organism : but even this great step had not to be waited for much longer. For this discovery we are indebted to Wohler, and with it he opened his long and brilliant scientific career. Wohler had discovered cyanic acid in i822,16 and was occupied in its investigation when he made the observation, in 1828, that urea, a known product of animal life, was formed upon the evaporation of a solution of its ammonium salt.17 It is true that the problem was not completely solved by this discovery. The synthesis from the elements was not yet possible, but still the most essential thing had been accom- plished. From inorganic compounds (amongst which many at that time classed cyanic acid)18 a substance had been prepared which had hitherto been found in the animal organism only. In spite of this, the revolution of ideas proceeded but slowly ; it was still believed that the vital force could not be dispensed with, and some decades afterwards, scientific discussions took place as to its existence. Nowadays, when the materialistic tendency becomes more and more ascendent, there are few to be found who ascribe the production of organic substances to forces different from those that govern the production of mineral substances. It is true that the experimental science has made great progress in this respect also, since it has succeeded in preparing from their elements many organic sub- stances. Thus Kolbe effected the complete synthesis of tri- chloracetic acid,19 and Berthelot the syntheses of formic acid and of alcohol. The latter chemist inaugurated, with these researches, his brilliant series of synthetical investigations.20 It may appear remarkable to many persons, into whose hands a treatise on organic compounds published in the third decade of this century, or earlier, may chance to fall, that even at that time, when this department of chemistry was in 16 Gilb. Ann. 71, 95. 17 Fogg. Ann. 12, 253 ; Quart. Journ. Science. 1828, 1, 491. 18 Compare, for example, Dumas, Traite. I, 409. 19 Annalen. 54, 145. 20 Berthelot, Chimie organique fondee sur la Synthese, Paris (1860) ; see also his more recent investigations, Bull. Soc. Chim. [2] 7, 113, 124, 217, 274, 303, 310, etc. LECTURE VII.] HISTORY OF CHEMISTRY. 117 so backward a state of development, experiments were made in order to obtain some information as to the constitution, or mode of arrangement of the atoms, of compounds. A pursuit of this kind may be regarded as idle speculation, and yet scien- tific chemistry was directed, at an early period, towards such considerations. This was owing to the phenomena of isomerism, into which, therefore, I must here enter with some detail. After chemists had begun to pay attention to the quanti- tative composition of substances, and especially after they had learned to regard constant proportions by weight of their con- stituents as a real characteristic of chemical compounds, it was assumed as a matter of course that the same composition per cent, always postulated the same properties. It was, of course, known that very many, and indeed most, substances occurred in several states : solid, liquid, and gaseous ; crystalline and amorphous; etc., but the sensation that the discovery of dimor- phism made, shows us how great the tendency was at that time to regard physical and chemical properties as functions of the composition per cent, (and of the temperature). It must naturally have created much surprise to see that sulphur can appear in two crystalline forms ; to hear that arragonite is pure calcium carbonate, and is therefore dimorphous with calc-spar ; etc.21 It was to be shown, however, in the same year as that in which the dimorphism of sulphur was recognised (1823) that even the chemical properties can change without alteration of composition. In the analysis of fulminic acid, Liebig obtained numbers which agreed exactly with those established for cyanic acid.22 This was at first believed to be an error, but subse- quent examination confirmed the observation, and the great difference between the two substances seemed wholly inex- plicable. Two years later, Faraday discovered another fact of the same kind.23 He was engaged in the examination of oil 21 Ann. Chim. [2] 24, 264. ** Ibid. [2] 24, 294 ; 25, 288 ; Schweigger's Journal. 48, 376. 23 Phil. Trans. 1825, 440; Ann. Phil. 27, 44 and 95 ; Schweigger's Journal. 47, 340 and 441. Il8 HISTORY OF CHEMISTRY. [LECTURE VII. gas, when he discovered a hydrocarbon that behaved very like ethylene; but it yielded no chloride of carbon when mixed with chlorine and exposed to sunlight, and it possessed, more- over, a density double that of ethylene.24 Further, an investi- gation of phosphoric acid was made at this period by Clark, who, through neglecting the fact that the salts contained water, was led to the opinion that there were two phosphoric acids, with different properties but with the same composition.25 Berzelius had previously observed the same thing with respect to stannic acid.26 He also showed in 1830 that the acid pro- duced along with tartaric acid during the manufacture of the latter, had the same composition as tartaric acid. He calls the new substance racemicacid (Drufsyra. Traubensiiure.}, and introduces the word isomer to designate substances of this kind. According to him, this word is only to be applied to compounds possessing the same composition and the same atomic weight, but with dissimilar properties.27 A year later, Berzelius designates as polymerism the phenomena observed by Faraday respecting the hydrocarbons. This name embraces those cases where the same composition is accompanied by dissimilar properties and different atomic weights.28 Metameric substances, on the other hand, are those which possess the same composition, the same atomic weight, and dissimilar properties, when the difference can be explained by a different arrangement of the atoms — i.e. by a different constitution.29 As an example, Berzelius very appropriately chooses stannous sulphate and stannic sulphite, which he writes : SnO + SO3 and SnO2 + SOZ. At that time, the different modifications of an element were also regarded as cases of isomerism ; and it was only in 1841 24 I mention here, in passing, that Faraday on the occasion of this investigation also discovered benzene. 25 Edinburgh Journal of Science. 7, 298; Schweigger's Journal. 57, 421. 26 Ibid. 6, 284. 27 Pogg. Ann. 19,' 305. 2S It appears from this that Berzelius at that time regarded the vapour densities of compounds also as a guide to their atomic weights. 29 Berzelius, Jahresbericht, 1833, 63- LECTURE VII.] HISTORY OF CHEMISTRY. 119 that Berzelius introduced the word allotropy to designate such cases.30 A number of examples belonging to this class were already known ; one of the most interesting being carbon, in the forms of diamond, graphite, and soot. It will be understood that the idea of metamerism could only be introduced after it had become possible to entertain any conception of the constitution of a substance ; while, on the other hand, the phenomena of isomerism necessarily led chemists to hypotheses respecting the mode of arrangement of the atoms. It is well known that there was in existence at that time a mode of regarding the facts which Berzelius endeavoured to extend more and more. I refer to dualism, which I have already had occasion to mention several times, and the conse- quences of which I shall now state more definitely. The phenomena of combustion had led Lavoisier to assume, in so far as it was possible to do so, that substances consisted of two parts. This way of regarding them was very advan- tageous and clear in the case of salts, which were looked upon as composed of base and acid. This view was in agreement with their whole behaviour, and rendered it possible to con- sider them all from one common standpoint. The arguments of Gay-Lussac and Thenard against the elementary nature of chlorine, which have been stated in a previous lecture,31 prove how deeply these ideas had become rooted, and how firmly it was the custom to base conclusions upon them. After the existence of the so-called hydrogen acids (i.e. of acids which do not contain oxygen) had been generally admitted, various opinions arose respecting the nature of their salts. A few investigators (Davy and Dulong, for example) regarded them as compounds of metals, just as they regarded other salts;32 but this view met with little approval at that time. Others remained true to the earlier conception, and with them common salt was still muriate of soda, which had the peculiarity, however, that it gave off its " water." Others, 30 Berzelius, Jahresbericht, 1841, Part II. 13. 3l Seep. 81. 32 See p. 83. 120 HISTORY OF CHEMISTRY. [LECTURE VII. again, no longer looked upon these substances as salts, but compared them with the oxides. In this connection, the double chlorides and iodides prepared in 1826 by Boullay, caused the latter to develop his ideas more fully.38 In accordance with these ideas, the chlorides, iodides, etc., of the alkali metals were bases, from which true salts were only produced by com- bination with the chlorides and iodides of the heavy metals ; and the latter, in turn, were analogous with the acids. Others still, and amongst them Berzelius,34 whose opinion at that time carried the greatest weight, regarded common salt and similar substances as compounds possessing a salt-like character ; but they separated them from ordinary salts. According to them, the whole group of salts consisted of two divisions — the amphid salts, to which the oxygen, sulphur, etc., salts belonged, and the haloid salts, which embraced the chlorides, iodides, etc. The latter were composed of two elements or radicals — of a metal and a halogen, as chlorine, iodine, cyanogen, etc., at this time came to be called. It remained, however, altogether un explained why substances with such similar properties as those of the amphid salts and the haloid salts, possessed such different constitutions. If it was desired to regard the oxygen salts as compounds of an acid with a base, then the establishment of these ideas was further attained as follows. Taking nitre as an example, KO would represent the base, and N2O535 the acid (or what we now call the anhydride). It thus came about that acetic acid was represented by C4H6O3, formic acid by C2H2O3, sul- phuric acid by SO3, etc. ; that is, instead of the actually exist- ing substances, others were represented, of which some were imaginary. The free acids were held to contain "a proportion of water which we cannot separate except by combining the acid with another substance " ; 3G and although Berzelius him- self had previously distinguished water of hydration from water contained in salts and not necessary for their existence,37 yet 33 Ann. Chim. [2] 34, 337 ; Journ. de Pharm. 12, 638. 34 Berzelius, Lehrbuch. Third Edition, 4, 6. :?5 Berzelius' atomic weights : see p. 100. a6 Berzelius, Lehrbuch. Third Edition, 2, 4. 3r Journ. de Phys. 73, 253. LKCTURE VII.] HISTORY OF CHEMISTRY. 121 this water also was neglected, as if non-existent, in most of the discussions as to the constitution of bases and acids. It may have been a consequence of this that the presence of water was assumed even in substances which belonged to other classes, when hydrogen and oxygen were found in them in the propor- tions necessary to form water, and that this water was then neglected in writing formulae for these substances. Many things might be adduced as having contributed to erroneous ideas of this kind ; such, for example, as the way in which Gay-Lussac and Thenard in 1811 interpreted their analytical results relat- ing to organic substances.38 According to these chemists, substances fall naturally into (i) those which contain just as much oxygen as is required in order to form water with the hydrogen present (carbohydrates) ; (2) those which contain less (resins, oils) ; and (3) those which contain more (acids). I regarded these perhaps seemingly detailed explanations as necessary, before I could enter more minutely into the views respecting the constitution of organic compounds. In passing on now to this most important question, I wish to show how dualism was gradually introduced here also, and how the radical theory arose as a consequence of this. Berzelius explained in 1819 that his electro-chemical theory/ could not be extended to organic chemistry, because under the influence of the vital force the elements there possessed entirely different electro-chemical properties. In decay, putrefaction, fermentation, etc., he observes phenomena which he regards as demonstrating the tendency of the elements to return to their normal condition.39 He did not, at that time, as yet consider it possible to regard all organic substances as binary groups. Dualism was, indeed, extended as far as possible ; the oxygen compounds were looked upon as " oxides of compound radicals, which, however, do not exist free, but are wholly hypothetical,"40 a mode of regarding the matter which was especially applicable to the acids. Accordingly, we now hear the radicals of acetic 38 Rech. phys. chim. 2, 265. 39 Essaietc., 96. 40 Berzelius, Lehrbuch. First Edition, 3, Part I. 149. 122 HISTORY OF CHEMISTRY. [LECTURE VII. acid, C4H6, of benzoic acid, C14H10, etc., spoken of, and these radicals are the remainders of .the acids after the deduction of their oxygen. It is easily understood that endeavours should be made from other points of view, and in other directions, to establish hypotheses regarding the nature of organic substances ; but I may pass over those which possessed no general significance and were applicable to a few substances only. I must, however, adduce one example of this kind, because the idea involved was for a long time held in respect in chemistry. It has to do with a conception of oxalic acid, which was then written C2O:1, the elements of water being neglected. Dobereiner, who carefully studied the behaviour of the oxalates in 1816, proved that some of them give off carbonic anhydride and carbonic oxide when heated, and on this account he thought himself justified in regarding the " acid of sorrel " as carbonate of carbonic oxide.41 This was an attempt to refer back complex substances to simpler ones, and it possesses a certain significance, in so far that it is based upon facts. More important by far is an observation of Gay-Lussac's concerning the composition of alcohol and of ether, which dates from about the same time,42 and which became the basis of the so-called Etherin Theory. The discoverer of the law of gaseous volumes points out that the densities of the vapours of alcohol, ether, and water, and the density of olefiant gas, stand to each other in such a relation that ether may be regarded as com- posed of half a volume of water vapour and one volume of olefiant gas, and alcohol as composed of equal volumes of the two. Dumas and Boullay adopted this observation as the basis of the views regarding the ethereal compounds, which they advanced in 1828 on the occasion of a detailed investigation of these substances.43 In their view olefiant gas is a radical ; that is, a group of atoms which enters into combinations in 41 Schweigger's Journal. 16, 105. 42 Ann. Chim. 91, 160 ; 95, 311. * Ibid. [2] 37, 15." LECTURE VII.] HISTORY OF CHEMISTRY. 123 the same way that the elements do. They compare it with ammonia, and are at pains to show that, just as the latter is the radical of the ammonium salts, defiant gas must be assumed in the ethers. In doing so, they try to carry the analogy so far that they even assert that ethylene possesses basic properties, and that the reason why it does not colour litmus tincture blue is merely because it is insoluble in water ; and that its alkaline nature is proved, moreover, by its property of neutralising hydrochloric acid, whereby hydrochloric ether, observed so long ago by Basil Valentine, is produced. They then show, by means of a table, how the radical C4H4 or 2C2H2 (olefiant gas) may be assumed in the formulae of the ethers analysed by them • whereby complete uniformity with the ammonium salts is attained : — 44 Olefiant Gas - - - 2C2H Hydrochloric Ether Ether- - - - 4C,H2 + H2O Alcohol - - - 4C2il2 + 2H2O Acetic Ether 4C2H, + C8H663 + H2O NH3 ... Ammonia NH34-HC1 - Sal-ammoniac 3 + C8H6O3+H2O Acetate of Ammonia Oxalic Ether - 4C2H2 + C4O3 + H2O Oxalate of Ammonia, etc.45 We find the opinion that the ethers are to be regarded as analogous to salts, first advanced by Dumas and Boullay, although it is true that they did not adopt the usual view, in accordance with which salts do not contain any water. The endeavour to classify the organic compounds in the same way as the inorganic ones, constituted the basis of their views, how- ever ; and, since this idea was found to be applicable to a whole class of substances, it was thus of great importance. The point of view was distinctly dualistic, but not quite in the former sense. Accordingly we find that Berzelius at first maintains a very cautious attitude towards it ; 40 he finds in it at best a 44 Here, as in all other cases, I quote the formulae of the authors, and hence I employ in this case Dumas' atomic weights which are referred to H= i : O= 16, C = 6, etc. 45 The table given in the paper quoted above contains obvious misprints ; compare Dumas, Traite. Organic Part, I, 68. 46 Berzelius. Jahresbericht, 1829, 286. 124 HISTORY OF CHEMISTRY. [LECTURE VII. symbolic mode of expression, which cannot be regarded as representing the actual composition of the substances. Only some years afterwards does he revert, for a short time, to Dumas' ideas, and he then calls the radical C4H8 Etherin.47 This appears to me to be the place to state the results of an investigation of Gay-Lussac's into the cyanogen compounds, which had been carried out as early as 1815, and contributed materially to giving a more definite meaning to the conception of a radical.48 Gay-Lussac repeated Berthollet's experiments on the composition of hydrocyanic acid and confirmed them, inasmuch as he established beyond all doubt that the acid is free from oxygen, and contains carbon, nitrogen, and hydrogen only. The examination of the salts led him to study the behaviour of mercuric cyanide at high temperatures, and thus to discover cyanogen. What is of importance for us in Gay- Lussac's work, is the way in which he regards the substances he describes. These are, in his view, compounds of a radical containing carbon and nitrogen (cyanogen) and identical with the gas obtained from mercuric cyanide. The possibility of preparing radicals was in this way demonstrated, and, in con- sequence, the conception attained a more real significance. It is further to be remarked that Gay-Lussac, in calling the radical of hydrocyanic acid, cyanogen, permitted himself a certain freedom, since it was not actually " the residue of an acid which has been deprived of its oxygen." Obviously the great French scientist compares hydrocyanic acid with hydro- chloric acid, and with hydriodic acid which he had himself discovered a short time previously. They are hydrogen com- pounds of elements or radicals, exactly as the ordinary acids are oxygen compounds. If it was now desired to define a radical, and to include cyanogen in the definition, it was no longer possible to say, with Lavoisier, that " it is the residue of a substance which has been deprived of its oxygen " ; but it was the other half of the definition that was to be accen- tuated; "a radical is a composite group which behaves like an 47 Annalen. 3, 282. 48 Ann. Chim. 95, 136. LECTURE VII.] HISTORY OF CHEMISTRY. 125 element." 49 As a consequence of Gay-Lussac's investigation, and of the isolation of cyanogen, this latter view had acquired an increased significance. Reflections of a similar kind do not appear to have occurred further to the chemists of that period. Generally speaking, radicals were only looked for in oxygen compounds, and especially in acids ; whereas Dumas' and Boullay's assumption of Etherin proves, on the other hand, that attention was not confined exclusively to these. Wohler and Liebig's investigation of bitter almond oil and its derivatives, in i832,50 had a pronounced effect on the views respecting radicals. It led these two chemists to the assump- tion of a radical containing oxygen, and thus added an entirely new significance to these ideas. Wohler and Liebig first show that the conversion of bitter almond oil into benzoic acid consists in the taking up of oxygen, since they establish the formulae C14H12O2 and C14H12O4 respec- tively for the two substances.51 In doing so, they assume, however, an atom of water, H2O, in the latter; but they neglect this and write the formula of benzoic acid C14H10O3. In this way they come to look upon both substances as com- pounds of the radical Benzoyl, CJ4Hi0O2, and to regard bitter almond oil as benzoyl hydride, and benzoic acid as an oxygen compound of the new radical. They show, in the course of the investigation, how the same radical may likewise be assumed in an extensive series of substances. By treatment of bitter almond oil with chlorine and bromine, they prepare benzoyl chloride and bromide, C14H10O2.C1. and C14H10O2.Br2. From these, by means of potassium iodide and of potassium cyanide, they obtain the iodine and the cyanogen compounds of the radical, C14H10O2.I2 and C14H10O2.Cy2. Finally, with ammonia and with alcohol, they obtain benzamid and benzoic ether. This investigation is regarded even now as one of the greatest achievements in the range of organic chemistry. The 49 Lavoisier, Oeuvres. I, 138. 50 Annalen. 3, 249. 51 Berzelius' atomic weights. 126 HISTORY OF CHEMISTRY. [LECTURE VII. impression which it produced at the time can be well under- stood. It was the first instance where, starting from one compound, an extensive series of well-defined substances had been obtained, the relations of which could easily be explained if the suggested mode of regarding them was adopted. We thus find Berzelius, then at the zenith of his fame and only seldom in accord with the views of others, bestowing abundant praise upon the investigation.52 He hopes that, as a conse- quence of it, a new day will dawn in chemistry, and he proposes to Liebig and Wohler to call the new radical Proin or Orthrin (break of day) because he believes that a clearer light will be cast upon our science by the assumption of ternary radicals. And Berzelius was right ! For even although the chief importance of the investigation does not rest with the radical composed of three elements, still the special part which had been universally attributed to oxygen since Lavoisier's time, even in this branch of chemistry, was taken away from it. Further, that the true signification of the word radical was to be sought for elsewhere, was demonstrated by the fact that, in the choice of a radical, the composition was left entirely out of consideration ; and the justification of this proceeding was found in the experimental results. Benzoyl was a radical be- cause, like an element, it combined with other elements, and because it could be transferred, without decomposition, from the compounds so formed, into others. It was the key to the interesting reactions of Liebig and Wohler, and it formed the foundation of the benzoic acid series just as cyanogen was the .basis of a large number of substances. Cyanogen and benzoyl are the pillars of the radical theory, which received confirmation by the discovery of cacodyl. I cannot enter into the details of this extremely difficult and brilliantly executed investigation of Bunsen's, but it is my duty to state the general results of his work. In 1760, Cadet had obtained a fuming liquid, possessing a 82 Annalen. 3, 282. LECTURE VII.] HISTORY OF CHEMISTRY. 12) nauseating smell, by distilling potassium acetate with arsenious anhydride.53 This liquid took fire spontaneously in air, and it was known to contain arsenic and to be poisonous. These properties appear to have deterred chemists from the study of it, for, with the exception of a few unimportant experiments by Thenard, they had not occupied themselves with it at all for seventy years, and had been content to mention it in text-books as Cadet's liquid. Dumas had then endeavoured, by distilla- tion, to separate a pure compound from the crude product, which was contaminated, amongst other things, with elementary arsenic. According to his analyses, it is represented by the formula C8H12As2 [C = 6, As = 75].54 Bunsen's first results appeared to confirm this,55 while later experiments eventually fixed it as C4H12AsoO [C = 1 2].56 Bunsen called the substance cacodyl oxide, and assumed the existence in it of the radical CjH12As2. He succeeded in preparing the chloride, bromide, iodide, cyanide, and fluoride by treatment with the corre- sponding acids ; the action of barium hydrosulphide produced the sulphide ; by oxidation Bunsen obtained cacodylic acid, C4H]2As2.O3 + H2O ; finally, he found it possible to isolate the radical cacodyl by decomposing the chloride by means of zinc, and, naturally, this assisted very materially in procuring recog- nition for his mode of regarding the matter. We can understand how keenly interest must have been aroused on hearing of the isolation of an organic radical containing a metal, and possessed, besides, of the extremely remarkable property of spontaneous inflammability. I have intentionally introduced here the account of this important research of Bunsen's (the completion of which falls at a later date) in order to be able to make clear the idea of a radical as it now gradually came to be conceived. This idea is essentially different from what was formerly understood by the term, and the new conception was brought to the front by 53 Mem. de Math, et de Phys. des savants etrangers. 3, 633. 54 Dumas, Traite. Organic Part, I, 135; compare Annalen. 27, 148. 53 Annalen. 24, 271. 56 Ibid. 31, 175 ; 37, i ; 42, 14 ; 46, i. 128 HISTORY OF CHEMISTRY. [LECTURE VII. means of a series of investigations, of which I have shortly stated the most important. In doing this, I was at pains to explain the development of the ideas, and I only desire now to be permitted to define the meaning of the term as it eventually became fixed in the minds of the chemists of the period. I begin with the celebrated definition of Liebig : 57 " We call cyanogen a radical," he says, in 1837, in his criticism of Laurent's theory, "(i) because it is a non-varying constituent in a series of compounds, (2) because in these latter it can be replaced by other simple substances, and (3) because in its compounds with a simple substance, the latter can be turned out and replaced by equivalents of other simple substances." These three requirements, of which, according to Liebig, at least two must be fulfilled in order that an atomic group may have any claim to the designation of radical, prove that only a study of the nature of a compound could lead to a knowledge of the radical contained in it. The behaviour of a substance towards elements and compound bodies required to be known, in order that its radical might be ascertained ; and from this it may be gathered what significance such a determination possessed. The choice. involved, to a certain extent, a resume of the whole investigation, since the decomposition products were known when the radical was known ; ihe latter was, of course, composite itself, but with its decomposition, those affinity relations ceased which connected with one another, substances containing the same radical. That the radical behaved like an element, had been confirmed over and over again. Not only did it enter into combinations with elements, but it could also be isolated from these combinations. How far this comparison was carried, is shown by a quotation taken from a joint paper by Dumas and Liebig : 58 " Organic chemistry possesses its own particular elements, which some- times play the part taken by chlorine and oxygen in inorganic chemistry ; sometimes, on the other hand, the part of the metals. Cyanogen, amide, benzoyl, the radicals of ammonia, of fatty 57 Annalen. 25, 3. 58 Comptes Rendus. 5, 567. LECTURE VII.] HISTORY OF CHEMISTRY. I2Q bodies, of alcohols and analogous bodies, these are the real elements with which organic chemistry operates, and not the ultimate elements carbon, hydrogen, oxygen, and nitrogen, which only appear when every trace of organic origin has disappeared." It will thus be understood that the atoms which constituted such a group were supposed to be held together by stronger forces than those which united the group to other atoms. The radical in this way attained a very real significance in the minds of the chemists of that period ; it actually existed in the com- pound, and hence,- in any particular substance, only a single radical could be assumed since there was only one present. And thus, with the constantly increasing importance which the radical of a substance attained in respect to views concerning its constitution, divergences necessarily arose in the choice of a radical, according to the decomposition products which were looked upon as the most important. The discussions thus called forth were very helpful in the further development of the science. Everyone tried to support his own view by evidence, and this could only be found in the reactions of the substance. We are indebted to these discussions, therefore, for a very intimate knowledge of certain classes of substances. The foregoing explanations will not, I hope, prove super- fluous. Their purport is to render clear the importance of the radical theory, the further development of which will be discussed in the next lecture. LECTURE VIII. FURTHER DEVELOPMENT OF THE RADICAL THEORY — VIEWS CON- CERNING ALCOHOL AND ITS DERIVATIVES— PHENOMENA OF SUB- STITUTION — DUMAS' RULE — THE NUCLEUS THEORY — THE EQUIVALENT OF NITROGEN. IN the preceding lecture I endeavoured to explain the sig- nificance of the radical, and I shall now deal more fully with its nature. Enough has been said already by way of prepara- tion for the controversies which were called forth by the choice of a definite atomic group as the radical of a compound ; and I now consider it my business to pass in review the most important of the discussions. It was especially with respect to the constitution of alcohol, and of the substances derived from it, that differences of opinion arose. Since it cannot be denied that the opinions regarding these compounds exercised an im- portant influence upon general views, and, further, since the most prominent chemists took part in the discussions, I shall endeavour to show, with respect to this group of substances, how various and how contradictory were the conceptions as to the arrangement of the atoms. I dealt, in the preceding lecture, with the so-called etherin theory, which originated in the comparison of the ethers with the salts of ammonia. In that comparison the radical N H3 was assumed to be present in these salts ; and even although the dualistic tendency of the whole mode of regarding them cannot be disputed, still the view is not in harmony with the other opinions respecting salts. Even then there was another theory as to the compounds of ammonia, in accordance with which they did not occupy any exceptional position, but were looked at from a purely dualistic point of view. The radical ammonium, which, as has already been stated, LECTURE VIII.] HISTORY OF CHEMISTRY. 13! was introduced into the science by Davy,1 constitutes the basis of this hypothesis. Ampere 2 and Berzelius 3 were afterwards chiefly instrumental in procuring its recognition. The superiority of this view, with its purely dualistic basis, as compared with the other, is indisputable ; for the compounds of ammonia (ammonium compounds) now appear as analogues of the ordi- nary salts, as their behaviour necessarily requires that they should. The analogy was illustrated thus : — 4 Sal ammoniac - - (N2H8)CL ! KC12 Muriate of Potash Sulphate of Ammonia (N31I8)OSO3 KOSO3 - Sulphate „ Nitrate ,, (N2H8)ON2O5 ! KON.2O5 Nitrate Acetate „ (NgHaJOQHgO^ KOC4H6O3 - Acetate It is clear that this way of regarding substances could also be applied to the compound ethers, if the radical C4H10 were assumed instead of C4H8. Berzelius took this step in i833.5 He was led to do so not only by his predilection for the ammonium theory, but also on account of newly discovered facts which I shall state here. During the same year, Magnus, by acting with sulphuric anhydride on alcohol and ether, had discovered ethionic and isethionic acids.6 The latter acid was obtained by the decom- position of the former by means of water, and, according to the analyses, it was isomeric with it. Its barium salt, in accord- ance with the prevailing etherin theory, had the formula QH8 + 2SO3 + BaO -f H2O assigned to it ; or, according to Mag- nus, it was to be regarded as a compound of ether and anhydrous sulphuric acid, with baryta. Liebig and Wohler's analyses of barium sulphovinate,7 which were confirmed by Magnus,8 had fixed its composition as C4H8 + 2SO;J + BaO + 2H2O ; 1 See p. 76. 2 Ann. Chim. [2] 2, 16, Note. 3 Gilb. Ann. 46, 131 ; Berzelius, Lehrbuch. Third Edition. 4 Berzelius adopts NH4 as an atom of ammonium ; and yet this does not occur in combination any more than an atom of ammonia NH3. Instead of it the double atom NH4 ( = N2H8) always occurs. The isomorphism of ammonium chloride and potassium chloride must have influenced him to write N-H4.C12 and not (NH4)2C12. 5 Pogg. Ann. 28, 626. 6 Annalen. 6, 152. 7 Ibid, i, 37. b Magnus, loc. cit. 132 HISTORY OF CHEMISTRY. [LECTURE vin. that is to say, it seemed to contain an atom of water more than the newly discovered compound. Berzelius now draws attention to the fact that the latter compound cannot be converted into the barium salt of sulphovinic acid by boiling it with water, and that therefore the view underlying these formulae (in accordance with which the substances contain ready formed water) is erroneous. According to him, alcohol and ether are the oxides of two radicals, C2H6 and €2H5 = C4H10. The compound ethers are thus still regarded as com- posed of ether and of acid ; but there is no longer ready formed water in them, and they now become comparable with salts : — Ether - - - - C4H10O Haloid Ether - - C4H10C1.> Acetic „ C4H100 + C4H60~ Nitric ,, -C4H]0O + N2O5 KO KC1, - KO + C4H603 KO+NoOg Potash Muriate of Potash Acetate ,, Nitrate Berzelius very clearly perceived the importance of his sug- gestion. He had now attained what he had long striven for. The dualistic conception was now applicable to organic com- pounds, or at least to the most fully investigated group of them ; and he does not conceal the pleasure which this occa- sions him. He states- that the organic substances^are now to be regarded (in the same way as mineral substances) as binary groups, but that in them compound radicals alone play the part of the inorganic elements — a view which Dumas and Liebig afterwards develop fully in a special treatise.9 (Compare p. 128.) A ground for discord was now provided, and it was not to be long before the contention should begin. The next incite- ment to it is given by Liebig, who throws down the glove to the etherin theory.10 According to him, this theory has no justification, and all the grounds that can be advanced in its favour rest upon fallacious experiments. Amongst these there is, in the first place, an observation of Hennell,11 according to which sulphuric acid absorbs etherin (olefiant gas), and pro- 9 Comptes Rendus. 5, 567. 240; 1828, 365. 10 Annalen. 9, I. n Phil. Trans. 1826, LECTURE VIII.] HISTORY OF CHEMISTRY. 133 duces sulphovinic acid directly. Liebig endeavours to prove that Hennell's etherin was contaminated with the vapours of alcohol and of ether, and that the pure gas is not absorbed by sulphuric acid.12 He then attacks the formulas of Zeise's platinochloride compounds, which, according to their dis- coverer, consist of etherin, platinous chloride, and potassium chloride.13 Liebig thinks he is justified in concluding from Zeise's analyses, and from the reactions of the substances, that it is not etherin but ether that they contain. Finally, he dis- putes the existence of the ethyl-oxalate of ammonia (ethyl- oxamate) which Dumas and Boullay had prepared from oxalic ether and dry ammonia gas.14 According to Liebig, the same substance is formed by the action of aqueous ammonia, and is identical with oxamide. By this attack all the supports of the etherin theory seemed to be destroyed. Liebig, in bringing the matter forward, admits his adherence to the hypothesis of Berzelius regarding the ethereal compounds, while he only differs from him in his view with respect to alcohol. In the latter substance also he assumes the radical C4H10, which he calls ethyl ; and with him alcohol is the hydrate of ether, C4H10O,H2O. In Liebig's opinion it was no objection that half as many atoms were thus assumed in one volume of alcohol as were assumed in the same volume of ether. Chemists were now further than ever from adopting Avogadro's hypothesis, as may be gathered from the following assertions of Liebig : — 15 " Apart from the contradiction which is involved, if ethei as an oxide is deficient in the property of uniting with water to form a hydrate, while, like other oxides, it is able to unite with acids, and like the metals, its radical is able to unite with the halogens, the specific gravity of alcohol vapour cannot be looked upon as any evidence for its constitution as an oxide of another radical. On the contrary, I believe the very circum- stance that ether and water vapour unite in equal volumes, and 12 Berthelot afterwards showed that absorption can be brought about by vigorous shaking. 13 Mag. f. Pharm. 35, 105; Pogg. Ann. 21, 533. 14 Ann. Chim. [2] 37, 15. 15 Annalen. 9, 16. 134 HISTORY OF CHEMISTRY. [LECTURE VIII. without condensation, is evidence in favour of the view that this compound, alcohol, is a hydrate of ether. ... In the formation of benzoic ether from absolute alcohol and benzoyl chloride, we perceive a simple decomposition of water, which does not extend further than to the water of hydration." Liebig went too far in his argumentation, and Zeise and Dumas justly protested against it. The former repeats his previous examination of the inflammable platinochloride, and finds his first results confirmed : there is no more oxygen present in the compound from which the water of crystallisation has been removed, and, therefore, ether cannot be assumed to be present in it, but only etherin.16 Dumas likewise upholds his earlier experiments.17 He points out the difference in the action of aqueous and of dry gaseous ammonia upon oxalic ether. In the first case only, is oxamide formed, whereas ammonia gas gives rise to the substance he had previously described, which he now calls oxamethan, and to which he assigns the formula C4O3,NH3,C4H4 [C = 6].18 As a conse- quence, Dumas also adheres to his old opinion. He draws attention likewise to the fact that it was with him the idea originated according to which ordinary ether ("sulphuric ether ") is the base of the compound ethers ; and he states that it is, in point of fact, the essence of the whole ethyl theory. He goes a step further, indeed, inasmuch as he regards ether itself as composed of water and olefiant gas. In this Dumas was right. There was one point, however, which had previously been brought forward by Liebig,19 on which Dumas and Liebig differed. This was the assumption by Dumas that in ether two of the hydrogen atoms play a different part from the others : and Liebig contested this particular point. The discovery of the mercaptans by Zeise 20 furnishes Liebig with an opportunity to adduce new proofs of 16 Annalen. 23, I. 17 Ann. Chim. [2] 54, 225. Annalen. 10, 277 ; *5> 52- 1S It seems as if Dumas, in accordance with his earlier experiments, assumed one atom of water more ; the analyses do not, however, afford any sharp decision respecting this. 19 Annalen. 9, 15. 20 Ibid. II, I. LECTURE VIII.] HISTORY OF CHEMISTRY. 135 the accuracy of his ideas.21 He regards these compounds as analogous to alcohol; they are composed of ethyl sulphide C4H]0S, and sulphuretted hydrogen H2S, and their extremely interesting metallic compounds show that in them there really are two hydrogen atoms which behave differently from the others. Two years later, in 1836, he collects together all the grounds for and against each view ; 22 and from the most various facts (but especially from the phenomena of substitution, already discovered by Dumas), he considers himself justified in drawing the conclusion that ether is not a hydrate but an oxide. The matter was not yet settled, however. In his examina- tion of wood spirit (carried out in conjunction with Peligot), Dumas had found new support for his views.23 He succeeded in settling the composition of this substance, a matter that different chemists had attempted to settle, but without success. He showed that in its whole behaviour it had the closest resemblance to alcohol, forming, as the latter does, ethereal compounds with acids ; and he assumes in it the radical C2H,, methylene, with which etherin is polymeric. The advantages of this way of regarding the matter, which does not lead to the assumption of hypothetical radicals, are again pointed out.24 The discussion between Berzelius, Dumas, and Liebig, of which I have given some examples, was of much service to our science. The facts were illuminated from the most different points of view, and this was far more favourable for progress than if a single theoretical opinion had come too prominently to the front. The chemists just mentioned were the chief exponents of chemistry at that time, and round them the other investigators gathered. Of these, only a few represented independent opinions, and chemists were divided, accordingly, into three camps. It is true that in 1837 a sort of armistice was concluded. At a personal interview, Liebig converted 21 Annalen. II, 10. 22 Ibid. 19, 270, Note. '23 Ann. Chim. [2] 58, 5 ; Annalen. 13, 78 ; 15, I. 24 Dumas and Peligot believe that they can isolate methylene. 136 HISTORY OF CHEMISTRY. [LECTURE VIII. Dumas to his opinions, and we find the two savants conjointly publishing a scientific treatise 25 in which they announce that thenceforth they wish to elaborate organic chemistry with united energies and from the same point of view; to have analysed in their laboratories all the substances not yet examined by them ; to open up, by the aid of their pupils, additional lines of investigation in the most varied directions ; and to subject the work of others to a rigid criticism and check. But the union only endures for a short time : it is terminated in a year, and each returns to special paths, which diverge more and more. In 1840 the two are again hostile towards each other, although perhaps more careful in their assertions and more polite than previously. Dumas had made observations in the meantime which caused him to break away from all traditions, to abandon dualism and the electro chemical theory, and to express views that Berzelius especially attacked most vigorously. The latter, who up to this time had taken the most important part in the development of the science, holding his opponents in check by means of his theories, and who had struggled for the ascendency with Liebig and Dumas, now tries, ineffectually, to oppose his ideas to those of Dumas and of Laurent. He relies upon unestablished hypotheses, which only later, at Kolbe's hands, receive a real foundation, and acquire thereby a scientific significance. Before turning to this period, with its theories of substitu- tion, of nuclei and of copulae, we must still consider a further development of radicals in the earlier sense, with which the various notions respecting alcohol and the compounds derived from it are brought to a close. Regnault, in his examination of the oil of the Dutch chemists, had found that this substance loses the elements of hydrochloric acid when it is distilled with potassium hydroxide, and that it yields a new substance of the composition C4H6C12.26 This he regards as the chloride of the radical aldehydene, C4H6, and 25 Comptes Rendus. 5, 567. <26 Ann. Chim. [2] 59, 358 ; Annalen. 15, 60. LECTURE VIII.] HISTORY OF CHEMISTRY. 137 confirms his view by preparing the bromine and iodine com- pounds of this radical, which he also assumes to be present in aldehyde and in acetic acid. He writes : — C4H6 - C4H6,C12 C4H6,Br2 - C4H6,C12 •+• H2Cl2 ~ C4H6,Br2 + H2Br2- (Ethylene Hydrocarbon (Ethylene Hypothetical radical Aldehydene Chloraldehydene Bromaldehydene Chloride of Hydrocarbon chloride) Bromide of bromide) C4H6,O +H2O - Aldehyde C4H(JO3 +H2O - Acetic acid This investigation of Regnault's, carried out in 1835, was prompted by Liebig, and was intended to prove to Dumas that the radical etherin is not present even in ethylene chloride ; it was intended to overthrow the etherin theory, and it may have had much influence in moving Dumas to give up his former views. It is true that in consequence of this investigation Liebig also abandoned the radical ethyl, and tried to explain the ethereal compounds on the assumption of a radical acetyl, C4H6.2r These substances are again compared with the salts of ammonia, but in the latter the radical amide is now assumed. Acetyl - Ac = C4H6 Ad = N2H4 - Amide Olefiant gas - AcH2 AdH2 - Ammonia Ethyl - - AcH4 AdH4 - Ammonium Ether - AcH40 AdH4O - Ammonium oxide Ethyl chloride - AcH4Cl2 AdH4Cl2 Sal ammoniac Alcohol AcH4O + H2<5 AdH4O + H2O Compound in sul- phate of ammonia Mercaptan - AcH4S + H,S AdH4S + H2S Hydrosulphuret of ammonia Isethionic acid - AcH2 + 2SO3 AdH2 + SO3 - Rose's anhydrous sulphate of am- monia. Acetic acid - - AcO + O2 Aldehyde - - AcO + H./5 27 Annalen. 30, 139. 138 HISTORY OF CHEMISTRY. [LECTURE VIII. Three views respecting the ammonium salts and the com- pound ethers had thus been advanced, and these differed from one another merely in the number of hydrogen atoms which were assumed in the radical. These were : — 1. The ammonia theory of Lavoisier, corresponding to the etherin theory of Dumas and Boullay ; 2. The ammonium theory of Davy, Ampere, and Berzelius, corresponding to the ethyl theory of Berzelius and Liebig; and 3. The amide theory of Davy and Liebig, corresponding to the acetyl theory of Regnault and Liebig. Liebig believed that, by his new theory, he had overcome all difficulties, and had settled all contentions with respect either to the ethyl theory or the etherin theory. He closes his paper with the following words : " From this point of view, both of the formerly opposed theories possess, as may easily be observed, the same kind of basis, and all further question as to the truth of the one view or of the other is thereby settled." In a certain respect, Liebig was right ; the.question whether ethyl or etherin was present in alcohol was no longer discussed ; not, perhaps, because ace.tyl was preferred, but because chemists now began to attach another signification to the radicals. The phenomena of substitution, which were already known, gradu- ally became of general applicability, and, after the discovery of trichloracetic acid, the hypotheses which Dumas and Laurent had advanced, acquired great influence. Not only was the radical theory, in the form in which it was then stated, threat- ened by these hypotheses, but dualism and the electro-chemical theory — the very foundations of the entire mode of viewing chemical matters — were attacked by them, and were eventually driven out of the science. These hypotheses led to the recog- nition of the radicals as variable and of chemical compounds as possessed of an individual or unitary character; and to the abandonment, as arbitrary, of the division of the latter into two parts. At a later period, in conjunction with the conceptions derived from the theory of the polybasic acids, they led to a revision of the size of the atoms in the case of compounds, to LECTURE VIII.] HISTORY OF CHEMISTRY. 139 the establishing of the chemical molecule, and to the theory of types. Simultaneously, the conception of the equivalent assumes a more fixed form, and is distinguished from that of the atom ; it is recognised that the atoms are not equivalent, but are of different values in combination ; the theory of atomicity is developed and this stimulates the determination of rational constitution as we now understand it. Let us next make something more than a mere bird's eye in- spection of this period, rich as it is in discoveries and hypotheses. We shall now subject it to a minute examination. In doing so, we find that the development of chemistry during the last sixty years is not inferior in interesting and important episodes to that of any period in the science. The participation in this development is a constantly increasing one, and it is a difficult task to seek out from the enormous mass of material which was elaborated during this period, the things that were important and conducive to progress, to state the development of the ideas in such a way that they shall be at once logical and in accord- ance with the actual facts, to do justice to every one, and yet not to lose the thread over details or questions of priority. The history of this epoch has never yet been described in a connected manner.28 In venturing upon the attempt to do this, I am well aware that an objective representation of the period is scarcely possible, and that I play the part rather of critic than of historian. Still I have endeavoured to make my exposition of some value from the fact that I have always been careful to arrive as nearly as I could at the truth, and not to permit myself to be led by prejudices or personal matters. The conception of equivalence might have led to that of replacement or substitution, since the quantities of two acids were equivalent when they saturated the same quantity of a base. The acid in a neutral salt could thus be replaced by its equivalent, without the neutrality being interfered with. The word " replacement " received further justification after Mitscher- 28 Wurtz's Histoire des doctrines chimiques only appeared during the printing of the first edition of these lectures, and Kopp's Entwickelung der Chemie in der neueren Zeit appeared some years afterwards. / 140 HISTORY OF CHEMISTRY. [LECTURE VIII. lich had studied the phenomena of isomorphism. It could then be said that certain elements in a crystal might be replaced by others, without alteration of the crystalline form. Such substi- tutions possessed the peculiarity, however, that they were not connected with any proportions by weight, and it may thus ap- pear all the more remarkable that they should render important assistance in the determination of atomic weights. The hypo- thesis underlying the phenomena of isomorphism was that one atom could only be replaced by one other ; that is to say, that the numbers of the atoms in isomorphous compounds must be identical. Since chemically similar substances had alone been compared, an extension of the prevailing views, based on the phenomena of isomorphism, would have been quite possible ; but this class of phenomena had never led to any attack upon the system. Such an attack now took place, however, and it was founded upon a series of facts which I must relate here. In the bleach- ing of wax by means of chlorine, Gay-Lussac had observed that for every volume of hydrogen eliminated, an equal volume of chlorine was taken up.29 He had also found the same thing in the action of chlorine- on hydrocyanic acid. In the course of their investigation of the benzoyl compounds, previously referred to, Wohler and Liebig, when acting with chlorine upon bitter almond oil, had discovered benzoyl chloride ; and they expressly remark that this substance is produced from the bitter almond oil by two atoms of chlorine taking the place of two of hydrogen.30 In 1834, Dumas examines the action of chlorine on oil of turpentine,31 and in this case also each volume of hydrogen eliminated is replaced by the same volume of chlorine. Then when he studies the products of the de- composition of alcohol by means of chlorine and of bleaching powder, in order to clear up the nature, and the mode of formation of chloral and of chloroform, he states the empirical rule, observed in a single case by Gay-Lussac, in the following general form : — 32 29 Gay-Lussac, Legons de Chimie.30 Annalen. 3, 263. 31 Ann. Chim. [2156,140. 32 Ibid. [2] 56, 113. LECTURE VIII.] HISTORY OF CHEMISTRY. 14! 1. When a substance containing hydrogen is exposed to the dehydrogenising action of chlorine, bromine, or iodine, for every volume of hydrogen that it loses, it takes up an equal volume of chlorine, bromine, etc. 2. When the substance contains water, it loses the hydrogen corresponding to this water without replacement. The second rule was advanced chiefly to explain the formation of chloral, and at the same time to justify the formula C8H8 + 2H2O [C = 6] adopted for alcohol by Dumas six years previously. According to Dumas, the phenomena of substitution furnish a new proof of the difference of the hydro- gen atoms, eight of which are united to carbon, and four to oxygen. In the case of the former alone does replacement occur, whilst the others are removed without replacement. Thus we have : — (C8H8 + 2H20) + 4C1 - C8H802 + 4HC1 Aldehyde. C8H8O2 + 1 2C1 = C8H2C16O2 + 6HC1 Chloral. By means of various examples, Dumas further endeavours to prove the general validity of the laws which he advances. In establishing the correct composition of the Dutch oil, he points out that the chloride of carbon obtained from it by means of chlorine, and examined by Faraday,33 supplies a new argument in favour of the accuracy of his views. He also finds similar support in the action of chlorine on hydrocyanic acid, on bitter almond oil, etc. But Dumas is not satisfied with this. He goes a step further still, and regards oxidations as cases of substitution, as, for example, the conversion of alcohol into acetic acid.34 In this case every volume of hydrogen eliminated is replaced by half a volume of oxygen. Accordingly we have : — (C8H8 + H402) + 04 = (C8H402 + H402) + H4O2 ; Alcohol. Acetic Acid. 33 Phil. Trans. 1821, 47. 34 Ann. Chim. [2] 56, 143. 142 HISTORY OF CHEMISTRY. [LECTURE VIII. and the formation of benzoic acid from bitter almond oil is explained in the same way : — C»H10O2.H, + O2 = C28H18O2.O + H2O. Bitter Almond Oil. Benzoic Acid. In order to be able to include the action of oxygen within his rule, he states the latter as follows : — When a compound is exposed to the dehydrogenising action of any substance, it takes up a quantity of this substance equivalent to the quantity of the hydrogen eliminated. The doctrine of Dumas seems to me to be of most im- portance in this connection. He shows us that equal volumes of hydrogen, chlorine, bromine, and iodine are equivalent, while they possess only half the value of the same volume of oxygen. The difference between the two is clearly visible here, and this is the beginning of the separation of atom and equivalent. The phenomena of substitution, or of metalepsy, as Dumas called it, were followed up further in the succeeding years by Dumas himself, as well as by Peligot,35 Regnault,3*3 Malaguti,37 and, especially, Laurent ;-and in particular, it is the independent extensions which the latter gave to Dumas' rule, that we shall now consider. Laurent has enriched chemistry with a very large number of experimental investigations, but these, in many cases, are unfortunately wanting in the necessary accuracy. He had at his command only very limited resources, and instead of confining himself, on that account, to a few branches, Laurent, who was very fertile in ideas, preferred to start a great many things and to carry them out in a superficial manner. He destroyed, in this way, his reputation as an experimentalist, and was met with hostility at the very outset of his scientific activity ; while he was afterwards treated with unnecessary severity, especially by Berzelius and Liebig. This naturally 35 Annalen. 12, 24 ; 13, 76 ; 14, 50 ; 28, 246. 3S Ibid. 17, 157 ; 28, 84 ; 33, 310 ; 34, 24, etc. 37 Ibid. 24, 40 ; 25, 272 ; 32. 15 ; 56, 268, etc. LECTURE VIII.] HISTORY OF CHEMISTRY. 143 reacted upon him, and he went his own way, and became gradually more incomprehensible, particularly in consequence of a nomenclature which was used almost exclusively by himself. Many of his clever and original ideas were thus lost to our science, or are now ascribed to the services of others. A great deal, on the other hand, was supplied to us first by Gerhardt, who was a friend and collaborator of Laurent for many years, and who combined a clear mode of expressing and of regarding chemical matters with probably more perspicuity and less genius. At an early period, Laurent had begun to occupy his attention with the phenomena of substitution. He first studied naphthaline and its derivatives ; 3S then, simultaneously with Regnault, the derivatives of ethylene chloride ; 39 afterwards the action of chlorine upon compound ethers,40 and upon the products of the distillation of tar, especially on phenol,41 etc. He very soon satisfied himself, by means of these various investigations, that the form which Dumas had given to the law of substitution was not a generally accurate one. He found that in very many instances there are more or fewer equivalents of chlorine or of oxygen taken up than there are of hydrogen eliminated, and vice versa ; and that this is the case even with substances which do not contain oxygen, so that the exceptions cannot be explained by Dumas' second rule.42 At the same time, however, Laurent points out that the substituted product, when it is produced by replacement of equivalent by equivalent, still exhibits certain analogies with the original substance ; arid he asserts that the chlorine introduced takes the place, and to a certain extent plays the part, of the hydrogen eliminated. His opinion may be stated somewhat in the following manner : — 43 Many organic substances when treated with chlorine lose a certain number of hydrogen atoms, which escape as hydrochloric 38 Annalen. 8, 8 ; 19, 38 ; 35, 292 ; 41, 98 ; 72, 297 ; 76, 298, etc. 39 Ibid. 12, 187 ; 18, 165 ; 22, 292. 40 Ibid. 22, "292. 41 Ibid. 22, 292 ; 23, 60; 43, 200, etc. 4'J Sec p. 141. 43 Laurent, Methode de Chimie. 242 ; E. 199 ; These de docteur, Paris, 20 Dec. 1837, n, 88 and 102 ; Ann. Chim. [2] 63, 384; Comptes Rendus. 10, 413; Revue Scientifique. I, 161. 144 HISTORY OF CHEMISTRY. [LECTURE VIII. acid; chlorine atoms equal in number to the hydrogen atoms eliminated are substituted for the latter, so that the physical and chemical properties of the original substance are not essentially altered. The chlorine atoms thus occupy the space left vacant by the hydrogen atoms. In the new compound the chlorine to a certain extent plays the same part as the hydrogen did in the original substance.44 Laurent endeavours to give expression to the observed facts and to the hypotheses based upon them, in the so-called nucleus theory.45 This theory is of importance in our science (although it never obtained any general recognition in it) because we have adopted, even if in another form, many of the ideas embraced by it, and also because it was adopted by Gmelin as the basis of the organic portion of his excellent handbook. On this account I shall state the chief points of Laurent's doctrines. According to Laurent, all organic substances contain certain nuclei, which he calls either fundamental (radicaux fonda- mentaux) or derived. The former are compounds of carbon with hydrogen, in which the mutual proportion of the number of the atoms is a simple one (i to 2, 3, 4, etc., 2 to 3, etc.). For any definite proportion, several nuclei exist which are polymeric amongst themselves. Besides, these fundamental radicals are so chosen that the hydrogen and carbon atoms contained in them occur in pairs. Subsidiary nuclei are formed from the fundamental nuclei by the substitution of other elements for the hydrogen ; for example, chlorine, bromine, iodine, oxygen, nitrogen, etc. Laurent afterwards assumes replacement by radicals or groups of atoms. In such reactions Dumas' rule always holds, that the hydrogen turned out is replaced by equivalent quantities of 44 I wish to recall the fact here, simply because a question of priority arose with respect to it (compare Comptes Rendus. 10, 409 and 511), that Liebig and Wohler, in their investigation of bitter almond oil, had already assumed that during the formation of benzoyl chloride the chlorine takes the place previously occupied by the hydrogen (compare p. 140). 45 Ann. Chim. [2] 6l, 125 ; compare also Gmelin, Handbuch. 4, 16 ; E. 7, 18. LECTURE VIII.] HISTORY OF CHEMISTRY. 145 other elements. But this is not the only kind of change that the nucleus can undergo, and here Laurent differs from Dumas. Thus an indefinite number of atoms may attach themselves to the radical, and be removed from it again without being replaced ; whereas, as already stated, an atom cannot be removed from the nucleus without the entry of its equivalent, since the destruction of the whole group would otherwise ensue. Such destruction infallibly occurs as soon as carbon, in the form of carbonic anhydride, carbonic oxide, etc., is removed from the compound ; in which case either a complete decomposition takes place, or a fresh nucleus is formed. The relation of this nucleus to the first one is not further defined, however. According to Laurent, the subsidiary nuclei show a great resemblance to the fundamental radicals in their physical and chemical characters ; the derived nuclei, obtained by attachment of new atoms, have, on the other hand, acquired a different character. Thus a union with hydrogen and oxygen (water) usually brings about the formation of alcohol ; a neutral oxide is formed by the taking up of two atoms of oxygen, a monobasic acid by the taking up of four atoms, and a dibasic acid by the taking up of six atoms of oxygen. Laurent further adopts a geometrical conception respecting organic compounds. In accordance with this conception the nuclei are prisms in whose angles the carbon atoms are located, whilst the edges are formed by the hydrogen atoms. These edges can be taken away and replaced by others, without the figure undergoing any considerable changes. But should the place be left unoccupied, the internal connection would come to an end and the whole would fall to pieces. Atoms can still be added to the prisms so as to form pyramids ; and the whole figure may be surrounded in this way, by which means its form is, of course, altered. These pyramids can be removed, where- upon the original prism makes its appearance again. In our matter-of-fact science we are not accustomed to such imaginative views, and so it may appear as if nothing of any value for chemistry lies hidden behind them. In order to disprove this, however, I shall translate the hypotheses of K 146 HISTORY OF CHEMISTRY. [LECTURE VIII. Laurent into our ordinary language. It will then be possible to obtain, moreover, a better grasp of Laurent's ideas. The nucleus theory clearly sprang from the radical theory, but only by an essential reconstruction of the latter. The radical of Laurent is not an unalterable group of atoms, but it is a compound, which can be altered by substitution in equiva- lent proportions and does not lose thereby its characteristic properties. Thus Laurent is able to derive all his radicals from hydrocarbons, a proceeding which is, of course, in complete con- tradiction to the older ideas. These radicals can unite with other atoms, and in the substances so produced the nuclei are present as such ; the nuclei pre-exist in the substances, and Laurent, therefore, entirely agrees on this point, with his predecessors. By means of these two hypotheses, he is able to explain all the facts — not only the cases which follow Dumas' rule, but those also which are at variance with it, and of the latter he had found a large number. At the same time, his point of view furnishes reasons why both kinds of reactions are possible. On the assump- tion of the alterability of the radical, it may easily be understood that a group included far more compounds than was possible with the older radical theory. Laurent was thus able to discover far more of what we should now call " generic relationships," and that was an unquestionable advantage. Since he assumed the number of carbon atoms in the nucleus to be constant, sub- stances were arranged into series according to the number of atoms of carbon they contained, and this supplied the basis of an excellent systematic classification. With him there was no connecting link between the series so formed ; and in this respect Laurent's division differs from the classifications of the present day, which accentuate all relationships as much as possible. No such mode of treatment could, indeed, have been carried out at that time. After these explanations, I may state that much that was new and good was advanced in Laurent's nucleus theory. Its importance lies principally in the fact that it was capable of a general application, and, as Gmelin proved, could be admirably employed as the basis of a detailed text-book. In this respect it LECTURE VIII.] HISTORY OF CHEMISTRY. 147 is distinguished by very marked advantages over the radical theory, which, owing to the very definite form that had been given to the radical, could only be of service in certain direc- tions, while it wholly overlooked various relationships. I think it may be advantageous to show, by means of a few examples, the mode in which Laurent applied his theories, and also what his formulae were for different compounds. In doing this I shall choose familiar groups of substances. Nucleus, Etherene C4H8 [C= i2]46 Etherene Hydrochlorate C4H8 + H2C12 Chloretherase C4H6C12 Chloretherase Hydrochlorate C4H6C12 + H2C12 Chloretherese - C^C^ (then unknown) Chloretherese Hydrochlorate C4H,C14 + H2C12 Chloretherise C4H2C16 (then unknown) Chloretherise Hydrochlorate C4H2C16 + H2C12 4T Chloretherose C4C18 Etherose Chloride C4C18 + C14 Chloral C4C16O + H2O Bromal C4Br6O + H2O Chloracetic Acid (then unknown) C4H2C14O + O2 Nucleus, Methylene C2H4 Chloroform - C2C14 + H2C12 Bromoform C2Br4 -f H2Br2 Cyanogen C2Az2 Hydrocyanic Acid - C2 Az2 + H2 Cyanic Acid C2Az2 + O 4S Nucleus C14H14 Bitter Almond Oil - C14H10O2 + H2 Benzoic Acid - C14H10O2 + O Hydrobenzamide C14HIOAz4/3 4- H2 49 46 Ann. Chim. [2] 63, 388 ; Annalen, 22, 303. 47 The principle of the nomenclature employed originated with Dumas (Ann. Chim. [2] 57, 305). Laurent almost always used this nomenclature. 48 The symbol Az (Azote) was written for the nitrogen atom in France at that time (as it still frequently is). I have used it here for a reason that will be perceived further on, 49 Ann. Chim. [2] 62, 23. 148 HISTORY OF CHEMISTRY. [LECTURE VIII. I have intentionally chosen these particular examples. They lead us to a point in Laurent's views which we have, up to the present, only very superficially touched upon. Since he assumed substitutions of hydrogen by nitrogen, we may inquire what the equivalent of nitrogen was. We recognise, from his deriving cyanogen from methylene, that Laurent assumed one atom ( = 14 parts by weight) of nitrogen as equivalent to two atoms or parts by weight of hydrogen. This hypothesis did not accord with fact in the case of hydrobenzamide, which Laurent had obtained by treating bitter almond oil with ammonia. If the new substance was to be referred to the same nucleus as that from which he derived benzaldehyde, then two-thirds of an atom, or 9.33 parts by weight of nitrogen, must be equivalent to two parts by weight of hydrogen. Laurent did not know how to get out of this dilemma. The question was decided by Bineau.50 In a detailed paper the latter endeavoured, in 1838, to give a solu- tion to the problem of the determination of the equivalent of nitrogen. After a discussion of the fact that the common method of fixing this number is very arbitrary (considering that the quantity of a substance which, in its lowest stage of oxida- tion, unites with ioa parts by weight of oxygen is usually assumed to be equivalent to that quantity of oxygen, whereas it would be quite as justifiable to start from any other stage of oxi- dation), he adopts other considerations in respect to the matter. He obtains his evidence chiefly from the hydrogen compounds. He compares ammonia with water, and asks how many atoms of oxygen are necessary in order to oxidise completely the hydrogen united to one atom of nitrogen. It is known that i J atoms are required for this, consequently Bineau finds 14 parts by weight of nitrogen to be equivalent to 24 of oxygen and to 3 of hydrogen ; in other words, the equivalent of the former, compared with 1 6 parts by weight of oxygen, is 9.33 = Az§. He introduces for nitrogen the symbol N, and points out that hydro- benzamide now accords with Dumas' rule. It can very easily be understood that Laurent accepted Bineau's determination. 50 Ann. Chim. [2] 67, 225, LECTURE VIII.] HISTORY OF CHEMISTRY. 149 On scarcely any side did the nucleus theory meet with acceptance. Dumas made use of much of it in advancing the theory of types, and although in doing so he mentions Laurent, still opinions which unquestionably were first stated by the latter, were ascribed to Dumas. No doubt, they met with acceptance earlier on account of the superior authority and position of Dumas. Liebig, however, expressed himself very forcibly against Laurent,51 and he was not altogether unjustified in his charges. In the application of his theory Laurent incurred blame for many arbitrary proceedings, which Liebig knew how to point out with much effect. Liebig further attacks the facts dis- covered by Laurent and employed by him in support of his opinions ; and these are likewise not always able to withstand the keen criticism applied to them. Much more violent still were the attacks of Berzelius,52 which he erroneously directed against Dumas. The view that negative chlorine could take the place of positive hydrogen, without altering the nature of the product, was wholly inadmis- sible with the author of the electro-chemical theory. He makes every conceivable endeavour to bring the constantly increasing number of substitution products into harmony with his theories. I shall postpone, however, the detailed consideration of his views until a subsequent lecture, and shall close this one with an observation of Gerhardt's,53 which enables us to recognise the clear and intelligent perception of this chemist, who at the time was still quite a young man. Laurent's formula for Dutch oil was C4H6C12 + H2C12. By treatment with chlorine this substance was said to be converted into chloride of carbon, C4C112.54 According to Gerhardt, Laurent's formula is inaccurate because it assumes the decom- position of hydrochloric acid by means of chlorine, accompanied by the re-formation of hydrochloric acid. 51 Annalen. 25, I. 5'2 Compare Comptes Rendus. 6, 629 ; and Berzelius' Jahresbericht 1840, 361. 53 J. pr. Chem. 15, 17. 54 Phil. Trans. 1821, 47. LECTURE IX. GRAHAM'S INVESTIGATION OF PHOSPHORIC ACID— LIEBIG'S THEORY OF POLYBASIC ACIDS, AND HIS VlEWS WITH RESPECT TO ACIDS IN GENERAL — ADOPTION OF THE DAVY-DULONG HYPOTHESIS — DIS- COVERY OF TRICHLORACETIC ACID —ATTACK UPON THE ELECTRO- CHEMICAL THEORY— REPLIES OF BERZELIUS— COPUL^E. I WISH to begin this lecture with some general remarks which may serve as an appendix to what was stated in the preced- ing lecture respecting the phenomena of substitution, and especially to the conceptions of them which were entertained by Dumas1 and by Laurent.2 I desire to point out how, in consequence of the observed phenomena of substitution, the conception of the equivalent assumed a more definite form. Thus in assuming, with Dumas, that the quantities replacing one another are equivalent (and this was an assumption that had some justification, according to Laurent's views which rendered possible a direct com- parison between the original and the final products), a series of experiments was all that was necessary in order to determine the equivalents of the substances so replacing one another. One section of chemists actually did work in this direction, and, in consequence of this, it is necessary to observe par- ticularly to which school the author of any paper published at that time belongs ; for at this very period Gmelin's school, which likewise wrote or desired to write in equivalent formulae, began to acquire much influence. Whilst the adherents of the substitution theory (who made use of the equivalents), in spite of numerous shortcomings and mistakes, always endeavoured to separate from each other the conceptions of atom and 1 Dumas, Traite, Organic Part. I, 75. 2 Annalen. 12, 187. LECTURE IX.] HISTORY OF CHEMISTRY. 151 equivalent, and to follow up both of them consistently, almost the opposite might be said of their opponents. The latter (/ knew that one atom of alumina, A12O3, requires three times as much sulphuric acid to saturate it as one atom of potash, KO ; they also considered that one atom of phosphoric acid, P2O£, required three times (in reality twice) as much base to form a neutral salt as one atom of hydrochloric acid ; and yet they did not hesitate to employ the name " equivalents " for these quantities. Just because our chemistry of to-day is based essentially upon the difference between the conceptions of atom and equivalent, everything that could lead to a distinction between these conceptions must be specially emphasised. I wished, therefore, to point out that a new means of determining equiva- lents was furnished, in the fourth decade of the century, by the phenomena of substitution, and that this involved a real advance in the question which is especially interesting us at present. But it was further shown, from an entirely different point of view, that the atoms of compound substances are not necessarily equivalent. Decisive reasons were advanced to establish the differences that exist, in this respect, in the case of the acids, which formed one of the most fully investigated classes of substances. The experiments relating to this matter were carried out at an earlier period than the advancement by Dumas of the theory of types, which was the next step in the development of the substitution theory, and on this account I think it should be treated of first. Even if both matters seemed at that time to be extremely diverse, still the exercise of some influence is not only conceivable, but it can actually be observed ; and, for this reason, the chronological order must not be altogether lost sight of. It is seldom that, in order to produce any .great result, so few investigations have been necessary as was the case in the founding of the theory of polybasic acids. The experiment and the idea whereby this wide field of investigation was opened up to experimental science, while new and secure footing was afforded to theory, are alike elegant and precise. Only a few I52 HISTORY OF CHEMISTRY. [LECTURE IX. persons took a part in this important crisis in chemistry, but they were valiant champions who conquered for us this domain. And, once set foot upon, the ground was secure, despite the contradiction of an authority whose words, although they had been observed in other cases in a scrupulously conscientious manner, were now spoken to the winds. It is to Graham that the first impulse towards an alteration in the views regarding acids was due. Graham's investigation of phosphoric acid, and the way in which he states his results — the former free from preconceived notions and hypotheses, and the latter clear and definite — show us that we have to do with an acute and clear-minded thinker. When regard is paid to the ideas that were necessarily introduced as the direct result of the investigation, and when those intellectual advances are considered which were occasioned — not exclusively, it is true, but still to a great extent — by Graham's labours, it must be conceded that so much has seldom been accomplished by a single investigation. As a consequence of Clark's investigation of phosphoric acid,3 the view had been arrived at that this acid existed in two isomeric conditions, which, in their salts in particular, were stated to present great differences.4 Common sodium phos- phate gave a yellow precipitate with neutral silver salts, and the liquid possessed an acid reaction ; the pyrophosphate, on the other hand, precipitated white silver pyrophosphate and the neutrality remained. It was known, of course, that the one sodium salt crystallised with more water than the other, but this was regarded as water of crystallisation, and no im- portance was attached to it ; so that the two acids were looked upon as isomeric modifications.5 Graham rectified this mistake. He succeeded in throwing light upon this hitherto obscure subject by proving that the water which was contained in the hydrated acids should not be disregarded as inessential to their 3 Edinburgh Journal of Science. 7, 298 ; Schweigger's Journal. 57, 421. 4 See also Stromeyer, Schweigger's Journal. 58, 123. 5 Compare Ber- zelius, Lehrbuch. Third Edition, 2, 60. LECTURE IX. J HISTORY OF CHEMISTRY. 153 constitution, but that, on the contrary, it assumed the function of the base.6 Graham showed, in 1833, how ordinary phos- phoric acid and all its salts may be regarded as compounds of one atom of phosphoric acid, P2O5, and three atoms of base which can be completely or partially replaced by water. Thus, according to him, ordinary (neutral) sodium phosphate consists of one atom of phosphoric acid combined with two atoms of soda and one of water : on mixing its solution with silver nitrate, the silver salt with three atoms of silver is precipitated, whilst sodium nitrate and nitric acid remain in the solution together. Exceptions to Richter's law (already observed in similar cases by Berthollet), where the mixture becomes acid when the . solutions of two neutral salts are mixed,7 were thus explained. In this case two atoms of soda and one of water were exchanged for three atoms of silver oxide. Another very important result of Graham's investigation is met with in the analysis of pyrophosphoric acid and its com- pounds. Graham shows that on heating the sodium salt men- tioned above to over 350°, the water it contains is driven off, and that sodium pyrophosphate, identical with the salt already known, is in this way produced. This salt is not, however, isomeric with the original one, as had been supposed, but differs from it by containing one atom of water less ; and this is of essential importance in regard to the nature of the acid. The white precipitate, too, which is produced by silver salts, only contains two atoms of silver oxide ; and it is thus a quite general property of pyrophosphoric acid, to saturate only two atoms of base (or of water), which distinguishes it very sharply from ordinary phosphoric acid. In the latter the ratio of the oxygen in the base to that in the acid is as 3 to 5 ; in the other acid it is as 2 to 5. Graham finds further that on heating the acid sodium phosphate, which consists, according to him, of one atom of phosphoric acid, one atom of soda, and two atoms of water 6 Phil. Trans. 1833, 253 ; A.C.R. No. 10 ; Annalen. 12, I. 7 Ber- thollet, SUit. Chim. I, 117 ; E. I, 85. 154 HISTORY OF CHEMISTRY. [LECTOR?: IX. (as base), both water atoms are driven off, and a hitherto unknown salt, sodium metaphosphate, is produced. The acid contained in this salt is characterised by being saturated by one atom of base, whilst, in the free state, it contains one atom of water. The silver compound was again different from either of the others. In this case the ratio of the quantities of oxygen in base and acid was as i to 5. Finally, it was shown in the investigation, that meta- and pyrophosphoric acids, as well as the majority of their salts, pass into ordinary phosphoric acid or a salt derived from it, when boiled with water, or still better, when fused with sodium carbonate. Two important theoretical conclusions can be directly de- duced from Graham's investigation. (i.) In acids there is a certain number of atoms of water, and salts are formed by the replacement of these. (2.) The atoms of the acids are not always equal in number to the atoms of the bases, and in some, the ratio is even variable. Thus Graham showed how, from the same phos- phoric anhydride, to prepare three hydrates which were able to take up quite different quantities of base. Liebig, in 1838, stated these conclusions with great clear- ness and precision.8 A man of his genius could not, however, rest satisfied with publishing thoughts that were merely conclu- sions drawn from the experiments of others. We are indebted to Liebig for an excellent investigation of a series of organic acids, from which it appeared that phosphoric acid does not stand alone with respect to its behaviour towards bases, but that in the cases of certain other acids, one atom likewise pos- sesses the property of saturating several atoms of base. Found- ing, as he did, upon a broader basis, he was then able to introduce the idea of the polybasic acids. Liebig's experimental investigation embraces fulminic, cyanic, meconic, comenic, tartaric, malic citric, and other acids. He finds relations amongst the salts of each of these acids, 8 Annalen. 26, 113 ; compare Comptes Rendus. 5, 863. LECTURE IX.] HISTORY OF CHEMISTRY. 155 which are similar to those in the case of phosphoric acid. But he endeavours, especially, to range the three cyanogen acids, i.e. cyanic, fulminic,9 and cyanuric, side by side with the three phosphoric acids. In the former, as well as in the latter, there is present, according to him, a group of atoms which has the power of saturating sometimes one, sometimes two, sometimes three atoms of base. But whilst the atomic weight is not altered in the case of the phosphoric acids, it increases in that of the cyanogen acids in the same ratio as the saturating capacity, so that the resulting salts are polymeric with one another. In the latter case, the quotient obtained by dividing the quantity of oxygen in the acid by that in the base, remains unchanged, whereas this, according to Graham, is not the case with the varieties of phosphoric acid. Liebig writes : 3MO.P2O5 Phosphate. 3MO Cy6O3 Cyanurate. 2MO.P2O5 Pyrophosphate. 2 MO Cy4O2 Fulminate. MO.P2O5 Metaphosphate. MO Cy2O Cyanate. Of distinctly greater importance are the considerations which lead Liebig to propose a separation from the other acids, of those which behave in a way analogous to phosphoric acid. The course of his argument in this matter is approxi- mately as follows : The relations are not so complicated in the cases of all the acids which share with phosphoric acid the characteristic property of neutralising several atoms of base by One atom of acid, as they are in the case of phosphoric acid itself; and hence it is not so easy to establish in all cases that they belong to this category. In the case of phosphoric acid, no matter what number may be chosen as its atomic weight, it can never be shown that one atom of acid saturates one atom of base in all three modifications.10 What, now, are the characteristics that enable us to recognise that we have to do with a substance belonging to this group ? 9 Liebig assigns to fulminic acid the formula 2rLO.Cy4O2 [H = i, C=I2, 0=16]. 10 Here, as in the foregoing, the word acid is to be understood as referring to the anhydride. 156 HISTORY OF CHEMISTRY. [LECTURE IX. Liebig has recourse to experiment in order to decide this highly important question. He compares the behaviour of phosphoric acid with that of sulphuric acid, a compound con- cerning which he has no reason for reckoning it in this class. In doing so he says : n - " If to acid sulphate of potash, we add another base which is not isomeric with potash and which forms with sulphuric acid a salt free from water of halhydration,12 soda for example, the acid salt separates into two neutral ones, Glauber's salt and sulphate of potash, which crystallise apart from each other. "If, on the other hand, a certain quantity of potash is added to acid phosphate of soda, phosphate of soda and potash is formed, wholly analogous in its composition to the acid salt. It contains three atoms of base ; two of these are soda and potash ; one of the two atoms of water previously contained in it, is replaced by potash, the second atom remains in the composition of the new salt. "This behaviour distinguishes phosphoric acid and arsenic acid from the great majority of all other acids : their power of forming salts of the same class with different bases, differing from those which are called double salts, depends essentially upon their property of combining with several atoms of base. I regard this character as decisive respecting the constitution of these, and of all acids which form compounds similar to those of phosphoric acid" A criterion is thus found for separating phosphoric acid and its analogues from the other acids, and Liebig employs it in order to establish the fact that all the substances examined by him belong to this class. The grounds upon which he also decides to include tartaric acid in this group are very interest- ing and important. This acid was at that time written C4H4O6, so that its atom saturated only one atom of base. The exist- 11 Annalen. 26, 144-145. 12 Liebig regards as water of halhydration, that water in salts which can be separated and replaced by equivalents of neutral salts. LECTURE IX.] HISTORY OF CHEMISTRY. 157 ence of Rochelle salt and of tartrate of potash and ammonia, which can be obtained from the acid potassium compound by neutralising with the corresponding bases, prove to Liebig that tartaric acid also possesses the property of neutralising two atoms of base ; and this leads him to double its atomic weight, i.e. to write it C8H8O10. The talented author of this famous paper thus understood very well that the considerations here advanced furnish a new aid to the determination of the atomic weight. Liebig justifies the separation of the acids into different groups, in the following words : — 13 " The acids might be divided into monobasic, dibasic, and tribasic. By a dibasic acid there would be understood one whose atoms unite with two atoms of base in such a way that these two atoms of base replace two atoms of water in the acid. The conception of a basic salt thereby remains un- changed. . . . Accordingly, when two and more than two atoms of base combine with one atom of an acid, and only one atom of water is separated during the operation (fewer there- fore than the number of equivalents of the fixed base), then a really basic salt is produced." . . ,14 This great step was thus taken. The way had been pre- pared for it by the labours of Graham ; the change was carried through and established by Liebig's investigations. If we desire to be strictly just (and we set some value upon this), we must not suppress the fact that Liebig published the first paper on this subject along with Dumas in i837.15 This was the only fruit of the proposed association of these two chemists. In the same paper in which Liebig develops, in detail, the theory of the poly basic acids (in accordance with which the acids fall into several classes), he endeavours, by means of a 13 Annalen. 26, 169. 14 In the property of forming pyro-acids Liebig also finds a ground for classing acids as polybasic (loc. cit. 169). 15 Comptes Rendus. 5, 863. According to a letter which Liebig addressed to the French Academy in 1838 (Comptes Rendus. 6, 823 ; Annalen. 44, 57), it appears that the share of Dumas in this investigation was very unimportant. 158 HISTORY OF CHEMISTRY. [LECTURE IX. "hypothesis," to get rid of the division, which had subsisted up to this time, into hydrogen acids and oxygen acids. This hypothesis is a reversion to the ideas of Davy and of Dulong.1*'1 A similar attempt had previously been made by Clark, although much less elaborated. According to Griffin,17 Clark stated views of this kind in his lectures as early as 1826. As he himself wrote to Mitscherlich in i836,18 he finds grounds for his opinion in the isomorphism of sulphate of soda and permanganate of baryta. At that time, the formulae assigned to these compounds were : NaOSO3 and BaOMn2O7, according to which they contained an unequal number of atoms. Clark proposes to double the atomic weight of sodium (that is, to assume it to be four times as great as at present) and to assign to it the number that Berzelius adopted in iSiQ.19 Since he further regards the acids as hydrogen compounds, from which salts are produced by the replacement of the hydrogen by metals, sulphuric acid, with him, is H2SO4 and permanganic acid HMnO4; and therefore sulphate of soda is NaS2O8, and permanganate of baryta is BaMn2O8 ; and by this means he attains similarity in the number of atoms in both compounds. Quite different grounds, which are of much superior value, and are more numerous, lead Liebig to revive again the Davy- Dulong hypothesis. Graham had shown that pyro- and meta- phosphoric acids can exist in aqueous solution without at once passing into ordinary (tribasic) phosphoric acid. Liebig con- sequently inquires whether these three acids really differ from one another by an atom of water in each case, and whether it is the gain or loss of water which brings about the changes in the basicity of phosphoric acid. He does not believe that convincing grounds can be found for the adoption of this hypothesis ; so that the contrary supposition, in accordance 16 Compare p. 83. 17 Griffin, The Radical Theory in Chemistry, London (1858), 4 et seq. 18 Annalen. 27, 160. 19 Compare p. 94. LECTURE IX.] HISTORY OF CHEMISTRY. 159 with which the salts are formed by the replacement by metals of the hydrogen of the (hydrated) acid, is not to be rejected unconditionally. The accuracy of this idea being premised, the acids would not contain any ready-formed water, and they could not be any longer regarded as consisting of anhydride and water, any more than the salts were compounds composed of acid (anhydride) and base. Liebig finds an important support for the latter hypothesis (in accordance with which the metals, as such, should be assumed in salts) in the behaviour of tartar emetic at a high temperature. According to the analysis, the formula C8H8KSb2Oi4 represents the compound dried at 100°. It was assumed to contain an atom of anhydrous tartaric acid, an atom of potash, and an atom of oxide of antimony, so that its formula was written C8H8O10+ KO + Sb2O3 (assuming that the formula of tartaric acid was doubled). According to Liebig this substance, on being heated to 300°, loses two more atoms of water, a property which it does not share with any other salt of the same acid. The assumption of the presence of water in the acid, hitherto regarded as anhydrous, appears objectionable to Liebig on account of the consequences which would follow from it, and so he believes that nothing remains but to ascribe the formation of water to the reduction of the oxide of antimony. The actual existence of a base, in the metallic condition, combined with an oxygen acid (even if only for certain compounds) would no longer require to be regarded as a mere supposition.20 On another occasion, when discussing these relations, Liebig writes the formula for tartaric acid, C8H4O12.H8, and that of tartar emetic which has been heated to 300°, C8H4O12 < £ , and I must not omit to draw attention to the fact that the displacement of three atoms of hydrogen by one atom of antimony, is here assumed.21 20 Arnalen. 26, 159. 21 The paper in question is by Dumas and Liebig, Comptes Rendus. 5> 863. l6o HISTORY OF CHEMISTRY. [LECTURE IX. Liebig admits that it is difficult to understand how potash is reduced by means of sulphuric acid, an assumption which must be made in case sulphate of potash is to be regarded as a potassium compound ; but he instances a case in which a hypothesis of the kind is indispensable to the explanation of the facts. The decomposition of thiocyanate of silver by means of sulphuretted hydrogen, with the formation of sulphide of silver and free acid, would be contrary to all views regarding affinity if the salt corresponded to the formula AgS + Cy2S ; whereas the reaction becomes a normal one, assuming the formula to be Ag. Cy2S2. Unsatisfactory as it is at first sight, the hypothesis of the reduction of the oxides by means of acids, further gives an explanation of the behaviour of many acids which exhibit a higher capacity for saturation towards silver oxide than they do towards soda, although the latter is endowed with more strongly basic properties. Finally, Liebig points out that by the assumption of the hypothesis of Dulong, the grouping of the hydrogen acids and of the oxygen acids into one class, which is almost enforced by the similarities in their reactions, is attained. Thus lime always gives up the same quantity of water no matter whether it is neutralised with sulphuric acid or with hydrochloric acid. The mode of explanation then adopted, in accordance with which the water, in the one case, was present in the acid ready- formed, and in the other case was produced during the action, according to Liebig takes no account of the analogy of the two cases. He tries to pull down the barrier, and his words are sufficiently significant to deserve a place here.22 " We employ, therefore, two modes of explanation for one and the same phenomenon ; we are forced to ascribe to water the most varied properties ; we have basic water, water of hal- hydration, water of crystallisation ; we observe its entrance into compounds in which it ceases to assume any one of these three forms ; and all this is for no other reason than that we have drawn a distinction between haloid salts and oxygen salts, 22 Annalen. 26, 179. LECTURE IX.] HISTORY OF CHEMISTRY. l6l which we do not observe in the compounds themselves ; in all their relations they possess properties of the same kind." Liebig then returns to Davy's ideas, and in doing so he states his views as follows : — 23 "Acids are, accordingly, certain compounds of hydrogen, in which the hydrogen can be replaced by metals. " Neutral salts are those compounds of the same class, in which the hydrogen is replaced by the equivalent of a metal. Those substances which we at present call anhydrous acids, acquire the property of forming salts with metallic oxides, for the most part, only on the addition of water ; or they are com- pounds which decompose the oxides at a high temperature. "On bringing together an acid and a metallic oxide, the hydrogen is separated, in the majority of cases, in the form of water. It is a matter of complete indifference for the constitu- tion of the new compound, in what manner the formation of this water is conceived : in many cases it is formed by the reduction of the oxide ; in others it may be produced at the expense of the elements of the acid — we do not know which. " We only know that, without water, no salt can be pro- duced at ordinary temperatures, and that the constitution of the salts is analogous to that of the hydrogen compounds which we call acids. The principle of the theory of Davy, which must be especially kept in sight in criticising the theory, is that he makes the capacity of saturation of an acid dependent upon the hydrogen or upon a part of the hydrogen which it contains ; so that, if the other elements of the acid, collectively, are called the radical, the composition of the radical does not possess the most remote influence upon this capacity." These statements are recognised, on the whole, as correct even at present. Together with what I have stated above con- cerning the polybasic acids, they constitute the basis of our views regarding acids. No doubt the characters which dis- tinguish polybasic from monobasic acids were considerably extended by Gerhardt and Laurent, so that the conceptions and 23 Annalen. 26, 181. L 1 62 HISTORY OF CHEMISTRY. [LECTURE IX. definitions assumed a much more fixed and decided form. The distinction between the basicity and the atomicity of an acid was learned at a still later date, and rules were formulated whereby these also can be ascertained numerically. But these developments fall into a period which is too far removed from the one now under consideration for us to be able to discuss them here at present. It will easily be understood that Berzelius could not share Liebig's opinions. To recognise them would have been to abandon dualism, the basis of his own theories. It is true that the new way of looking at substances was not purely unitary ; the acids were supposed to consist of radical and hydrogen, and the salts, of radical and metal, so that there still existed a division into two parts ; but this was in a sense in which Berzelius could not admit it. The mode of salt forma- tion, as Liebig conceived it, must especially have been in opposition to his views. There were no longer twro com- pounds of the first order — an electro-positive and an electro- negative constituent — which united ; the formation of salts consisted, instead, in the replacement of hydrogen. How could this be reconciled writh. the electro-chemical theory, in accord- ance with which compounds are only formed by the union of atoms one with another? Hence we find Berzelius also pro- testing 24 against the theory of hydrogen acids, if I may thus designate Dulong's ideas. His reasons wrefe not sufficient, however, to dissuade the greater number of chemists from adopting this theory, and on this account I shall not enter more minutely into the matter, but shall again turn to the facts which were to lead to a unitary system. I refer to the phenomena of substitution, or to the replaceability of hydrogen by electro- negative elements. Besides Dumas and Laurent, Regnault and Malaguti were especially engaged in the investigation of this subject. The results obtained by these chemists — the theories of Laurent, as well as Liebig's views concerning acids — had not been without 24 Berzelius, Jahresbericht 1839, 264 ; Annalen. 31, I. LECTURE IX.] HISTORY OF CHEMISTRY. 163 influence upon Dumas. An extremely interesting discovery, which he makes in 1839, obliges him to expound his views at this time with respect to substitution ; and also to retract, partially at least, the statements previously advanced and to substitute in their place new ones of far greater significance. In this way, from the empirical rules of substitution, the theory of types arises. By the action of chlorine, in sunlight, upon acetic acid, Dumas had obtained a crystalline substance whose composition could be expressed by the formula C4CleH2O4,25 and which could, therefore, be regarded as acetic acid, C4H8O4, in which six atoms or volumes of hydrogen were replaced by six atoms of chlorine.26 The interesting and important part of this reaction lay in the properties of the new compound, which Dumas called chloracetic acid. This acid had the same saturating capacity as acetic acid, so that Dumas was able to assert that by the entrance of chlorine in place of the hydrogen, the chief character of the compound was not altered ; or, as he expresses himself, " that in organic chemistry there are certain types, which persist even when an equal volume of chlorine, bromine, or iodine is introduced into them in place of the hydrogen which they contain." It will thus be seen how Dumas, in consequence of his dis- covery of chloracetic acid, is led to the same point of view which had already been taken up by Laurent, but which the former had at first put aside as extending beyond the limits of fact.27 It is, however, an injustice to Dumas to represent his theory of types as merely an application or perhaps an expansion of Laurent's ideas. Laurent was a clever speculative thinker ; but he did not hesitate to state a hypothesis for which a complete scientific proof could not, at the time, be adduced, and this, I think, was the case with respect to his views concerning sub- stitution. That this, at least, was the impression made upon his contemporaries is to be seen from Liebig's criticism of 25 In the French papers Dumas retains the atomic weight C = 6. 26 Annalen. 32, 101. 27 Comptes Rendus. 6, 689. At that period Dumas called Laurent's theory an extension of his ideas which did not concern him. 164 HISTORY OF CHEMISTRY. [LECTURE IX. Laurent's theories.28 The facts were still wanting, which showed, in a definite and decisive manner, the analogy between the original substance and the final product. Our science cannot advance by means of ideas alone : it is only when an opinion is called forth, and to a certain extent necessitated, by experi- ment that a further development is involved in it. It was not Dumas' position and name alone that now procured favour for the theory which, a year before, had scarcely been taken notice of. The chemists of that period had no such respect for authority. The discovery of chloracetic acid lies between the promulgation of the nucleus theory and that of the theory of types ; and even if " a theory can be made up of words " still, in chemistry, greater value is, fortunately, placed upon a decisive experiment than upon daring speculations. An analogy between acetic and chloracetic acids could not remain unobserved ; and especially after Berzelius, who had his reasons for not admitting any similarity between them, had given prominence to their differences ; and, with a certain amount of irony, inquired for their kindred relationships.29 Dumas shows the reactions which they undergo by the in- fluence of potash, and points out their similarity.30 We have — C4H2H804 « C204 + C,H2H, 31 Besides carbonate of potash, there is formed, in the one case, marsh gas, and in the other, chloroform ; i.e., two sub- stances which again exhibit the same difference in composition, the one from the other, as the two acetic acids do ; and of which the latter, as Dumas also particularly shows,32 can be obtained from the other by the action of chlorine. By the discovery of trichloracetic acid a basis is furnished 28 Annalen. 25, I. 29 Berzelius, Jahresbericht 1840, 367 etc. 30 Annalen. 33, 179. 31 This latter reaction appears to have been dis- covered previously by Persoz (Introduction a 1'Etude de la Chimie mole- culaire), as was pointed out by Pelouze and Millon (Annalen. 33, 182). a2 Annalen. 33, 187 and 275. LECTURE IX.] HISTORY OF CHEMISTRY. 165 upon which Dumas erects his theory of types. 33 Thus, accord- ing to him, all substances which contain the same number of equivalents combined in the same way, and of which the chief characters are similar, belong to the same chemical type. These are, for the most part, compounds which can be obtained from one another by very simple reactions, such as acetic acid and chloracetic acid ; chloroform, bromoform, and iodoform ; ethylene and the products arising from it by substitution by means of chlorine. Dumas thinks he has found, in the conception of the chemical type, the basis of a new classification, which includes the recently observed facts ; but he employs, at the same time, the molecular type introduced by Regnault,34 which he calls also the mechanical type. To this type the following com- pounds belong : — Marsh gas C2H2H6 Methyl ether - C2O H6 Formic acid - - C2H2O3 Chloroform - C2H2C16 Chloride of methyl C2C12H6 Chloride of carbon C2C12C16 These substances, which may be regarded as arising by substitution from one another, and which may possess very different properties, are classed in one natural family. The point of view which led to the establishment of Regnault's types, is a much more comprehensive one than that from which Dumas was induced to advance his chemical types ; the substances embraced under the latter heading constituting merely a subdivision of those which must be classed under the same mechanical type. Dumas also sees this clearly, for he says : 35 " On every occasion when a substance undergoes change without quitting its molecular type, it is changed in accordance with the law of substitution. On every occasion when a substance passes, on undergoing modification, into 33 Annalen. 33, 259; 35, 129 and 281 ; 44, 66. 34 Ibid. 34, 45. 35 Ibid. 33, 279. 1 66 HISTORY OF CHEMISTRY. [LECTURE IX. another molecular type, the law of substitution is no longer adhered to during the reaction." And further: "Alcohol, acetic acid, and chloracetic acid belong to the same natural family ; acetic acid and chloracetic acid to the same species." It may therefore be said that, so far as the idea goes, mechani- cal type and nucleus amount to the same thing; both comprise the substances which arise from one another or which can, at least, be looked upon as arising from one another by equivalent substitution. Dumas, as may have been observed, has now arrived at the opinion that his law of substitutions is not applicable to all reactions, and that an equivalent of another element is not always taken up even for the hydrogen removed. He is all the more obliged to admit this, since he now no longer assumes the existence of ready formed water in organic substances (alcohol, for example36) whereby his second rule ceases to hold.37 He is obliged, in consequence, to recognise, and he does it explicitly, that the phenomenon of substitution is not a general one ; he even finds in this one of its most essential features. 38 While he thus limits the applicability of the law of substi- tution, the validity of the law in another direction is enhanced. It is not only the hydrogen of an organic substance that can, according to Dumas, be replaced, but all the elements which it contains. True substitution of the oxygen, of the nitrogen, and even of the carbon may be accomplished ; 39 and these elements may be replaced, not only by others, but also by compound groups such as cyanogen, carbonic oxide, sulphurous acid, nitric oxide, nitrous acid, amide, etc. The assumption of the replaceability of carbon, which at that time met with the most vigorous contradiction as a silly hypothesis, and was, in Germany, made the subject even of ridicule,40 was a conse- quence of the experiments of Walter41 who had obtained sulpho- camphoric acid by the treatment of camphoric acid with 36 Annalen. 33, 261. :<7 Compare p. 141. 38 Annalen. 33, 264. » Ibid. 33, 269. 40 Ibid. 33, 308. 41 Ibid. 36, 59. LECTURE IX.] HISTORY OF CHEMISTRY. 167 sulphuric anhydride, carbonic oxide being evolved. Dumas regarded this acid as camphoric acid in which one atom of carbon was replaced by the group SO2. If the conception of the molecular type is regarded in its widest sense, it may be said that this idea of Dumas as to the replacement of carbon, was wholly justified by subsequent experiments. Wohler has pointed out a substitution of carbon by silicon;42 and by means of reactions quite analogous to those employed for converting a hydrocarbon into the corre- sponding alcohol, Friedel and Crafts have transformed ethyl silicide into silico-nonyl alcohol which, as the name indicates, they look upon as nonyl alcohol in which one atom of carbon is replaced by one atom of silicon.43 More recently, silicon compounds have been discovered which are not only to be regarded as analogous to certain carbon compounds, but which behave in a manner similar to the latter. This is particularly the case with triethyl silicol.44 I may also remark here that the view of Dumas concerning the replacement of carbon was in contradiction to the nucleus theory of Laurent, and rendered difficult the classification of organic substances according to the number of their carbon atoms. The ideas of the two chemists approach each other more closely as regards the conception of the radical than as regards its composition. Dumas now expressly slates also that the radical is not an unalterable group, but that in it, just as in all compounds, the atoms are replaceable by others. Gerhardt had, however, advanced similar views two years earlier, and we shall, therefore, have to deal fully with this point afterwards in another lecture. The first, and probably the most important consequence of the theory of types was that it demanded a unitary mode of regarding substances. The compound was no longer to be regarded as consisting of two parts. It constituted, rather, a 42 Annalen. 127, 268. 43 Comptes Rendus. 6l, 792 ; also Annalen. 138? J9 > compare further Friedel and Ladenburg, Annalen. 143, 118; 145, 174 and 179; 147, 355 ; and Comptes Rendus. 66, 816. 44 Laden- burg, Annalen. 164, 300. 1 68 HISTORY OF CHEMISTRY. [LECTURE IX. uniform whole, which might undergo change from the fact that one atom could take the place of another. Dumas compares it with a planetary system ; the atoms here represent the indi- vidual planets, and they are held together by affinity instead of by gravitation. The atoms in this system can be replaced by others : so long as the number of the equivalents and the relative positions of the atoms are preserved, the system is unchanged. According to the theory of types, the properties of a com- pound were affected far more by the arrangement than by the nature of the atoms; and this doctrine which Dumas now defends as confirmed by experiment, leads him to an attack upon the electro-chemical theory. This is how he expresses himself :— 45 " One of the most immediate consequences of the electro- chemical theory is the necessity of considering all chemical compounds as binary substances. It is necessary to find out, in every one of them, the positive and the negative constituents, or the groups of particles to which these two distinctive char- acters are ascribed. No view was ever more fitted to retard the progress of organic chemistry." And in another place : 46 "In general, when the, substitution theory and the theory of types assume similar molecules, in which some of the elements can be replaced by means of others without the edifice be- coming modified either in form or outward behaviour, the electro-chemical theory splits these same molecules, simply and solely, one may say, in order to find in them two opposite groups, which it then supposes to be combined with each other in virtue of their mutual electrical activity." Dumas does not deny the influence of electrical forces upon chemical reactions. On the contrary, chemical and electrical forces might, according to him, even be identical. What he attacks is the electro-chemical theory of Berzelius, in accordance with which hydrogen is supposed to be always positive, and chlorine always negative. He believes that in the formation or decomposition of compounds he can recognise 45 Annalen. 33, 291. « Ibid. 33, 294. LECTURE IX.] HISTORY OF CHEMISTRY. 169 the action of electrical forces ; but what he declares to be erroneous, and irreconcilable with the phenomena of substi- tution, is the assumption that the electrical state of the atoms is unchanging. The fatal moment had now arrived ; it was a question of defending dualism, and the electro-chemical theory, which was in the fullest agreement with it, and which had prevailed, almost unattacked, for nearly twenty years, against the views that had been stated in opposition to it. Ways and means had to be devised whereby the newly-discovered facts with respect to substitution could be brought into harmony with the electro-chemical ideas. Before the storm actually broke, Berzelius had perceived the threatening clouds gathering about him, and had taken his precautions. As soon as Laurent had assumed, in his first papers, that the hydrogen of the nucleus (or fundamental radical) could be replaced by chlorine, Berzelius, quite cor- rectly perceiving the danger to his theories which such views might possess, repudiated energetically the statements of Laurent.47 The entrance of electro-negative elements into radicals is put aside as an untenable hypothesis ; and even the oxygen radicals, which he had greeted with so much pleasure a few years previously, are discarded. This assumption is, according to Berzelius, " of the same kind as that which would regard sulphurous acid as the radical of sulphuric acid, and manganese peroxide as the radical of manganic acid. An oxide cannot be a radical. The conception of the word radical is such that it represents the substance which, in an oxide, is combined with oxygen." Berzelius now only recognises radicals which contain carbon and nitrogen, carbon and hydrogen, or carbon, nitrogen, and hydrogen. Sulphur "cannot enter into the composition of a radical any more than oxygen can." " The ternary radicals " must therefore be regarded " either as compounds of a binary sub- 47 Berzelius, Jahresbericht 1839, 358. 170 HISTORY OF CHEMISTRY. [LECTURE IX. stance with a simple one, or as compounds of two binary substances." 4S The radical CUH10 is now constituted the basis of the com- pounds discovered by Liebig and Wohler, and this is justified by the analogy which benzoic acid, benzoyl, and the hydrocarbon C14H]0 exhibit with manganic acid, manganese peroxide, and manganese. Thus : C14H10O3 Benzoic acid - - MnO3 Manganic acid. C14H10O2 Benzoyl MnO2 Manganese peroxide. C14H10 Mn Manganese.49 Berzelius regards benzoyl chloride as similar to chromyl chloride, adopting for the latter the formula of H. Rose.50 He writes : 2CrO3 + CrCl6 . Chromyl chloride. 2C14H10O3 + C14H10C16 - Benzoyl chloride. Quite analogous with this, is the formula of phosgene, which substance Dumas regarded as carbonic acid in which one atom of oxygen is replaced by two atoms of chlorine.51 Berzelius writes CO2 4- CC14, Phosgene. At this time Berzelius is putting dualism more prominently forward than ever, and in his view, these formulae are entirely justified. "Since, in accordance with our present views, the forces which bring about chemical combinations do not act between more than two substances of opposite electro-chemical tendencies, all compound substances must permit of being split into two constituents, of which the one is electro-positive and the other electro-negative."52 In consequence of these views, all substances which, besides carbon and hydrogen, also contain oxygen, chlorine, bromine, or sulphur, break up into several parts which often appear to be chosen quite arbitrarily; further the atomic weight is fre- quently doubled or trebled, so that the subdivision into binary 48 Annalen. 31, 13. 49 Compare also Berzelius, Lehrbuch. Third Edition, 6, 205. 50 Pogg. Ann. 27, 573. 51 Dumas, Traite, I, 400. 5a Annalen. 31, 12. LECTURE IX.] HISTORY OF CHEMISTRY. iyi radicals may be carried out. Highly complicated formulae are thus obtained, of which I can only instance a few : Malaguti's chlorinated ether : C4H0O3+2C4H6C16.53 Malaguti's chlorosulphuretted ether : (C4H603 + 2C4H6C16) + (C4H603 + 2QHA) etc.54 His conception of chloracetic acid is very important as we shall find further on ; he regards it as a compound of oxalic acid with chloride of carbon : C2C13 + C203 whilst acetic acid remains as the trioxide of the radical acetyl, C4H6 or C4H:,. Even in 1840 he still contests the similarity in the constitution of the two compounds, and does not permit himself to be shaken in this by their analogous behaviour with potassium hydroxide.55 This view was not long tenable, however, in face of the constantly increasing number of substitution products, very many of which exhibited unmistakable analogies with the original substances. When Melsens succeeded, in 1842, in reconverting chloracetic acid into acetic acid, by treatment with potassium amalgam,50 and thus proved that chlorine can be replaced by hydrogen again so that the original substance is reproduced, even Berzelius was compelled to make an admis- sion. He says : r>~ — " If we recall to memory the decomposi- tion of acetic acid by means of chlorine with formation of Chlorkohlenoxahdure (chloracetic acid), another view as to the composition of acetic acid presents itself as possible. In accordance with this view it would be a coupled 5S oxalic acid whose copula is C2H;>, as the copula of the Chlorkohlen- oxalsciiire is C2C13 ; consequently the action of chlorine upon acetic acid would consist in the conversion of the copula C2H3 into C,C1,." 53 Berzelius, Jahresbericht 1840, 375 ; here also Berzelius still writes H instead of H._,. 54 Annalen. 31, 113 ; 32, 72. 5r> Ibid. 36, 233. 5(i Ann. Chim. [3] 10, 233. 57 Lehrbuch. Fifth Edition, I, 460 and 709. r>8 Berzelius here employs a word that had been introduced into the science by Gerhard t. I72 HISTORY OF CHEMISTRY. [LECTURE IX. Perhaps Berzelius did not notice that he had thus conceded the chief point in the theory of substitution which, a few years before, he had vigorously contested. Chlorine could replace the hydrogen of the " copula," and the constitution of the compound was not essentially altered thereby. Berzelius now wrote : C2O8 + C2C13 Chloracetic acid. C2O3 + C2H3 Acetic acid. Was a fundamental principle of the electro-chemical theory not violated by this concession? I think it was. It was now necessary to assume either that forces different from the electrical forces were present in the copula, or that the electri- cal properties of the elements are altered in the compound ; and both of these assumptions were much at variance with the former ideas of Berzelius. The substitution theory had thus come off victorious. Berzelius, it is true, never admitted his defeat, but, as a matter of fact, he had given in. The electro- chemical theory was now abandoned. In its last throes it had produced the idea of copulae : had these latter any vitality ? At first it did not seem so ; they were looked upon as the idle invention of a wearied intellect. With this I partly agree ; but they still possessed a spark of life, otherwise they would not have been capable of development, and it would not have been possible even for a man like Kolbe to advance them to what they afterwards became. This subject will be dealt with in a subsequent lecture. LECTURE X. INFLUENCE OF THE SCHOOL OF GMELIN — THEORY OF RESIDUES — COUPLED COMPOUNDS — GERHARDT'S DETERMINATION OF EQUIVA- LENTS— DISTINCTION OF ATOM, MOLECULE, AND EQUIVALENT BY LAURENT — NEW CHARACTERISTICS OF POLYBASIC ACIDS — MOLECULES OF THE ELEMENTS. THE battle was over, and the victory won. It had been shown that, starting from the decompositions which substances undergo under the influence of the \fllyanic current, we are not in a position to explain the manifoTd reactions of organic chemistry, and in particular, the phenomena of substitution. The founda- tions had been shaken by the fact that positive hydrogen was replaceable by negative chlorine, and the whole edifice — the electro-chemical theory — collapsed. Organic chemistry had shown that laws which had been advanced without reference to the facts that it presented, did not harmonise with the teach- ings of these facts. It was now a question, however, as to whether organic chemistry could also render any positive ser- vice— whether it would be possible, starting from the facts which it had already furnished, or which it would furnish in the future, to establish new principles that might serve as the basis of a chemical system. The way of looking at substances from the electro-chemical point of view, and dualism also, were maintained in the case of inorganic compounds. In order to do this, a sharp distinction between the latter and organic compounds became necessary, so that it might be possible to apply a doctrine to inorganic compounds which had proved inapplicable to organic com- pounds. If this doctrine was to be completely supplanted, however, and if organic chemistry was to enjoy the fruits of its victory, it was necessary that this younger branch of the science 174 HISTORY OF CHEMISTRY. [LECTURE X. should offer to the opposing party definite principles upon which it might be reconstructed. It was not in a position to do this at first, since, up to this time, more attention had been paid to overthrowing the old system than to building up a new one. It is true that endeavours were made on various sides to embrace all organic compounds in one uniform conception. The radical theory, the nucleus theory, and the theory of types had arisen, and each had its supporters ; but the very fact of there being so many views, proved the insufficiency of any of them. We find, then, a great deal of confusion ; the adherents of the different systems were in continual strife, and a becoming demeanour was not always maintained. Consequently, it is difficult to say which were the prevailing ideas at the beginning of the fifth decade of the century. Even the views as to the principles of each mode of regarding con- stitution were widely at variance from one another. The school of Gmelin had greatly augmented its adherents, and to the latter the atomic theory appeared too hypothetical. We cannot be surprised to find that chemists now begin to lean more and more in this direction, since even the expression " atomic weight " is gradually supplanted by the " equivalent," and the latter is employed, as it had been by Wollaston, in the sense of combining weight.1 Upon the overthrow of the system of Berzelius (that is to say, of the only system which, in any uniform sense, embraced the whole science), and with the origination of the most various hypotheses and theories, which were not capable of any general application and did not seem to have any promise of a long existence, there arose in the minds of many a certain aversion to all speculation, which was looked upon as premature and hurtful to the science. Nothing was in keeping with the times except the temperate considera- tion of observations, and Gmelin was the right man to represent a tendency of this kind. He united boundless industry with wide knowledge, and he understood how to turn both of these qualities to account in his Hand-book. In adducing facts 1 Compare Liebig, Annalen. 31, 36. LECTURE X.] HISTORY OF CHEMISTRY. 175 completeness and conscientiousness were his watchwords, and these were adhered to. Since formulae, for Gmelin's school, merely represented the composition of substances by means of a contracted style of writing them, these chemists were at liberty to choose their " equivalents " or "combining weights" at will from the pos- sible multiples. The guiding principle appeared to them to be simplicity in symbolising, and therefore their numbers possess little real significance as regards development. I shall merely remark that they adopted the formulae of Berzelius for most compounds, and, in doing so, they regarded the double atom as an equivalent. They arrived at this by halving the atomic weights of oxygen, sulphur, carbon, selenium, etc., with refer- ence to hydrogen, chlorine, bromine, iodine, nitrogen, phos- phorus, and the metals.'2 It must not be supposed, however, that, at the beginning of the forties, the atomic weights of Berzelius were no longer used. On the contrary, they were still employed by Liebig and his numerous and important followers,3 and it was only towards the end of the decade (after the appearance of Gerhardt's paper) that the latter also made use of Gmelin's equivalents. I have already indicated in a previous lecture what the reasons were that brought about this revolt from the atomic theory.4 Although I said there that none of the physical rules which expressed relations between the atomic weight and certain properties of matter, appeared capable of general application, on the other hand, the law which had induced Dalton to advance the atomic theory, that is, the law of multiple proportions, was still un- attacked. The examination of numerous organic compounds whose investigation was already completed, had only tended to confirm it. No doubt, it had become necessary to admit that a much larger number of atoms can unite with one another 2 Compare Gmelin, Handbuch. First Edition, 34. At this time Gmelin also halves the atomic weight of phosphorus, and so writes phosphoric acid PaO.-» as Berzelius did. 3 The sign for the double atom is not employed in Liebig's Annalen, and HLO is therefore printed instead of HO, as Liebig observes from want of the necessary types. 4 Compare p. 1 06. 176 HISTORY OF CHEMISTRY. [LECTURE X. than either Dalton or Berzelius had considered possible, and the law had, on this account, obviously to some extent lost its definite character. Even at that time the question might have been asked, whether the statement could be called a law, since chemists were not in a position to determine anything as to the limits of the power of combination of atoms ; and also whether every compound could ultimately be referred to unvarying weights of the constituents, if any large multiple might be chosen at will. Ideas of this kind do not appear, however, to have arisen at that period,5 and so there always remained this one generalisation for those who tried to retain the atomic theory and to advance speculations as to the constitution of compounds. Amongst these, Dumas had in the recent years played the most important part, by founding his theory of types. In this theory, which was, no doubt, partly borrowed from Laurent, there was much that was eminently suited for a classification of organic compounds ; but still the use of it was only recognised more generally after its fusion with the radical theory, that is, after radicals had been introduced into the types. This could only take place after the conception of the radical had been com- pletely transformed, and it is now my business to show how and by whom this development was brought about. On studying the writings of the founders of the radical theory, one might be tempted to assert that it was they who not only established the conception of the radical in its first signification, but that, at the same time, they had also done the most important service in respect of the subsequent acceptation of the word. Thus, the following passage from Berzelius is noteworthy : — 6 "We shall assume that, by means of any circumstance whatsoever, we could clearly see the relative position of the simple atoms in the compound atom of the salt [sulphate of copper]. It is clear that, whatever this may be, we should then find in it neither oxide of copper nor sulphuric acid, for 5 Compare, however, Berzelius in Liebig's Annalen. 31, 17 ; also Dumas, ibid. 44, 66. 6 Jahresbericht 1835, 348. LECTURE X.] HISTORY OF CHEMISTRY. 177 the whole is now a single coherent substance. We can picture to ourselves ... the elements in the atom of the salt as coupled together in various ways ; for example, as one atom of sulphide of copper combined with four atoms of oxygen, that is to say, as the oxide of a compound radical ; as one atom of binoxide of copper and one atom of sulphurous acid ; as one atom of copper and one atom of a salt-former, SO4; and, finally, as one atom of oxide of copper and one atom of sul- phuric acid. So long as the simple atoms remain together, one of these notions is as good as another. If it is a question, however, of the behaviour when the compound atom is decom- posed by electricity, or by the action of other substances (in the wet way for instance), then the relation is quite different. The compound atom in that case never undergoes decom- position in accordance with the first two views, but it does according to the two latter. The copper can be exchanged for other metals according to the view Cu + SO4 ; but if the copper is taken away without being replaced, as is the case by the action of electricity, then that part of the atom of the salt which remains over, breaks up into oxygen and sulphuric acid. If, on the contrary, the salt of copper is decomposed either by a very feeble electrical force or by means of other oxides, into oxide of copper and sulphuric acid, both of these remain after- wards, and from them the salt can be compounded again. There must naturally be a reason for these circumstances, and the reason can scarcely be other than this, that when sulphuric acid and oxide of copper unite to form a compound atom of the salt, the relative positions of the atoms in the united binary substances do not materially change, and the latter can thus be combined or separated as often as is desired. . . . From this, however, it easily follows that in decomposing to form other binary com- pounds of the elements, the atoms must undergo a trans- position of their relative situations, so that their capacity for combining anew is either diminished, or, as is usual, ceases entirely. Nitrate of ammonia, which is decomposed into nitric acid, ammonia, and water, and is recompounded from these, can be decomposed by heat into nitrous oxide and water, M 178 HISTORY OF CHEMISTRY. [LECTURE X. without our afterwards being able to recompound it from these substances. The reason for this fact must be that, in the latter mode of decomposition, the atoms of the elements are trans- posed into other relative situations which are obstructive to their reuniting." These ideas are as clear, as impartial, and as unprejudiced as could possibly be desired. The same holds also for the following statements of Liebig : 7 " A theory is the explanation of positive facts, which does not permit us, from the behaviour of a substance in various modes of decomposition, to make deductions backwards as to its constitution, with conclusive certainty, simply because the products vary with the conditions of the decomposition. " Each view as to the constitution of a substance, is true for certain cases, but unsatisfactory and insufficient for others." Even if I also admit that the principles of the newer radical theory are expressed in these statements of the two great teachers,8 still it would appear to me presumptuous that any one should declare the latter, on this account, to be the authors of the views which are now to be discussed. By their activity in other directions in the domain of theoretical chemistry, they have shown that, with them, the radical is a definite, unchang- ing group, and that they only considered a single view as to the constitution of compounds to be admissible. I would recall the numerous discussions regarding the conceptions of alcohol and its derivatives. Would these have been possible if opinions such as those quoted above had been guiding and predominant ideas with Berzelius, Liebig, and Dumas? Prior to the dis- covery of the phenomena of substitution, this certainly was not the case. In treating of Berzelius and of Dumas, we have already discussed the influence which these facts exerted upon the conception of the radicals. We have still got to consider 7 Annalen. 26, 176-177. 8 Thus Gerhardt, for example, begins. the exposition of his theoretical views by directing attention to the various formulae that are possible for sulphate of baryta ; that is to say, he adduces considerations quite similar to those quoted above as advanced by Ber- zelius. See Gerhardt, Traite de Chimie. 4, 561. LECTURE X.] HISTORY OF CHEMISTRY. 179 Liebig's relations to them. His opinion concerning Laurent's nucleus theory must not guide us here. The discovery of trichloracetic acid had not been without effect upon him also, and he not only admits the replaceability of hydrogen by nega- tive elements, but he also agrees with Dumas in his views respecting these facts. This is seen from the following : 9 "The remarkable observation has been made in inorganic chemistry, that the manganese in permanganic acid may be replaced by chlorine without altering the form of the com- pounds which permanganic acid can produce with the bases. There can scarcely be a greater dissimilarity in chemical pro- perties than that between manganese and chlorine. . . . Chlorine and manganese can replace each other in certain compounds without alteration in the nature of the compounds. I do not see why a similar behaviour should be impossible with other substances — with chlorine and hydrogen, for example, and this very view of these phenomena, in the form in which it has been advanced by Dumas, appears to me to furnish the key to most of the phenomena of organic chemistry." Dumas, it is true, goes too far for him. Liebig will not admit, for example, the replaceability of carbon, and he prints in his journal the well-known letter of S. C. H. Windier 10 which makes sport of Dumas in a somewhat harsh manner. However this may be, it might appear from these statements that Liebig had materially contributed by his views to the further develop- ment of the radical theory. This is not my belief, and I find support for my view in a paper on the theory of ether, which he published in the year i839.n Liebig here endeavours to solve the difficulties of the question as to the constitution of ether, by the assumption of the radical acetyl ; and, in doing so, he proves to us that radicals, with him, still retain their old signification. The whole plan of his Hand-book also shows the same thing.12 9 Annalen. 31, 119, Note. 10 Ibid. 33, 308. n Ibid. 30, 129 ; compare p. 137. 12 See Liebig's Handbuch der Chemie, Heidelberg (1843), especially 2, I etc, l8o HISTORY OF CHEMISTRY. [LECTURE X. In my opinion, the labours of Laurent and of Gerhardt chiefly contributed to make the radical what it still is. Laurent, by advancing the nucleus theory, emphasised the variability of the radicals, a matter that was afterwards brought forward by Dumas also.13 But it was Gerhardt who first indicated the possibility of the assumption of two radicals in one compound, and thereby destroyed all idea of the actual existence of sepa- rated groups. It is with the history of this part of the develop- ment of organic chemistry that we are now concerned. The influence? of his gifted teacher, Liebig, can scarcely fail to be observed in Gerhardt's earliest publications. We are aware that Liebig contests the presence of water in the acids. By a very happy extension of the idea, Gerhardt negatives the pre-existence of water in the majority of organic compounds. Its presence in alcohol, especially, appears to him just as unlikely as that of ammonia in the substances containing nitrogen from which ammonia is evolved by means of potash. He knows that there is a class of substances of simple com- position and of extraordinary stability, such as water, carbonic acid, hydrochloric acid, and ammonia, some of which are pro- duced in almost every, organic decomposition, without, however, our being able to recompound, from these substances, the sub- stances originally decomposed.14 The formation of one substance out of another was not, with Gerhardt, any ground for the assumption that the first was present in the second, ready formed : substances do not require to contain any water in order that water may be separated from them in certain reactions. The reason for the frequent formation of water and similar substances is to be found in their stability, and in the great affinity which their constituents possess for one another. Now this view was of essential significance, and it led Gerhardt, in 1839, to the theory of residues and of coupled compounds.15 He says16 that when two substances react upon each other, an element 13 Compare pp. 146 and 167. 14 J. pr. Chem. 15, 37. l5 Ann, Chim. [2] 72, 184. I6 Comptes Rendus, 20, 1031, LECTURE X.] HISTORY OF CHEMISTRY. l8l (hydrogen) separates out from the one, and unites with an element (oxygen) from the other to produce a stable compound (water), whilst the residues join together. Mitscherlich's nitro- benzene17 is to be regarded, according to Gerhardt, as pro- duced in this way out of a residue of benzene and a residue of nitric acid. The hydrocarbon gives up hydrogen, and the nitric acid gives up oxygen. Sulphobenzide 18 is also regarded in the same way ; it contains the residues C24H10 from the benzene, and SO2 from the sulphuric acid.19 The SO2 in this case is not identical with sulphurous acid, as the latter occurs, for instance, in sulphite of lead, but it is contained in the compound in quite a special form, namely as a substituted group. Although this latter view is a very peculiar one, still it was eminently suited to supplant the belief in the pre-existence of radicals. The residues were imaginary substances, which were all the more completely deprived of any reality by the fact that they were assumed to be different from the similarly composed atomic groups which occurred in the free state. About two years later (in 1841) Mitscherlich20 propounds similar ideas, which, however, he extends to a much larger class of substances. According to him, also, compounds do not contain any ready-formed radicals which play the part of elements in decompositions ; in the appearance of water, he sees the reason for the observed direction taken by decom- positions; and therefore he does not seek for this reason in the constitution of the substances employed. The products which are obtained by the action of acids upon bases or alcohols (salts and esters), are likewise considered from this point of view, and it is shown that they decompose again into their constituents by taking up water. The idea of the residues was well adapted to explain the phenomena of substitution ; the latter, according to Gerhardt, obeyed the following rule.21 The element eliminated is replaced 17 P°gg- Ann. 31, 625. 18 Compare Mitscherlich, Pogg. Ann. 31, 628. 19 Dumas' atomic weights : C = 6, O= 16, 8 = 32 etc. 20 Pogg. Ann. 53, 95. 21 Ann. Chim. [2] 72, 196. 182 HISTORY OF CHEMISTRY. [LECTURE X. either by an equivalent of another element, or by the residue of the reacting substance. The application of this rule was limited, however, for Gerhardt, besides substitutions, is also acquainted with additions, and these of two kinds. Firstly there are those in which the saturating capacity is altered, and the formation of salts is reckoned amongst these ; and then there are addi- tions in which this is not the case. Gerhardt directs his attention chiefly to the products of the latter kind of addition, and calls them coupled compounds (corps copules). To this class there belong, in particular, the substances produced by the action of sulphuric acid upon organic compounds, such, for example, as benzoyl sulphuric acid and its salts, discovered by Mitscherlich.22 The acid is produced by the action of sul- phuric acid upon sulphobenzide. According to Gerhardt, the two substances become coupled, whereby the saturating capacity of sulphuric acid, which was still regarded at that time as monobasic, remains unchanged. Thus : — C94H10(SOa) + SO,HaO- CMH10(S02) . SO3 . H2O. Sulphobenzide. Sulphuric acid. Hyposulphobenzidic acid. Sulphovinic acid (ethyl sulphuric acid) is regarded as a coupled compound of sulphate of ethyl and sulphuric acid, and is written C8H10(SO.2)O.2. SO8H2O ; whereas sulphobenzoic acid, the basicity of which is supposed to be equal to the sum of the basicities of its constituents, and in the formation of which the saturating capacity has remained unchanged, is reckoned amongst the conjugated acids, a class of substances which were first distinguished by Dumas.23 It may, however, also be regarded as produced by the coupling of a substituted benzoic acid, CasH^SC^C^, with sulphuric acid. The view announced above with respect to coupled sub- stances, is very soon abandoned by Gerhardt. He retains the name, but gives it a different signification. But I have inten- tionally stated the older notion, because the word was also employed by Berzelius and by Kolbe, who again bestowed a <22 l>0gg- Ann- 3I> 283 and 634. M Dumas and Piria, Annalen. 44, 66 ; Ann. Chim. [3] 5, 353. LECTURE X.] HISTORY OF CHEMISTRY. 183 special signification upon it. It appeared to me of interest to follow the historical course of an expression which has been employed in senses so numerous and so varied. Gerhardt, in 1843, regards all compounds as coupled which are prepared by the action of acids upon alcohols, hydrocarbons, etc., and in whose formation the substances unite with the elimination of water.24 Coupled compounds were, accordingly, no longer addition products, obtained by the union of two com- pounds, but they were formed by the joining together of two residues. They were, therefore, substitution products — a view which Gerhardt, however, does not adopt. With him, they still constituted a special class and were not compared with the original substances, principally, no doubt, because they possessed a different saturating capacity. With respect to the latter also, Gerhardt has now become of a different opinion, and states that the basicity of the coupled compound is equal to the sum of the basicities of the substances coupled, less one. From this statement, which is advanced as an axiom, the dibasic character of sulphuric acid follows. Coupled with neutral substances, such as alcohols, or hydrocarbons, sul- phuric acid gives rise to monobasic acids, whereas acetic acid, nitric acid, hydrochloric acid, etc., do not possess this pro- perty, and are hence regarded by Gerhardt also as monobasic. In 1845, Gerhardt endeavours to show the general appli- cability of the law of basicity mentioned above.25 He now designates as coupled, all compounds which are formed by the union of two substances with the elimination of water and decompose into their constituents again by taking up water ; and, therefore, he reckons in this class, the neutral ethers, the acid ethers, etc., and formulates the law 3 = (b + V)--Lt where B represents the basicity of the coupled compound, and b and b' the basicities of the substances which take part in its formation. Gerhardt expressly remarks here that this equation 24 Comptes Rendus. 17, 312. -5 Comptes rendus des travaux chimiques par Laurent et Gerhardt, 1845, 161. 184 HISTORY OF CHEMISTRY. [LECTURE X. holds only for the coupling of a single equivalent,26 and that the equation must be applied twice in order to ascertain the correct basicity of the product obtained by the coupling of two equiva- lents of one substance with one of another. Thus sulphuric acid, for example, which Gerhardt now regards as dibasic, can form with neutral substances both acids and neutral products. The ether of sulphuric acid belongs to the latter class of sub- stances ; it is produced from two equivalents of alcohol and one equivalent of acid. Its basicity, B, is obtained from the following equations, in which B^ represents the basicity of ethyl- sulphuric acid : — - i =o. Strecker, in 1848, thought he had brought the rules into a more general form when he made a statement to the following effect : — 27 The basicity of the coupled compound is equal to the sum of the basicities of the compounds, less one-half of the number of hydrogen equivalents removed,28 or the basicity is diminished by one unit for each pair of hydrogen atoms removed. But in this way of stating the matter, the same result as was required- by Gerhardt's rule was, necessarily, always obtained. It can only be regarded as a simplification (not as a wider generalisation) in which a single application was sufficient in all cases. Although it was afterwards shown that even this form of the law of basicity does not always lead to accurate conclu- sions,29 and although the exceptional position which was given to certain classes of substances in consequence of the idea of coupled compounds, was more recently recognised as incor- rect,30 still, it cannot be denied that the assumption of the copula played a definite part in the historical development of 26 Gerhardt at that time employed the word equivalent in Gmelin's sense, so that for us it often means atom and often molecule. 2" Annalen. 68, 51. 28 Strecker assumes the equivalent of water = 9, compared with that of hydrogen chosen =i. 29 Compare Becketoff, Bulletin phys.-math. de 1s Academic de St Petersbourg, 12 (1854), 369. 30 Compare Kekule, Annalen. 104, 130. LECTURE x.] HISTORY OF CHEMISTRY. 185 chemistry. In particular, these views led to new characteristics for the recognition of polybasic acids, and this was of great importance at that time, when so few of these acids were known. The underlying idea with respect to coupled substances, that the majority of compounds could be regarded as made up of the residues of other substances, was of very great value for the progress of the science, simply because it was opposed to the rigidity and immutability of the radicals. How fertile these ideas were is shown, for example, by the discovery of the anilides and the anilid-acids. According to Gerhardt the amides are to be looked upon as compounds of the residues of ammonia and of acids ; thus he supposes oxamide to be formed according to the equation : 31 C1Ha04+2NHI = C1Ha02.(NH)2 + 2HaO [€=12, O=i6], that is to say, by the replacement of two oxygen atoms by twice the imid- residue NH. Regarding a similar replacement as also possible by means of the residue of aniline, the nature of which had been settled by Hofmann's comprehensive and interesting researches,32 and then trying to show this replace- ment by direct experiment, he succeeds in preparing oxanilide, the formation of which is represented by the following equa- tion : C2H2O4 + 2C6H7N - C2H2O2 (C6H5N)2 + 2H2O. He carries the analogy between ammonia and aniline still further by the discovery of the anilid-acids, which he regards as analogous to the amid-acids.33 Thus he writes the formula of sulphanilic acid, which he obtains by the action of sulphuric acid upon oxanilide, SH2O8 . C6H5N, and, in its existence, finds a new proof of the dibasic character of sulphuric acid. It may perhaps seem strange that Gerhardt introduces the residues NH and C6HBN into these compounds, instead of NH2 and C6H6N. In this he may have been influenced by Laurent, who had already tried, a few years previously, to 31 Comptes Rendus. 20, 1032. 32 Annalen. 45, 250 ; 47, 37. 33 Journ. de Pharm. [3] 9, 405 ; 10, 5 ; compare Annalen. 60, 308. 1 86 HISTORY OF CHEMISTRY. [LECTURE X. replace the amid- by the imid- group.34 Gerhardt was able to associate himself with this conception without involving himself in any further consequences, since his formulae did not attempt to express the arrangement of the atoms but were only contracted equations. They were not intended to represent what the com- pounds are, but merely what their nature is and what becomes of them : 35 they were meant to indicate the modes of formation and of decomposition of substances. Gerhardt was the first to advance the view that we should not conclude from the decom- position products as to the arrangement of the atoms, because the latter are set in motion by the reaction.36 According to him, several formulae were therefore possible for the same sub- stance, and different residues (radicals) might be assumed in it according to the decompositions which it was desired to emphasise. In this way the point of the controversy, which had been carried on so fiercely and so long, as to the nature of the radicals, was demolished. Afterwards, Gerhardt comes to employ empirical formulae, which had been recommended by Liebig on account of the constantly growing divergences of opinion as to rational constitution.37 Conjointly with Chancel, he introduces, in 1851,- the synoptic formulae,38 which never met with any general acceptance because they were incon- venient and not easily understood. If the form was new, still the idea was simply the old one. This mode of writing formulae was likewise intended only to represent the formation and de- composition of substances, and it too consisted of contracted equations. The great advantage of this way of regarding the matter lay in the possibility of advancing several rational for- mulae for one substance, whereby new analogies and differences made their appearance and gave rise to a large number of investigations.39 34 Comptes Rendus. I, 39. 35 Gerhardt, Introduction a 1'etude de la Chimie, 1848. x Compare Baudrimont, Comptes Rendus. 1845. 37 Annalen. 31, 36. 38 J. pr. Chem. 53, 257. 39 It may be remarked here, in passing, that Gerhardt, a few years later, again replaces imid, NH, by amid, NH2, after Laurent (J. pr. Chem. 36, 13), in 1844, had adopted and sought to establish Hofmann's conception of aniline as phenamid. LECTURE X.] HISTORY OF CHEMISTRY. 187 Gerhardt's activity was perhaps of still greater importance in connection with another question ; that is, with the fixing of the atomic and molecular weights. Although the movement for the revision of these most important numbers originated with him alone, still he was influenced in the further elabora- tion of the work by Laurent, with whom he was at that time in very intimate communication. Indeed, I might almost say that it was Laurent who first clearly stated 40 what Gerhardt wished to advance. It is, however, extremely difficult to separate from one another the services of the two chemists, because they published a great deal conjointly, and probably discussed everything together. I would ask, therefore, that my statements with respect to this matter may not be taken too literally. Gerhardt's first paper on the subject in question dates from the year i842.41 In this paper he frequently employs the word equivalent in a sense in which it was introduced into chemistry by Wollaston and Gmelin, although this sense is one of which we cannot any longer approve. With Gerhardt the word is one whose signification he does not seek for in its origin, other- wise he could not call H2SO4 and HC1 one equivalent each ; for he himself wishes to prove that sulphuric acid is dibasic, in which case it is not equivalent to hydrochloric acid. What Gerhardt wishes to determine are atomic and molecular weights, which, however, he does not yet understand how to distinguish from each other, and for which he uses the word " equivalent," at the same time designating as " atomic weights " the numbers which he is attacking. Gerhardt's equivalents, so-called, are not really equivalent, but merely comparable quantities ; and in determining them, the most various points of view which can be of consequence in estimations of atomic and molecular weights and of equivalents, are taken into consideration. It must appear striking and peculiar to any unprejudiced person, that the numbers which Gerhardt proposes as the 40 Ann. Chim. [3] 18, 266. 41 J. pr. Chem. 27, 439 ; also Ann. Chim. [3] 7, 129; 8,238. 1 88 HISTORY OF CHEMISTRY. [LECTURE X. "equivalents" of the elementary substances (with the exception of the numbers for the metals), agree almost completely with the atomic weights of Berzelius, of the year 1826. It is also noteworthy that Gerhardt does not mention Berzelius, and is obviously quite unaware that he, to a large extent, adopts his numbers. On the other hand, the Swedish chemist does not appear to have noticed this agreement, since he violently attacks Gerhardt's paper.42 What I consider as most remarkable, how- ever, is the fact that, at the time when Gerhardt makes his proposal, very eminent chemists (I only mention Liebig and his pupils43) are actually employing atomic weights for the most important elements, such as carbon, oxygen, hydrogen, chlorine, etc., with the ratios which Gerhardt recommends as new ; but that, a few years afterwards, the equivalents of Gmelin, against which Gerhardt's paper was directed, are almost univer- sally adopted. The valuable part of Gerhardt's paper consisted, however, much less in the suggestion which he makes as to the " equiva- lents " of the elements, than in his views respecting the equiva- lents of compounds. In following up, by means of equations, the decompositions of organic substances, he arrives at the general proposition that the quantities of carbonic acid, water, and ammonia produced in these decompositions are expressible as multiples, by whole numbers, of C2O4, H2O2, and NH3.44 Hence, according to him, these quantities must represent an equal number of equivalents, whereas it was at that time as- sumed that the equivalents of carbonic acid and of water were only half as great as these formulae would indicate. Guided by quite similar considerations, he fixes the equiva- lents of carbonic oxide and of sulphurous acid as G_,O2 and S2O4, and, in doing so, adds, as an important support of his assumptions, that these quantities occupy exactly the same space in the gaseous state. He is thus able to assert that the 42 Berzelius, Jahresbericht 1844, 319. * Liebig (Annalen. 31, 36) points out that it will probably never be possible to ascertain the true atomic weights, and that it is therefore better to employ the equivalents. « C = 6, 0 = 8, N=I4, H=i. LECTURE X.] HISTORY OF CHEMISTRY. 189 equivalents of carbon, oxygen, and sulphur are not 6, 8, and 1 6, as Gmelin's school assumed, but twice these numbers, that is, 12, 16, and 32; and he proves, by many examples, that there do not exist any equivalent formulae, constructed on the principles advanced by himself, which contain less of the re- spective elements than these latter quantities. He also proves that an even number of atoms of carbon, of oxygen, and of sulphur always occurs in compounds containing these elements, if they are represented by means of Gmelin's equivalents. Consequently Gerhardt doubles the equivalents of carbon, of oxygen, of sulphur, etc., with reference to those of hydrogen, of chlorine, of nitrogen, etc., whereby he obtains Berzelius' numbers. He differs very materially from the followers of Berzelius in the formulae which he proposes for organic com- pounds. According to him, these had been doubled, as com- pared with many inorganic substances, consequently he halves them ; they were, as he expresses it, referred to H = 2, or to 0 = 200, whilst, for the majority of inorganic compounds, the number chosen for compariso'n was H= i or O= 100. Hence substances were divided into those like water, carbonic oxide, carbonic acid, etc., which occupy two volumes (H = i = i vol.), and those like alcohol, ethylene, chloride of ethyl, etc. (that is, all the substances which were then called organic, and of which the vapour densities were by no means always known), which correspond to double this volume. I may be permitted to leave off the consideration of Ger- hardt's views here, in order to glance backwards and seek for the reasons which had caused chemists to write "four-volume" formulae for organic substances. This way of writing them must appear all the more remarkable since both Berzelius and Dumas, at first, at least, believed that they must choose the atomic weights of compounds so that they should represent equal volumes in the state of vapour. Relatively few vapour densities were known at that time, and hence the rule was broken in many cases without the fact being known. Another very important reason lay in the widely spread assumption that the acid was the substance united to IQO HISTORY OF CHEMISTRY. [LECTURE X. the base in a salt, or, as it can also be expressed, that the usually hypothetical anhydrides (instead of the hydrates) were regarded as acids. Thus, on the basis of the atomic weights of Berzelius, the analysis of acetate of potash led to the formula K.C4H6O4, and from this, after deduction of the potash, KO, the atom of acetic acid remained as C4H6O3, which did not permit of any further division. Those who regarded atom and equivalent as identical had to find confirmation of the accuracy of this formula in the fact that this quantity of acetic acid is neutralised by one equivalent of potash, KO ; and so the formulae of all monobasic acids were necessarily doubled. The establishment of the theory of polybasic acids caused Liebig to double the formulae of several dibasic acids, as for example, that of tartaric acid (see p. 159). This, in turn, affected the atomic magnitudes of neutral substances, such as alcohol, compound ethers, etc. To the first of these the formula C2H6O, corresponding to two volumes, had at first been as- signed, and, in order to arrive at this formula, Berzelius assumed a radical in alcohol different from that in ether.45 But Liebig, with whom alcohol was the hydrate of ether, adopted the group ethyl, C4H10, as the basis of both,46 and the close relations between alcohol and acetic acid then became promi- nent for the first time. Ethylene was now written C4Hf, and chloride of ethyl C4H10C12 ; that is to say, all the compounds of the ethyl series contained four atoms of carbon. Quite similar reasons caused the doubling of the other formulae. Gerhardt wishes, as we have stated, to halve these, and he is moved to do so from other points of view besides that of the volume relations. According to him, the salt-forming metallic oxides do not consist, as Berzelius assumes, of one atom of metal and one of oxygen, but they are comparable with water (which he now writes H2O) and contain two atoms of metal ; 4~ whereas in the hydroxides one atom of metal and one of hydrogen are united with one atom of oxygen.48 He is there- 45 Compare p. 132. ^ Compare p. 133. 4" Compare also Griffin, Chemical Recreations, Seventh Edition (1834), 92-93 and 228-229, 48 Ann, Chim. [3] 18, 266. LECTURE X.] HISTORY OF CHEMISTRY. 191 fore obliged to halve the atomic weights of the metals, and to assume K = 39, Na=23, Ca = 2o, etc. He calls that quantity of a monobasic acid an equivalent which yields a neutral salt by the replacement of one part of hydrogen by 39 parts of potassium, whilst he assumes the equivalents of the dibasic acids to be twice that quantity.40 The formula of acetic acid therefore becomes C2H4O2 while that of oxalic acid, C2H2O4, remains unchanged. That Gerhardt, in spite of these well-considered and excel- lent observations, which are now for the most part adopted, had not reached the point of view which we take up, is shown by a part of his paper 50 where he thinks he should point out that, as a consequence of his proposals, the atomic theory, the theory of volumes, and the equivalent theory all coincide. According to our present views this is not attainable. The various conceptions were first separated from one another in 1846, by Laurent,51 who thereby rendered Gerhardt's numbers admissible. He showed that these values were not by any means equivalent, and consequently did not deserve the name. They express, as he points out, those quantities which enter into reaction, and accordingly represent molecular weights. Although Gerhardt's endeavours in the determination of equivalents tended in the direction of employing comparable quantities only, this was first stated and elevated to a principle by Laurent. According to the latter, it is necessary to start from a " terme de comparaison " and to refer the formulae of all compounds to it. Since he is quite clear as to the fact that the quantities contained in equal volumes do not always pro- duce the same chemical effect, he considers the question whether he will compare substances in the gaseous state according to the space which they occupy, or whether he will compare their equivalents. He rejects the latter comparison, on account of the difficulty associated with the determination of equivalents of substances which are not analogous, and decides in favour of the first comparison ; that is to say, he 49 Ann. Chim. [3] 7, 129. 50 Ibid. [3] 7, 140. 51 Ibid. [3] 18, 266. IQ2 HISTORY OF CHEMISTRY. [LECTURE X. chooses the formulae (and therefore the molecules) of sub- stances so that they represent two volumes (H= i = i vol.) in the state of vapour. In doing so he is obliged, however, to admit certain exceptions to which he draws attention. Thus it was known from Bineau's investigations52 that the formulae NH4C1, for sal ammoniac, and SO4H2, for sulphuric acid,53 correspond to four volumes, but, in spite of this, these quan- tities are regarded by Laurent as representing their molecular weights. There were definite grounds, in these instances, which appeared to make this assumption necessary. The isomorphism of sal ammoniac with potassium chloride excluded the formula NjHgClj; the dibasic character of sulphuric acid, which Laurent looked upon as proved, demanded a molecular weight at vari- ance with Avogadro's hypothesis. Even if this hypothesis, therefore, was regarded as the chief standard in the fixing of formulae, still the results obtained were subject to modification by chemical reactions, and by physical properties such as specific heat, specific volume, crystalline form, etc. Further, the law of the even number of atoms, which had been outlined for special cases by Gerhardt54 in 1843, played an important part in these determinations. Laurent now states this law, to the effect that in all compounds the sum of the atoms of hydro- gen, chlorine, bromine, nitrogen, etc., must always be an even number. The law becomes of increased significance from the fact that Laurent applies it in order to prove that the molecules of these elements, which he calls dyads,55 consist of two atoms. Gerhardt's ideas were greatly elucidated by Laurent, who made them more generally accessible and comprehensible by laying greater weight upon the terms he employed, and by defining these terms precisely. As a result of this, an important advance was effected, because the separation of atom, molecule, and equivalent was now really accomplished ; and hence it again became possible to employ Avogadro's hypothesis (thirty-five years after its promulgation) as the 52 Ann. Chim. [2] 68, 416. 53 Compare ibid. [3] 18, 289. 54 Ibid. [3] 7, 129. 55 Ibid. [3] 18, 266; compare also Laurent, Methode de Chimie. 77 ; E. 62. LECTURE X.] HISTORY OF CHEMISTRY. 193 basis of a system. With Laurent, the molecule is the smallest quantity of a substance which is required in order to give rise to a compound, and which, in the form of vapour, always (or at least with few exceptions) occupies double the volume of an atom of hydrogen. The atom is the smallest quantity of an element which occurs in compounds, whilst the equivalents represent quantities of analogous substances which have the same value in reactions.56 I shall try to give an idea of the importance to chemistry of these definitions, by stating some of the conclusions that were drawn from them in view of the experimental work of the period. The consistent application of the idea of equivalents necessitated the assumption by Laurent and Gerhardt of several equivalents for many of the metals.57 "The idea of the equivalent includes the notion of an identity of function ; we are aware that one and the same element can play the part of two or of several others, whence it must occur that different weights also correspond to these different functions. On the other hand we find different weights of the same metal, such as iron, copper, mercury, etc., replacing the hydrogen of acids, and, in doing so, forming salts which contain the same metal but possess different properties. These metals have, consequently, different equivalents." This idea was not new,58 but as really equivalent formulae had never been employed, it had not up to this time had any further consequences. Now, when Laurent and Gerhardt in- troduce this mode of writing formulae, it acquires a certain value. Thus, for example, these reformers of chemistry seek to find analogies where they have hitherto remained concealed; the formulae of the sesquioxides can be assumed to be similar to those of the normal bases, and a uniformity can thus be introduced into the way of regarding salts which has not been possible hitherto. It is known that the neutral sulphate of the protoxide of iron (ferrous sulphate) contains, for the same 56 Ann. Chim. [3] 18, 296 ; also Comptes rendus des travaux chimiques par Laurent et Gerhardt 1849, 257. 57 Ibid. 1849, l etc. 68 Compare p. 103. N 1Q4 HISTORY OF CHEMISTRY. [LECTURE X. quantity of sulphur, one and a half times as much iron as the neutral sulphate of the peroxide of iron (ferric sulphate). Thus we may express it that 28 parts of iron in the ferrous salt can take the place of one part of hydrogen, while the latter can also have its place taken by i8| parts of iron, in the ferric salt; both quantities are, therefore, equivalent to one part of hydrogen. If we distinguish, as Laurent and Gerhardt did, by the terms ferrosum (Fe = 28) and ferricum (fe = §.28) the equivalents of iron in the ferrous and in the ferric salts respectively, then the formulae of ferrous sulphate (Fe2) SO4 and of ferric sulphate (fe2) SO4 become comparable with each other. A similar thing holds for other metals such as copper, mercury, tin, etc. ; in the salts corresponding to the lower and to the higher oxides of these metals, different equivalents, of which the one is double the other,59 must be assumed in each case. Complete analogy is attained in the mode of writing salts when equivalent formulae are employed for acids also. We then have : — Ferrous sulphate Cupric chloride Mercurous chloride S04(Fe2) . Cle(Cu2) Cl2(hg2) In this mode of writing formulae, the differences between monobasic and polybasic acids disappear ; and it certainly is an advantage of the molecular formulae that they permit these highly important peculiarities to become prominent. Laurent and Gerhardt recognised this very fully, and it was the latter, especially, who endeavoured, and with much success, to bring about the separation of these classes of substances by more definitely advancing new characteristics.60 The formation of double salts with non-isomorphous bases did not appear to Gerhardt sufficient for fixing the basicity of an acid ; he points out that dibasic (and polybasic) acids can 59 Laurent's views with respect to the mode in which the existence of several equivalents for the same element may be explained are given in his Methode de Chimie, 127 ; E. 103. 60 Laurent and Gerhardt, Comptes rendus mensuels des trauvaux chimiques 1851, 129 ; J. pr. Chem. S3, 460. LECTURE X.] HISTORY OF CHEMISTRY. 195 form two (and more) ethers, of which one (or more) is acid, and one is neutral. The molecule of the latter, if it is assumed to correspond to two volumes, contains one alcohol residue in the case of monobasic acids, and two (or several) such residues in the case of dibasic (or polybasic) acids. Further, the amid- compounds, and also the anilid-compounds (discovered a short time previously61) furnish additional evidence. Thus whilst the monobasic acids produce only one amide, one nitrile, and one anilide, the acid ammonium salts of dibasic acids, by the loss of water, give rise to the formation, besides, of an amid-acid and of an imide ; and they alone can yield anilid-acids. Laurent had already drawn attention, a few years earlier, to another difference between these substances.62 According to him, only the formulae of the dibasic and polybasic acids permit the assumption of their containing water, whereas in one molecule of a monobasic acid the constituents of only half a molecule of water require to be present, for which reason, the latter are not able to form anhydrides. Thus nitric acid is :— HNO3 = (HOj) + NO2i ; while sulphuric acid is :— H2SO4 = H2O + SO3. With Laurent, hypochlorous anhydride, which was already known at that time, is C1HO in which one atom of hydrogen is replaced by one atom of chlorine.63 Laurent's views as to the molecules of the elements were also very important. It was a consequence of Avogadro's hypothesis, with which Laurent identifies himself,64 that the molecules of certain elementary substances should be regarded as composed of at least two atoms. Laurent tries to support this view by means of chemical reasons. According to him the so-called "dyads," such as hydrogen, chlorine, bromine, 61 Compare p. 185. 62 Ann. Chim. [3] 18, 266. 63 These views were looked upon as refuted by Gerhardt's discovery of acetic anhydride etc., but strictly speaking, they were not. It was, in fact, an extension of the word anhydride to apply it to the substances discovered by Gerhardt. The relation between C2H4O2 and C4H6O3 on the one hand is not quite the same as that between C4H6O4 and C4H4O3 on the other. w He does not seem to be aware, however, that it was first announced by Avogadro. 196 HISTORY OF CHEMISTRY. [LECTURE X. iodine, nitrogen, phosphorus, arsenic, and antimony, always occur in even numbers only ; and this rule, if it is to hold for the molecules of the elements as well, renders the existence of single atoms in the free state impossible. Besides this, Laurent brings forward the well-known effects of the nascent state, and explains them by the assumption that at the moment of the separation of the elements from their compounds the single atoms are isolated, and therefore combine far more easily with others than in those cases where it is a question of molecules or of atomic groups which must first be decomposed before reaction can take place. Laurent and Gerhardt, with their far-reaching reforms, met with almost no immediate recognition ; on the contrary, it seems as if the conception of the equivalent, in its first uncer- tain form, had now found more adherents than formerly, and as if Gay-Lussac's law of volumes appeared to chemists to be less adapted than ever to constitute the basis of a system. For this reason there was not, in general, the slightest ten- dency exhibited to assume, with Laurent, the divisibility of the molecules of the elements. It is no doubt true that urgent grounds, 'and especially, chemical grounds, were still wanting. It was a very happy idea that Laurent and Gerhardt had hit upon when they stated that the formulae of substances must represent comparable quantities ; but the standard was still wanting. The spaces occupied by the gases were known only in relatively few cases, and even amongst these there were some cases where the molecular weights deduced were un- serviceable because they were at variance, or at least they appeared to be at variance, with the chemical properties. A series of facts which confirmed these ideas, and eventually pro- cured for them general recognition, was still wanting. We are indebted for our knowledge of them to Williamson, who showed how to find the molecular weight by chemical methods, and thereby rendered a service to our science which cannot be too highly estimated. Even although he did not instigate the reform of chemistry, still it was his investigations which first made its accomplishment necessary and possible. LECTURE XL REASONS FOR THE ASSUMPTION OF THE DIVISIBILITY OF ELEMENTARY MOLECULES— FIXING OF THE MOLECULAR WEIGHTS, BY WILLIAM- SON, BY MEANS OF CHEMICAL REACTIONS — THEORY OF THE FORMATION OF ETHER — FUSION OF THE RADICAL THEORY WITH DUMAS' TYPES — SUBSTITUTED AMMONIAS — POLYATOMIC RADICALS — GERHARDT'S THEORY OF TYPES AND SYSTEM OF CLASSIFICATION. IT has often been stated that chemistry should advance by development from within itself alone, and that the influence of the other sciences is injurious if it is exerted in any other direc- tion than that in which chemical facts appear to lead us. I understand perfectly well that in chemistry a theory is not chosen which is in contradiction to certain facts, in order to attain a more complete agreement with physical laws, and I appreciate this from the didactic point of view in particular ; but I consider it just as proper and essential in our science to modify our views so as to produce harmony with recognised natural laws, and also with theories and hypotheses, as soon as the facts permit of our doing so. It therefore appears to me appropriate, now when in our historical development we have arrived at the middle of a far-reaching reform}/to advance some of the not exclusively chemical reasons which told in favour of the system of Laurent and Gerhardt. In doing this, I confine myself to facts which appear to support the divisibility of the molecules of the elements, for the special reason that this hypothesis, even after it had been repeatedly announced, still met with no approval. Amongst the results, of extreme value in chemistry, which Favre and Silbermann announced in 1846 in their research on heats of combustion,1 there is a very remarkable fact which 1 Comptes Rendus. 23, 200, 198 HISTORY OF CHEMISTRY. [LECTURE XI. deserves to be stated here. They observed that on burning carbon in oxygen, less heat is produced than when nitrous oxide is employed. They considered that this striking fact could only be explained on the hypothesis that in both cases, besides the formation of carbonic anhydride, a decomposition took place ; that is to say, a change occurred involving the separa- tion of atoms which had previously been combined. Accord- ingly they thought that even in the free particles of oxygen gas, several (two) atoms must be assumed, and that the quantity of heat required for their decomposition must be greater than that required for the separation of the oxygen from the nitrogen in nitrous oxide. By the application of a hypothesis relating to chemical com- bination and decomposition, Brodie 2 arrives at the divisibility of molecules of hydrogen and of oxygen. It appears to him that the contrast which was drawn, in accordance with the views prevailing at that time, between the formation of compounds and the separation of elements, has no natural foundation. In his view, every combination is merely the consequence of a decomposition, and this can only be occasioned by new com- binations. He tries to .prove the accuracy of this view by means of various examples, and, in doing so, introduces certain signs and expressions designed to give an idea of the contrast (more generally, of the relation) between the atoms entering into combination. According to Brodie, there exists between these a relation (polarity) of such a kind that the one is desig nated as positive or negative as contrasted with the other. This relation, which Brodie also calls chemical difference, depends upon the peculiarities of all those particles with which the atom is for the time being combined. To permit of a better comprehension of this, I give here some of the examples adduced by Brodie. Silver does not unite directly with oxygen, whereas silver chloride is decomposed by boiling with potash, silver oxide being formed. According to Brodie, this is due to the fact that it is only by their combina- 2 Phil. Trans. 1850, 759. LECTURE XI.] HISTORY OF CHEMISTRY. 1 99 tion with chlorine and potassium respectively that the silver and oxygen acquire the polarity necessary for their combination. He writes : — According to Faraday,4 perfectly dry calcium carbonate is not decomposed even at the highest temperatures, whereas in the presence of water the decomposition begins at once ; and similarly, according to Millon,5 sulphuric anhydride can be distilled over carbonate of potash, the formation of a salt only taking place on the addition of water. Brodie writes : — KG + HSO4 = HO + KSO4. In this case, especially, it can be distinctly recognised why Brodie considers that combination is always accompanied by decomposition, whilst the first example seems to justify the converse statement. But the existence of free elementary atoms is incompatible with this, so that Brodie is at pains to show that these always appear in pairs, and then unite with one another. The most striking of the examples which he adduces is that of the evolution of hydrogen which takes place when copper hydride, discovered by Wurtz,6 is treated with hydro- chloric acid : — Cu2H + HC1 = C+u,Cl + HH. That a similar evolution is not observed on treating the metal with the acid is looked upon as owing to the fact that the same species of polarity is always associated with the hydrogen in hydrochloric acid, whilst the affinity of copper for chlorine is not sufficient to decompose hydrochloric acid.7 3 H = i, O = 8, K = 39, Ag=io8, C = 6, etc. 4 The statement is now regarded as erroneous ; compare Gay-Lussac, Ann. Chim. [2] 63, 219. 5 I have been unable to find this fact, quoted by Brodie, in Millon's papers. 6 Annalen. 52, 256. 7 Compare Wurtz, Lecons de philosophic chimique, 64. 200 HISTORY OF CHEMISTRY. [LECTURE XI. The formation of nitrogen by heating ammonium nitrite is also explained in the same way : — NO4H4N = 4HO + NN. This mode of regarding the matter is very specially suitable for explaining the reductions by means of hydrogen peroxide, which were at that time partially known and had been chiefly studied by Brodie himself.8 He regards the formation of oxygen as a consequence of the different polarities which this element possesses in the two oxides. We have, for example : — HOO + Ag, AgjO = HO 4- OO + Ag. The reduction of potassium permanganate and of potassium bichromate is supposed to proceed in a similar manner. Two atoms are always set at liberty simultaneously, and these, in consequence of their chemical difference, then unite with each other.9 Further, the discovery of ozone by Schonbein ; 10 the recognition of its nature as an isomeric modification of oxygen ; n and in particular the proof that it is condensed oxygen (a fact which was first stated by Andrews and Tait 12 on the strength of some highly interesting experiments, but was especially demonstrated by Soret 13) only find an explanation in the hypothesis of the divisibility of the elementary molecules. If ozone, as appears from Soret's experiments, possesses a relative density, one and a half times as great as that of oxygen, then the smallest particles of this latter gas must contain at least two atoms, whilst those of ozone consist of three atoms. But if this assumption is admitted in the case of oxygen, it cannot easily 8 Phil. Trans. 1850, 759. 9 Compare also the explanation which Wurtz gives of the fact that the combination of nitrogen and oxygen takes place much more easily in presence of hydrogen (Wurtz, Le9ons. 65). 10 Pogg. Ann. 50, 616 ; 59, 240 ; 63, 520 ; 65, 69, 161, 190, etc. ; compare "Ueber das Ozon," Basel 1844. n Archives des sciences phys. et natur., Geneve 12, 315 ; 17, 61 ; 18, 153. 12 Annalen. 104, 128 ; 112, 185 ; Ann. Chim. [3] 52, 333, and 62, 101. 1S Annalen. 138, 45 ; Supplementband LECTURE XL] HISTORY OF CHEMISTRY. 2OI be avoided in the cases of the other elements. The different vapour densities which were found in the case of sulphur 14 can only be explained by the assumption that, at low temperatures, the molecule consists of three times as many atoms as at a very high temperature. It is certainly not without interest, that, in 1857, Clausius was led, by the mechanical theory of heat, to the divisibility of the physical molecule.15 Since, according to this theory, the kinetic energy of translatory motion in equal volumes of two gases at the same pressure, is proportional to the absolute tem- perature, Clausius concluded that the kinetic energy of trans- latory motion of the single molecules of all gases at the same temperature is the same. This assumes the fulfilment of Avo- gadro's hypothesis. The number of facts drawn from different branches of natural science which tell in favour of this hypothesis might be multi- plied still further : but I confine myself to the statement of those that I have mentioned, and pass on to chemical arguments which, after all, were the only ones that eventually led to the re- cognition of the hypothesis. Amongst these, the experiments which were now carried out and led to the conception of the chemical molecule, distinctly take the first place. I will not assert that this conception was not already extant ; but it now appeared in a much more definite form. This latter assertion will certainly be justified when I collect here the facts and hypotheses which existed and exercised an influence upon the determination of molecular weights by chemical methods, prior to Williamson. The atomic theory gave the first clue to these magnitudes. The formula of every compound required to be expressible by means of multiples, by whole numbers, of the atomic weights ; but this had no obligatory consequences so long as the atomic weights had not been determined with certainty, because, in case of necessity, even the atomic weight of a constituent could 14 See Dumas : Ann. Chim. [2] 50, 170, and Deville and Troost, Comptes Rendus. 49, 239. 15 Pogg. Ann. 100, 353. 202 HISTORY OF CHEMISTRY. [LECTURE XI. be altered. It was unquestionable, however, that even with only a moderately consistent application of atomic considerations, there was a certain connection between the different formulae. A relation of this kind was attained, especially by Laurent and Gerhardt, in the case of organic compounds in particular. It appears to me to follow from the nucleus theory that Laurent assumed the number of carbon atoms in the radical to remain unchanged until carbon was separated in some form or other.1'1 It was Gerhardt who first clearly stated this rule 17 (which, how- ever, is not always accurate, as, for example, in the formation of polymeric substances), and it will readily be conceded that by means of it, the molecular weights of many series of com- pounds were fixed from a knowledge of these magnitudes in the cases of a few substances. The so-called law of the even number of atoms gave a further guide to the smallest formula, and, in consequence of it, Laurent and Gerhardt found frequent occasion to alter the formulae previously adopted. The conception of the polybasic acids exercised a very im- portant influence in the fixing of formulae, since even Liebig for example (who first definitely grasped this conception) was induced to double the formula of tartaric acid,18 in order to make it conform with the chemical character of this substance. In the case of one special class of substances — the acids — the recognition of a criterion of their polybasicity, for which we are indebted to Liebig, Laurent, and Gerhardt, was just as far- reaching as the later experiments of Williamson in the cases of other groups of compounds. The phenomena of substitution, as may readily be under- stood, contributed something towards rendering the molecules of different compounds comparable with one another. The formula of a substance frequently required to be multiplied by two or by three so that it might not be necessary to assume fractions of atoms in the products arising from the substance by the action of chlorine, etc. It only became necessary to attend to these considerations in consequence of the unitary 16 Compare p. 145. 17 J. pr. Chem. 27, 439. 18 Annalen, 26, 154. LECTURE XI.] HISTORY OF CHEMISTRY. 203 notions introduced by Dumas at the same time, and of the rule of Gerhardt mentioned above. This does not apply to the opponents of these views, as I shall prove by means of an example. Kolbe and Frankland prepared methyl (ethan), in 1848, by treating ethyl cyanide with potassium, and they assigned to it the formula C2Hj [C = 6].19 They subjected this substance to the action of chlorine in order to convert it eventually into methyl chloride. Instead of the latter they obtained a compound having the same composition as ethyl chloride, but which instead of becoming liquid at 12°, remained gaseous at - 18°. This they regarded as isomeric with ethyl TT chloride, and they formulated it C2H3 . C2Q, that is, as a coupled compound of methyl with another atom of methyl in which one atom of hydrogen is replaced by chlorine. With Kolbe and Frankland, therefore, the existence of the first sub- stitution product C4H5C1 was no reason for assigning the for- mula C4H6 to the original hydrocarbon. Laurent was of a different opinion. Prior to the isolation of the alcohol radicals, he had proposed for them, in case they should be discovered, the formulae now adopted.20 Afterwards, when Kolbe had discovered a general method for the preparation of the alcohol radicals by the electrolysis of salts of the fatty acids,21 Laurent and Ger- hardt return to this view in a detailed manner and designate them as homologues of marsh gas.22 A. W. Hofmann allies himself with them in so far that he leaves the possibility of isomerism between the alcohol radicals and the homologues of marsh gas an open question.23 It is otherwise with Frankland, who defends the formuke corresponding to two volumes for the alcohol radicals, as opposed to those corresponding to four volumes, and formulates methyl, C2H3, and ethyl hydride, C4H6, afterwards just as he had done before.24 The conception of the chemical molecule was more readily appreciable by Laurent, Gerhardt, and Hofmann. They en- 19 Annalen. 65, 279. 20 Ann. Chim. [3] 18, 283. 21 Annalen. 69, 257. 22 Comptes rendus mensuels des travaux chiniiques. 1850. 2 Annalen. 77, 161. 24 Ibid. 77, 221. 204 HISTORY OF CHEMISTRY. [LECTURE XI. deavoured to make their ideas in this connection plain to their opponents, but still they do not appear to have succeeded in convincing them. Telling experiments were still wanting. Gerhardt required hundreds of examples in order to show that H4O2, and not H2O, corresponds to the formulas N2H6 and H2C12. Laurent's proof with respect to the doubling of the molecular weights of hydrogen, chlorine, etc., is also clumsy. In spite of this, it cannot be denied that these chemists at that time possessed accurate fundamental ideas, and I do not doubt that they also would have been able to deduce the same con- clusions, productive as they were for our science, from the facts which were now discovered and so excellently turned to account by Williamson. By the action of potassium ethylate upon ethyl iodide, Williamson had hoped to be able to effect the synthesis of an alcohol ; 25 ethyl was expected to take the place of the potassium, with the formation of an ethylated ethyl alcohol — an expecta- tion quite in conformity with the views of the period. A short time previously, Wurtz had discovered ethylamine,26 which he regarded as a substituted ammonia — a view which was con- firmed by the interesting mode of formation of this and many analogous substances, discovered by Hofmann.27 Frankland had already tried by means of zinc ethyl, a substance discovered by himself,28 to introduce alcohol radicals into organic sub- stances.29 But Williamson's experiment gave unexpected results : instead of an alcohol, he obtained ether. He under- stands, however, how to adapt his ideas, which had been turned in an altogether different direction, to his results, and he at once recognises the full importance of his experiment. He explains the formation of ether under the conditions which he had observed, and then the formation of ether in general ; and he proves the accuracy of his view by means of a series of brilliant experiments. Different views were at that time held as to the formulae of 25 Annalen. 77, 37. 26 Comptes Rendus. 28, 223. 27 Annalen. 66, 129, etc. 28 Ibid. 71, 213 ; 85, 329. » Ibid. 85, 354. LECTURE XL] HISTORY OF CHEMISTRY. 205 alcohol and of ether. In conformity with Liebig's ethyl theory, alcohol was pretty generally written C4H12O.2, and ether C4H10O [C— 12, O=i6]. Now, with the halving of the atomic weights, the formulae had, in many cases, been halved : alcohol became C4H6O.2, and ether C4H5O [C = 6, O = 8], whereas Gerhardt assigned to these substances the formulae C2H6O and C4H10O [C=i2, O=i6], Further, Laurent had already drawn atten- tion to the fact, in i846,30 that the formulae of alcohol and ether, as well as those of potassium oxide and hydroxide, are derivable from that of water.31 He wrote : — HHO EtHO EtEtO KHO KKO Water. Alcohol. Ether. Potassium Potassium Hydroxide. Oxide. Williamson perceived that the last view alone was in agree- ment with his experiment. He formulated as follows the equation which represented the reaction he had discovered : — H5I = ?«0 + KI [C = 1 2]. In order to meet the opposing view, in accordance with which the equation ought to be written : — C4H5O.KO + C4H5I - 2(C4H60) 4- KI [C = 6] since the assumption was that potassium alcoholate is a com- pound of potassium oxide with ether, and that the latter separated during the decomposition, whilst a second " atom " of the same substance was simultaneously produced from the ethyl iodide, Williamson carried out the reaction with methyl iodide. He expected to obtain methyl ethyl ether, whereas, in accordance with the view just referred to, a mixture of methyl ether and ethyl ether ought to be produced. The experiment was, therefore, decisive and justified Williamson's hypothesis. Both by the action of methyl iodide upon potassium ethylate 30 Ann. Chim. [3] 18, 266. 31 Griffin claims priority of the view n accordance with which the alkalies do not contain any water. See Griffin, Radical Theory, 9. 206 HISTORY OF CHEMISTRY. [LECTURE XI. and by that of ethyl iodide upon potassium methylate the so- called mixed methyl ethyl ether was produced :— 50 + 1C H3 = >250 + IK K ^ -ti These experiments proved to Williamson that ether is produced from alcohol by the replacement of an atom of hydrogen by ethyl, and, therefore, that it contains more carbon in its molecule than alcohol does. The formation of the mixed ethers furnished him with a reason for excluding every other view. It now became a question to explain the formation of ether under known circumstances, especially by the treatment of alcohol with sulphuric acid ; and the solution of this problem was also furnished by Williamson, after chemists had been engaged upon it for decades. At first the process of etherifica- tion had been explained by the dehydrating action of sulphuric acid.32 This view agreed very well with Dumas' etherin theory. Hennell, although an adherent of the etherin theory, considered this view irreconcilable-with the formation of sulphovinic acid observed by him ; 33 and it was also in contradiction with the fact that water distils over at the same time as the ether. It was Liebig, especially, who established by means of numerous experiments a new theory of etherification.34 He ascertained that the formation of ethyl sulphuric acid precedes that of ether, and, according to him, the sulphuric acid does not with- draw water from the alcohol, but ether, which latter unites with the sulphuric acid. Indeed ethyl sulphuric acid was at that time looked upon as a compound of these two substances, that is, as an acid salt of ethyl oxide. Thus ethyl sulphuric acid = C4H10O + 2SO3 + H,O [C=i2, 0=16,8 = 32]. 32 Compare especially Fourcroy and Vauquelin, Scherer's Journal 6, 436 ; also Gay-Lussac, Ann. Chim. 95/31 1, and [2]2, 98. 33 Phil. Trans. 1828, 369. a4 Annalen. 9, 31 ; 13, 27 ; 23, 31. LECTURE XL] HISTORY OF CHEMISTRY. 207 Ethyl sulphuric acid, according to the experiments of Liebig, breaks up between 127° and 140° into ether and sulphuric acid. The remarkable phenomenon that a substance is produced and undergoes decomposition in the same operation, was explained by Liebig on the assumption that the formation only occurred at those places where the alcohol dropped in, and where, there- fore, the temperature was lowered to the boiling point of this liquid. The process of etherification therefore consisted, accord- ing to Liebig, in the combination of the sulphuric acid with the ether of the alcohol, and the decomposition into its constituents of the compound so formed, at parts of the liquid where the temperature was higher. Ether then distils over, and, along with it, the water which was separated on the formation of the ethyl sulphuric acid. In opposition to this theory Berzelius 35 advanced another view, which was supported and developed by Mitscherlich in particular.36 According to this view, the sulphuric acid acts by contact, taking no part in the reaction, and simply decom- posing the alcohol into ether and water by catalytic action. This was a mode of expressing the facts by means of specially chosen words, but it can scarcely be called an explanation. Liebig's hypothesis really embraced an explanation which was pretty generally accepted. It was only called in question after Graham had shown, in i85o,3T that alcohol and ethyl sulphuric acid are both necessary for the formation of ether ; and that the latter of the two does not yield ether even when heated alone to 143°, but decomposes, in presence of water, into alcohol and sulphuric acid. Williamson clearly understands how to apply these facts. The formation of ether is explained by the following equation : — C2H5c,n , Q>H5 ^ _ H^x-v C2H5~ H oU4 + H -HbU4 + C2H5U' sufpSc Alcohol. SuSrk Ether, acid. 85 Berzelius, Jahresbericht 1836, 241. 36 Pogg. Ann. 53, 95 ; 55, 209. 37 Journ. Chem. Soc. 3, 24 ; Annalen. 75> IQ8 ; compare Journ. de pharm. [3] 18, 30. 208 HISTORY OF CHEMISTRY. [LECTURE XI. whilst the formation of the ethyl sulphuric acid is represented by this one : — 5 o = o + 'so* If the latter equation is read from right to left, it explains Graham's experiment of the decomposition of ethyl sulphuric acid into alcohol and sulphuric acid. As soon as the doubled formula for ether was admitted, it was comprehensible why ether is not produced by heating ethyl sulphuric acid. Williamson, however, not contented with having shown the accuracy of his opinions by their being in harmony with the known facts, devises new experiments by which he can test them.38 The method he adopts is the same as previously. He chooses the two substances which act upon one another from groups containing different numbers of carbon atoms. He now causes ethyl sulphuric acid and amyl alcohol to interact, and this reaction gives rise to the expected ethyl amyl ether : — c|f 5so4 + C5^no = ^so4 + CH"O. He studies, besides, the action of sulphuric acid upon mix- tures of ethyl and amyl alcohols, and is able to show the formation, in this case, of three ethers — ethyl ether, amyl ether, and ethyl amyl ether. He finds " in these reactions the best evidence of the nature of the action of sulphuric acid in form- ing common ether, or in accelerating the formation of the so- called compound ethers ; for acetic ether is formed from acetic acid, just as ethylic ether from alcohol, by the replacement of hydrogen by ethyle. And if the circumstance of containing hydrogen, which is replaceable by other metals or radicals, be the definition of an acid, we must consider alcohol as acting the part of an acid in these reactions." A further consequence of Williamson's experiments was the fixing of the molecular weight of acetic acid. According to Williamson, this acid is formed from alcohol by the replace- ment of two hydrogen atoms of the ethyl group by an atom of 38 Journ. Chem. Soc. 4, 229 ; Annalen. 8x, 73 * LECTURE XT.] HISTORY OF CHEMISTRY. 2OQ oxygen, the radical C2H5, ethyl, being converted by oxidation into C2H3O, othyl. Acetic acid is now regarded as water in which one atom of hydrogen is replaced by othyl. The formula C6H12O3 fixed upon for acetone by Kane,39 did not appear to be in harmony with this ; it had, however, already been halved, and Williamson endeavoured to explain the for- mation of the substance during the distillation of the acetates, by means of the following equation : — C2H300 C2H300 _ CO . K00 C2H30 K K K f CH3 ' According to him, the potassium peroxide is replaced during the reaction by methyl derived from the othyl. Here, also, he checks his opinion by applying the method already mentioned. He distils mixtures of acetates and valerianates and obtains mixed ketones : — C2H800 C5H90Q COKOQ C.H.O K K K f CH3 ' Williamson concludes this paper, which did a great deal to advance chemistry, with the words : " The method here em- ployed of stating the rational constitution of bodies by com- parison with water, seems to me to be susceptible of great extension ; and I have no hesitation in saying that its introduc- tion will be of service in simplifying our ideas, by establishing a uniform standard of comparison by which bodies may be judged of." The " terme de comparaison " which Laurent had already sought for in vain, had now been found. Substances were to be conceived as formed from water in accordance with the pro- posal, which Williamson makes in 1851 in his famous paper upon salts,40 to regard this substance as the type for all com- pounds. His method of fixing molecular weights is a purely chemical one ; he regards compounds as formed from water by the replacement of one or of two atoms of hydrogen. Inas- much as he tests all his views by means of facts already known, as well as by means of new experiments, these views obtain, I ya Annalen. 22, 278. 40 Journ. Chem. Soc. 4, 350. O 210 HISTORY OF CHEMISTRY. [LECTURE XI. might almost say, absolute confirmation. His experiments, moreover, are not chosen at random, but are always sug- gested by the same kind of logical deductions. Williamson furnished thinking chemists with the means of determining molecular weights by chemical methods. The very general applicability of the method which he discovered, is demon- strated by the fine experiments of Gerhardt, who was led by it to the preparation of the mixed anhydrides,41 and of Wurtz, who employed it with similar success in fixing the formulae of the alcohol radicals.4- The mode, also, in which Friedel and Crafts43 determined the molecular weight of silicic ether depends upon a train of ideas that only became clear to chemists after the publication of Williamson's investigations. The fact must not remain unmentioned here that, only a few months after Williamson's publication, Chancel, on yth October 1850, published a paper44 in which, by means of similar experiments, he arrived at the same results as William- son. Chancel distils potassium ethyl sulphate with potassium ethylate, and with potassium methylate, and thus obtains ethyl ether and ethyl methyl ether. His mode of fixing the mole- cular weight of dibasic- acids is peculiar to himself, although it coincides, in principle, with Williamson's method. By dis- tilling potassium ethyl sulphate with potassium methyl car- bonate, and with potassium methyl oxalate, Chancel prepares ethyl methyl carbonate and ethyl methyl oxalate. The reac- tions are represented by the following equations : — co' (CH K + so< (cfnj K = co° (cn The experiments of Williamson and of Chancel were of the greatest importance in the development of the science, for our present views are founded upon the conception of the chemical molecule. Unfortunately it is not always possible to determine 41 Ann. Chim. [3] 37, 332 ; Annalen. 82, 127 ; 83, 112 ; 87, 57 and 149. 42 Ibid. 96, 364. ^ Ann. Chim. [4] 9, 5. 44 Coniptes Rendus. 31, 521. LECTURE XL] HISTORY OF ' CHEMISTRY. 211 this with the same exactness as in the cases considered above. It became manifest, however, that with few exceptions, to which we shall afterwards return, the chemical molecule agrees with the molecular weight deduced from the observed volume, on the basis of Avogadro's hypothesis. This led to the assumption of the general identity of the physical and the chemical molecule, whereby a new means, and an almost always sufficient one, is furnished for ascertaining the highly important molecular weight. But Williamson's experiments and opinions also exerted an influence in another direction ; that is, with respect to the views concerning the constitution of compounds. The way was now prepared for a fusion of the newer radical theory or theory of residues with Dumas' theory of types, from which Gerhardt's theory of types arose. In this development the labours of other chemists were, however, of at least as great importance, especially as these had already been partially carried out. Accordingly, we shall now direct our attention more particularly to the latter. In 1849, Wurtz, by treating cyanic ether, cyanuric ether, and the substituted ureas prepared from these by himself, with potash, obtained bases extremely like ammonia, which he compared with the latter inasmuch as he regarded them as ammonia in which an atom of hydrogen is replaced by a radical such as methyl, ethyl, amyl, etc.45 This way of regarding these substances involved an important step in advance, since it was the first successful attempt to introduce radicals into the types.46 The fact that Liebig, as early as 1839, expressed a similar view concerning these substances, which were then only hypothetical,47 satisfactorily proves the clear perception of this gifted scientist, but cannot detract from the merit of Wurtz. The views of Wurtz respecting the constitution of these 45 Comptes Rendus. 28, 224, 323 ; 29, 169, 186 ; Annalen. 71, 330. 46 Compare, however, Laurent, Ann. Chim. [3] 18, 266. 47 Hand- worterbuch der Chemie, by Liebig, Poggendorff, and Wohler, I, 698. 212 HISTORY OF CHEMISTRY. [LECTURE XI. artificial bases, received important support from Hofmann's method of preparing them.48 Hofmann succeeded, by treating the alkyl iodides with ammonia, in introducing the radicals into the latter, and his experiments possess all the greater importance from the fact that he also showed how to prepare secondary and tertiary compounds, as well as substances cor- responding to ammonium chloride and ammonium hydroxide. Thus : ^TT 4. IP TT — Tsl^2^5 ITT IN -n-3 T iv^o-Tj-5 — -LN TJ j m H £-2^5 N7?B +ICH3 =NC H3,HI v^rj-p, TJ- H C9 C"H3 N C9H, + IC5Hn = NC H3, HI QH5 (C2H5)2 Nf TT , Jf~< TT XTT' "PT ^ -LJ-Q T iv^qrir = IN V_x -tlo \^2"5/2 (t~* TJ \ N C H3 I + AgHO = Agl + N^2«'20. C H ^ W3 ^5*^11 f TJ Mi*1!! I shall not leave the fact unmentioned that Paul Thenard had discovered the organic phosphorus compounds in i845,49 but that these only received their correct explanation now.50 Of other investigations, carried out at the beginning of the fifties, which contributed to the establishment of the new theory of types, I mention the discovery of the acichlorides by Cahours ; 51 that of the anhydrides of monobasic acids by Gerhardt ; Williamson's researches on dibasic acids ; and, finally, the preparation of the acid amides of Gerhardt and Chiozza.52 48 Annalen. 66, 129; 67, 61, 129; 70, 129; 73, 180; 74, I, 33, 117; 75. 356; 78, 253; 79, ii. 49 Comptes Rendus. 21, 144; 25, 892. 50 Compare Frankland, Annalen. 71, 215. 51 Ibid. 70, 39. Liebig and Wohler had prepared benzoyl chloride and benzoyl-amide twenty years before ; see Annalen. 3, 249. 52 Comptes Rendus. 37, 86. LECTURE XI.] HISTORY OF CHEMISTRY. 213 Gerhardt had at once grasped the bearings of Williamson's investigations, and was only able to perceive in them confirma- tion of the views already upheld, at an earlier date, by himself and Laurent, but never stated with the same precision.53 He perceived that Williamson's reaction for the formation of ether might also be applied to the monobasic acids, and that the oxides or anhydrides of the latter should be obtained in this way.54 The experiment succeeded, and thus it was reserved for Gerhardt, who, as well as Laurent, had denied the existence of anhydrides of monobasic acids, to disprove this view by his own experiments. He had, it is true, previously only stated the impossibility of withdrawing a molecule of water from one molecule of acid, and this statement still held ; for he showed that two molecules of a monobasic acid are always concerned in the formation of the anhydride, and the proof was furnished by Williamson's method. By treatment of potassium acetate with acetyl chloride, Gerhardt obtained acetic anhydride : — CaH'£ ] O + C2H3OC1 - g^g j O + KC1, and by employing benzoyl chloride he obtained the intermediate anhydride of benzoic and of acetic acids : — Before turning to the researches carried out jointly by Gerhardt and Chiozza on the anhydrides and amides of dibasic acids, I must give an account of the conception which William- son introduced concerning these acids, whereby the stimulus was given that led to these researches. The extension which Williamson gave in 1851 (that is, a year after his first investi- gation on etherification) to the views already arrived at, was an extremely important one. Even if it may perhaps be said that in his previous publications he leaned towards the views of Laurent and Gerhardt, and merely confirmed these by means of new, although certainly most decisive experiments, he now appears in a perfectly independent and original manner. 53 See his claim : Annalen. 91, 198. 54 Ann. Chim. [3] 37, 332. 2l4 HISTORY OF CHEMISTRY. [LECTURE XI. Williamson shows how the existence of dibasic acids depends upon the presence of radicals with basicity greater than one.55 The formation of the substituted ammonias, in the reaction discovered by Wurtz, which he formulates : — 2 r» _i_ Qi"i /rr»\ 2 n _i_ ^2**| rr (H2)°2 + N (CO' = (CO)°2 + N H'J gives him occasion to express himself as follows : — " One atom of carbonic oxide is here equivalent to 2 atoms of hydrogen, and by replacing them, holds together the 2 atoms of hydrate in which they were contained, thus necessarily forming a bibasic compound, ^ -^ ' CX, carbonate of potash." JS.O By further assuming that carbonic oxide can double its atomic weight, without alteration of its basicity (or equivalence), he obtains in C2O2 the radical of oxalic acid, and is able to represent the formation of oxamide by means of the equation : — (C2H5)2 Q N H H _ /C2H5 o\ (CO)2 (C0)2 U'2 + ? 2 2 4 ~ V H U/ + N2H4' The conception of sulphuric acid as a dibasic hydrate of the radical SO2 is highly important, and the experiments which Williamson carries out in support of this view, are most inter- esting. Besides the known chloride SO2C12, which Regnault had prepared from sulphurous anhydride and chlorine,50 he succeeds in isolating also chlorosulphonic acid, by treating sulphuric acid with phosphorus pentachloride.57 Thus : H r] O u <=;n S02 + PC15 - bY? + POC13 + HC1. O g H By means of this experiment he disproves the view of Ger- hardt, in accordance with which the formation of the anhydride is always supposed to precede that of the chloride in the case 55 Journ. Chem. Soc. 4, 350. 56 Ann. Chim. [2] 69, 170; 71, 445. 57 Proc. Roy. Soc. 7, 1 1 ; Annalen. 92, 242. LECTURE XI.] HISTORY OF CHEMISTRY. 215 of dibasic acids.58 Thus Gerhardt, conjointly with Chiozza, had published, in June 1853 (and thus half a year before this last paper of Williamson's appeared), investigations on the derivatives of dibasic acids, and especially on anhydrides and chlorides, in which he thought it was shown, amongst other things, that the first action of phosphorus pentachloride con- sists in the removal of water, and that it is only in the second stage of the reaction that a substance containing chlorine is produced. Gerhardt and Chiozza arrived at this time, however, at very important results ; they regarded the dibasic anhydrides, for the first time, as water in which both the hydrogen atoms are replaced by a single radical, and they also showed how to prepare succinyl chloride and similar chlorides. In two sub- sequent papers 59 they deal with the investigation of the amides corresponding to the polybasic acids. They show that these are either derived from two molecules of ammonia which are held together by the replacement, by a dibasic radical, of one atom of hydrogen in each, or that they may be derived from one molecule of ammonia. The amid-acids correspond to the mixed type NH3 + H2O, which can only be produced by the polyatomic acid radical entering the molecule ; and the earlier statement of Gerhardt that only dibasic acids could give rise to the formation of amid-acids is thus explained. By these and similar experiments, but especially by adopt- ing the ingenious conclusions that Williamson had drawn from his investigations, Gerhardt is able to establish a complete classification of organic compounds according to a new principle,60 and he expounds this classification in the fourth volume of his excellent Hand-book. An important point in Gerhardt's system, consists in his showing the connection between substances of opposite char- acter by means of intermediate substances. Unlike the dualists, he does not contrast such substances as potash and sulphuric acid as absolutely opposite in character, 58 Comptes Rendus. 36, 1050. 59 Ibid. 37, 86 ; 38, 457. 60 Gerhardt, Traite de Chimie organique, 4, quatrieme partie. 2l6 HISTORY OF CHEMISTRY. [LECTURE XI. but he connects them by means of transition compounds, and thus obtains series in which he arranges substances. In arrang- ing these series he makes use of two generalisations, of which, however, one does not originate with himself. In 1842, Schiel had pointed out 61 that the alcohol radicals form a series whose separate members differ by n. CH2, and that the corre- sponding alcohols show a difference in boiling point of 18° for each CH2 as H. Kopp had already proved in the case of ethyl and methyl compounds.62 In 1843, Dumas showed63 that the fatty acids also possess, amongst themselves, the same differ- ence in composition. Gerhardt now employs this very striking regularity, which, as is well known, occurs amongst very many organic substances, and calls the compounds which differ by n. CH2, homologous. It had been found that such compounds possess great similarities to one another, and that their physical properties slowly and progressively change. This had especially appeared from Kopp's detailed and excellent investigations,64 Gerhardt establishes, further, the idea of isologous compounds : these substances are also chemically similar but their difference in composition is not n. CH2. Acetic and benzoic acids are well-known examples belonging to this class of substances. The homologous and isologous series constitute the one part of Gerhardt's classification ; the other part is represented by the heterologous series. All substances are referred to the latter which can be obtained from one another by means of simple reactions (by double exchange) ; these substances are allied in their mode of formation but they are chemically different. Gerhardt very appropriately regards this arrange- ment of the compounds as similar to a game of cards which is based upon the colour as well as upon the value of the separate cards. Just as in the latter every card which is wanting is characterised. by its vacant place as of a certain value and of a certain colour, so the chief properties, the formation, and the 61 Annalen. 43, 107. 62 Ibid. 41, 79. & Ibid. 45, 330. 64 Ibid. 41, 169 ; 50, 71 ; 55, 166 ; 64, 212 ; 92, I ; 94, 257 ; 95, 121, 307 ; 96, i, i53> 303 ; 98, 367 ; ioo, 19, etc. LECTURE XI.] HISTORY OF CHEMISTRY. 2iy decomposition of the terms which are wanting in the chemical classification can be stated beforehand. Gerhardt compares the members of one and the same heterologous series (representative, therefore, of the various homologous and isologous series) with four very minutely studied inorganic substances, as prototypes, viz., water, hydro- chloric acid, hydrogen, and ammonia — all compounds of hydrogen. A substance which was to be regarded as belonging to one of these types, was necessarily capable of being con- ceived as derived from it by the replacement of hydrogen atoms by radicals. Thus Gerhardt refers alcohols, ethers, acids, anhydrides, salts, aldehydes, ketones, etc., to the water type, and to this type there also belong the mercaptans, sul- phides, etc. The latter really correspond to the type of sulphuretted hydrogen, but this is merely a subdivision of the water type. Chlorides, bromides, iodides, and cyanides are referred to hydrochloric acid. Ammonia was the prototype of the amines, amides, imides, and nitriles, as well as of the corresponding phosphorus compounds. Finally, the hydro- carbons, the alcohol radicals and the radicals containing metals were referred to the hydrogen type, H2. The great step had, accordingly, now been taken ; radicals had been introduced into the mechanical types of Regnault and of Dumas. If we look back and inquire to whom we are chiefly indebted for this excellent extension of the earlier theory of types, the names of Laurent and of Wurtz especially deserve to be mentioned. As early as 1846, the former had referred alcohol and ether to water ; three years later, Wurtz discovered ethylamine which he regarded as a substituted ammonia. This view met with acceptance all the more rapidly that the simi- larity between the two substances is so startling. C5 I shall not omit to observe here again, that the conception of the radical was now adopted in the sense in which Gerhardt had defined 65 With respect to the share which Hunt took in the development of the theory of types, compare Silliman's Journal [2] 5, 265 ; 6, 173 ; 8, 92 ; 9, 65 ; also Phil. Mag. [4] 3, 392. His claim is printed in Comptes Rendus. 52, 247, and the reply of Wurtz in Repert. de Chimie pure. 3, 418. 2l8 HISTORY OF CHEMISTRY. [LECTURE XI. it in 1839. Radicals were residues of compounds, i.e., they were atomic groups which, in certain reactions, could be trans- ferred, undecomposed, from one substance to another ; they did not, however, on this account at all require to exist indepen- dently, and they were only intended to express the relation in which elements or atomic groups are replaced.613 The formulae which are thus obtained for compounds, do not indicate the arrangement of the atoms ; they are merely reaction formulae, which recall a series of analogies. It can thus be understood how Gerhardt could imagine several radi- cals and several rational formulae in the case of the same sub- stance. The determination of the true constitution of substances appeared to him a task not capable of accomplishment, since modes of formation and of decomposition can alone lead to a judgment respecting it, and the multiplicity of these does not permit of any conclusion as to the arrangement of the atoms. Thus, for example, barium sulphate is formed from sulphuric acid and baryta, from sulphurous acid and barium peroxide, and besides, from barium sulphide and oxygen. The constitution of the salt might, therefore, be represented symbolically by means of the three formulae :— Ba2O + SO3, Ba2O.2 + S(X, Ba.2S + O4 67 (0=i6,S-32, Ba = 68.5). By means of this single example Gerhardt considered himself able to prove that all endeavours directed towards the repre- sentation of the arrangement of the atoms by means of symbols, must lead to nothing. With Gerhardt, reactions are double decompositions ; and here the contrast is seen between his system and the dualistic system, in which all compounds are conceived as formed by additions. Gerhardt goes so far as even to assume a double decomposition, or as he calls it, a typical reaction, when two molecules unite to form a single one. Thus ethylene chloride, according to him, is produced from olefiant gas in consequence of the substituting action of chlorine. The chloride C2H3C1 is 66 Gerhardt, Traite de Chimie organique. 4, 569. 67 Compare p. 177. LECTURE XI.] HISTORY OF CHEMISTRY. 2IQ formed, and this remains united with the hydrochloric acid which is produced simultaneously.68 The general arrangement and the comprehensive character of Gerhardt's system leave nothing to be desired. Even although our views have been considerably changed and cleared up since that time, and although we are compelled, from our present standpoint, to look upon the types as insufficient, still Gerhardt's services to chemistry can never be questioned. Un- fortunately he was not long able to congratulate himself on the acceptance that his admirable Hand-book met with, as he died shortly after its completion. 68 Compare p. 149. LECTURE XII. MIXED TYPES — RELATION BETWEEN KOLBE'S VIEWS AND THE COPUL^E OF BERZELIUS — RADICALS CONTAINING METALS — CONJUGATED RADICALS— KOLBE AND FRANKLAND AND THE VIEWS REGARDING TYPES— POLYBASICITY AS AN EVIDENCE FOR THE ACCURACY OF THE NEW ATOMIC WEIGHTS — DISCOVERY OF THE POLYATOMIC ALCOHOLS AND AMMONIAS. I AGAIN desire to direct attention to the theory of types in the form in which it had been established by Gerhardt. The latter had divided organic substances into natural families, if I may so express myself, represented by the four types, water, hydro- chloric acid, ammonia, and hydrogen, which were also called by him types of double decomposition. In this connection it must be pointed out that Gerhardt assumes the existence of conjugated radicals, so as to be able to include substitution products also in the types, and " to connect with one another several systems of double decomposition of a substance."1 For this purpose he employed, in part at least, the same mode of regarding substances as Kolbe, an exposition of which I have still to give. It must now be pointed out that the conjugated compounds are no longer considered in the sense previously stated by Gerhardt ; and not only has the name of the " corps copules " been changed into " corps conjuges," but the signification of the thing itself has been altered. The law of basicity, already discussed in detail, no longer finds any application ; '2 mono- basic acids can now give rise to conjugated compounds by interacting with neutral substances ; and to this class of con- jugated compounds there are now reckoned all the substances produced by substitution (acids especially), and consequently 1 Gerhardt, Traite. 4, 604. 2 Compare p. 183. LECTURE XII.] HISTORY OF CHEMISTRY. 221 the substances obtained by the action of chlorine, bromine, iodine, nitric acid, sulphuric acid, etc., upon organic materials. Accordingly, the conjugated radicals were, as we should now say, substituted radicals, and they embraced, besides, the atomic groups containing a metal, such as cacodyl, etc. Whilst Gerhardt placed chloracetic acid C2C13^ I Q) picric acid C6H2(N02)8|0j sulphobenzoic acid C?H4(S04?1(X, etc. " ) *12J in this class of substances, other chemists, Mendius for example,3 only cared to call substances of the latter kind conjugated ; whereas others still, such as Limpricht and Uslar,4 would have wished to see almost all organic compounds placed in this category. A discussion took place in connection with this subject which ended with the introduction of the mixed types and the abandonment of the conjugated compounds. Gerhardt had already referred the amid-acids, which he places in his text-book amongst the "acides conjuges," to the type ammonia + water 5 in 1853. Returning to this idea, but, at the same time giving it an important extension, Kekule shows, in 1857, upon the assumption of mixed types, how a distinction between conjugated and other compounds becomes quite unnecessary.6 The possibility of this hypothesis rested upon the conception of the polybasic radicals introduced in 1851 by Williamson, and by means of this conception it be- came comprehensible how two molecules, previously separate, may be united into a single one. Williamson had explained the joining together of two molecules of water as depending upon the nature of the radical SO2, and in this way the condensed types arose. Kekule employs this view in establishing the mixed types. With respect to this he expresses himself as follows : — "A union of several molecules of the types can only occur when, by the entrance of a polyatomic radical in place 3 Annalen. 103, 39. 4 Ibid. 102. 239. 5 Compare p. 215. 6 Annalen. 104, 129. 222 HISTORY OF CHEMISTRY. [LECTURE xn. of two or three atoms of hydrogen, a cause is furnished for the holding of these molecules together." Since an unlimited number of heterogeneous molecules may unite in this manner, even the. most complicated compounds could be referred to types. There was thus no longer any necessity to have re- course to conjugated compounds, and Kekule further points this out : — " The so-called conjugated compounds are not com- posed in any manner different from other compounds; they can be referred in the same manner to types in which hydrogen is replaced by radicals ; they follow the same laws with respect to their formation and saturating capabilities as hold for all chemical compounds." With a view to facilitating a better comprehension of Kekule's ideas, I give below a few of the formulae proposed by him: — - w H r ^ 5 referred/H H HJ0 Benzene sulphonic acid. H HI H Qn -j referred /HJ SO.,1 to \H\ SO2|O CAH5 o /HJ referred to HJ H j O HJ Sulfobenzide. Nordhausen sulphuric acid H etc. sojo Isethionic acid. Carbyl-sulphate. Sulfobenzoic acid. H fH CTH40\0 referred IH\ H/O to (HJ° IH\O HJU The way in which Kekule employs the reaction with penta- chloride of phosphorus, in order to distinguish from each other the types H2 and H2O, is interesting, and I shall mention it here in passing. Kekule points out how the oxygen in water is replaced, by means of this reagent, by two atoms of chlorine, whereby the molecules corresponding to this type are broken LECTURE XII.] HISTORY OF CHEMISTRY. 223 up, whilst those deducible from the hydrogen type are pre- served :— c H i r H n V^ 47 1 J. r L /"\ V*» o -i- -1 r V^ A Ethyl sulphuric acid SOJ gives SO0CL H}° -id while benzene sulphonic acid SO2^Q gives SO9C1 H J ~HQ It was by means of the theory of mixed types (the last consequence of this mode of regarding substances) that Ger- hardt's system first attained that uniform character in which it dominated organic chemistry for several years. But after the idea of the types had been recognised,7 they themselves became unnecessary. The theory of types was only a formal con- ception, which lost its importance as soon as its real teaching had been grasped. It had been necessary, however, for the origination of the views as to atomicity then in process of development. Particularly active in this connection were Williamson, Wurtz, Odling, and especially Kekule — that is to say, the chemists who had already taken an important part in establishing the theory of types. Simultaneously, however, such important services were rendered, from an altogether different side (that is, from the opponents of Gerhardt and successors of Berzelius), both by means of theoretical specula- tions and of experimental investigation, that before we turn to the theory of atomicity and the views arising from it as to the mutual relations of the atoms, we shall look more closely at the labours of that school which had sprung up from the ruins of the system of Berzelius. In doing so, I may be permitted to go a long way back and state the facts which, in my opinion, led from the copulse of Berzelius to the important views of Kolbe. Magnus had shown, at the beginning of the fourth decade of the century,8 that the salts of ethyl sulphuric acid, dried 7 Ann. Chim. [3] 44, 304 ; compare also Hunt, loc, cit, 8 Pogg. Ann. 27 267. 224 HISTORY OF CHEMISTRY. [LECTURE XII. in vacuo over sulphuric acid, correspond to the old formula of Serullas,9 C4HS + 2SO3 + MO + H2O [C=i2, 8 = 32, O = i6]. Liebig confirmed this,10 and was thereby led to pronounce ethyl sulphuric acid isomeric with isethionic acid.11 He found very essential differences, however, in the behaviour of the two acids towards potassium hydroxide. Whilst the former acid was converted, on simply boiling with this reagent, into alcohol and potassium sulphate, the latter acid was only decomposed on fusion with it and gave rise to the formation of a sulphate and a sulphite. This reaction induced Liebig to assume the existence of dithionic acid in isethionic acid. Berzelius, who adopted Liebig's view, employed it in arranging into two classes, the substances produced by the action of sulphuric acid upon organic compounds.12 Kolbe, in 1844, tried to bring into harmony with the opinion of Berzelius,13 the ingenious views of Mitscherlich 14 in accordance with which (following the analogy of the ordinary acids) the sulpho-derivatives of the first class were regarded as compounds of sulphuric acid, and those of the second class as compounds of carbonic acid. He was at that time engaged upon an examination of.the substance discovered by Berzelius and Marcet 15 in acting with chlorine upon carbon bisulphide. He fixes its formula as CC12SO2 [C = 6, O = 8, S = i6] and calls it sulphite of perchloride of carbon. By treatment with potash he converts this substance into Chlorkohlenunterschwefel- sdure (trichlormethyl-sulphonic acid), which, in turn, is con- verted by means of the reactions of Melsens16 (that is, by the action of nascent hydrogen) into Chlorformylunterschwefelsaure (dichlormethyl-sulphonic acid), Chlorelaylunterschwefelsciure (chlormethyl-sulphonic acid), and Methylunterschwefelsdure (methyl-sulphonic acid). Kolbe regards these compounds 9 Ann. Chim. [2] 39, 153 ; 42, 222 ; Fogg. Ann. 15, 20. 10 Annalen. 13, 28. u Compare p. 131. 12 Annalen. 28, I. 13 Ibid. 54, 145. 14 Ibid. 9, 39; Pogg. Ann. 31, 283 ; compare also Mitscherlich, Lehrbuch, Third Edition, i, 107 and 586. 15 Gilb. Ann. 48, 161. 1G Compare p. 171. LECTURE XII.] HISTORY OF CHEMISTRY. 225 as hyposulphuric (dithionic) acid coupled with different radicals, and writes their formulae : — C2C13 + S2O5 + HO Chlorkohlenunterschwefelsaure. C2HC1., + S2O5 + HO Chlorformylunterschwefelsaure. C2H2C1 + S.,O5 + HO Chlorelaylunterschwefelsaure. C2H8 + S2O5 + HO Methylunterschwefelsaure. Kolbe succeeds in effecting the synthesis of trichloracetic acid in a similar manner, that is, by treating chloride of carbon with chlorine in sunlight, in presence of water. In this he finds a ground for the assumption by Berzelius of the presence of chloride of carbon in trichloracetic acid, and thereby secures an important footing for Berzelius' whole mode of regarding these compounds. At the same time the analogy of the sub- stance discovered by Dumas, with the compounds containing sulphur, prepared by Kolbe, is now furnished, since trichlor- acetic acid was written, after the style of Berzelius, C2C13 + C2O3 •4-HO. It was thus a conjugated oxalic acid, whilst the others were conjugated hyposulphuric acids. Kolbe admits, as Berzelius had done previously, the replace- ment of hydrogen by chlorine in the copula. That a substitu- tion of this kind should be possible without essential alteration of the properties, depended upon the assumption that the nature of the copula exercised only a subordinate influence upon the character of the compound. Kolbe, no doubt, per- ceives (what Berzelius never admitted) that he thereby adopts an essential point in the theory of substitution. It appears to me necessary to state distinctly that Kolbe, and also Frankland (who, at that time, agreed completely with Kolbe's views), adopted the conception of a radical in its earlier sense. They believe in the existence in compounds of certain atomic groups, and are, therefore, far from admitting, with Gerhardt, that different radicals may be assumed to be present in a substance. Both Kolbe and Frankland attack the problem of ascertaining the constitution of compounds, and, by doing so, they essentially distinguish themselves from p 226 HISTORY OF CHEMISTRY. [LECTURE XII. the adherents of the theory of types who, with the excep- tion of Williamson,17 write reaction or decomposition formulae only. With the assumption of distinct atomic groups in complex substances the idea of the possibility of their isolation was also combined, and thus we find Kolbe and Frankland, in 1848, engaged in experiments which have for their aim the separation of radicals ; ls in particular, it appeared to Kolbe extremely desirable to decompose acetic acid into methyl and oxalic acid, of which it was the conjugated compound. He succeeds in isolating one, at least, of the radicals lv) by aid of the electrical current. Under the influence of this agency, acetic acid splits up into methyl and carbonic acid. According to Kolbe, the reaction took place in such a manner that the conjugated groups first separated from one another, and that the oxalic acid was then converted into carbonic acid at the expense of the oxygen of the water ; and the simultaneous evolution of hydrogen appeared to confirm this view. The preparation of methyl cyanide by heating ammonium acetate with phosphoric anhydride,20 discovered a short time previously by Dumas, told in favour of the views of Kolbe and Frankland, and so did the conversion of the nitriles into the corresponding acids, which was carried out by the latter chemists themselves.-1 Upon his isolation of ethyl from ethyl iodide by means of zinc,'22 it appeared to Frankland that he had removed every doubt as to the accuracy of his and Kolbe's mode of regarding compounds. The ethyl theory was now to resume its old place, in the form stated by Liebig in 1835. According to Frank- land : " The isolation of four of the compound radicals belong- ing to the alcohol series, now excludes every doubt of their actual existence, and furnishes a complete and satisfactory 17 Journ. Chem. Soc. 4, 350. l8 Ibid. I, 60; Annalen. 65, 269. 10 Journ. Chem. Soc. 2, 157; Annalen. 69, 257. ao Comptes Rendus. 25, 383 and 473 ; Annalen. 64, 332. -l Annalen. 65, 288. " Journ. Chem. Soc. 2, 263; Annalen. 71, 171. LECTURE XII.] HISTORY OF CHEMISTRY. 227 proof of the correctness of the theory propounded by Kane, Berzelius, and Liebig fifteen years ago." 23 This series of investigations, carried out between 1844 and 1850, rehabilitated the theory of copulae. Even if it only appeared to be justified by the reactions in the case of a small class of substances, still it was justified for the most important compounds, and for those, in particular, which had influenced Berzelius in setting up his views. Experiment had shown that the assumption of methyl in acetic acid, of chloride of carbon in trichloracetic acid, of ethyl in alcohol, etc., had a real foundation, and it soon appeared to be clear that the way opened up by Kolbe and Frankland must further lead to many brilliant discoveries. While occupied with the isolation of ethyl from ethyl iodide, Frankland discovered zinc ethyl,-4 a substance which com- manded the greatest interest on account not only of its physical but also of its chemical properties. After the discovery of this compound, the efforts of no small number of chemists were directed towards making it available for synthetical purposes ; 25 and even although all the hopes which were based upon it were not realised, still there are few compounds which have been employed in so many ways in investigations in organic chem- istry. Intimately connected with the discovery of zinc ethyl is the preparation of the other organo-metallic compounds. We are indebted to Wohler.26 for the discovery of tellurium ethyl ; the antimony compounds were prepared by Lowig and Schweizer,27 and the tin compounds simultaneously by Frank- land 2S and by Lowig ; 2U mercury ethyl was prepared by Frank- land,30 and aluminium ethyl was prepared by Cahours31 but -3 Journ. Chem. Soc. 3, 46 ; Annalen. 74, 63. 24 Journ. Chem. Soc. 2, 297 ; Annalen. 71, 213. 25 Pebal and Freund, ibid. 118, I ; Wurtz, Comptes Rendus. 54, 387 ; Annalen. 123, 202 ; Rieth and Beilstein, ibid. 124, 242; 126, 241 ; Alexeyeff and Beilstein, Comptes Rendus. 58, 171 ; Butlerow, Zeitschrift fiir Chemie. 7, 385 and 702 ; Fried el and Lad en burg, Annalen. 142, 310; Lieben, ibid. 146, 180, etc. '-6 Ibid. 35, in ; 84, 69. 27 Ibid. 75, 315. -8 Phil. Trans. 1852, 418; Annalen. 85, 332. 29 Ibid. 84, 308. 30 Journ. Chem. Soc. 3, 324; Annalen. 77, 224. 31 Ibid. 114, 227 and 354. 228 HISTORY OF CHEMISTRY. [LECTURE XII. first studied by Buckton and Odling.32 Highly important was the discovery of potassium and of sodium ethyl which was made by Wanklyn,33 whilst Friedel and Crafts 34 showed how to obtain silicium ethyl, etc. I have intentionally referred to these compounds here because they exercised a distinct influence upon the further development of the theory of conjugated radicals. Kolbe was the first to explain correctly the nature of cacodyl ; he calls it methyl coupled with arsenic, As CH3 [C = 6] ;35 and even if we do not now employ the word "coupled," still we have in other respects retained this view with regard to the substance, and our conception of the relation of the metal to the radical has not become much clearer. Kolbe has a similar way of looking at other organic com- pounds ; all of them contain conjugated radicals, most of them with carbon as the copula. Thus, in acetic acid and the allied compounds, he assumes the radical C2 C0H3 which, following the example of Liebig, he calls acetyl ; 36 and he writes : — (C2H3fC2O, HO Aldehyde, (CoHjfGjOa, HO. Acetic acid, (C2H3)~C2C13 Regnault's chloride of acetyl,37 (C2H3)~C, • ° Acetamide [C = 6, O = 8]. Although this formula for acetic acid does not differ essen- tially from that of Berzelius, still there was much that was new and valuable in the considerations underlying these symbols. For example, Kolbe now draws attention to the fact that the four carbon equivalents of acetic acid (equivalents in Gmelin's sense) do not really possess the same function, but that two of them are contained in it in the form of methyl, while the other two serve as a point for engaging the affinity of the oxygen. The formulae of the other fatty acids are obtained from that 32 Proc. Roy. Soc. 14, 19; Annalen. Supplementband 4, 109. 33 Proc. Roy. Soc. 9, 341 ; Annalen. 108, 67. 34 Ibid. 127, 31. :Jr> Ibid. 75, 211, and 76, i. 'M Compare p. 137. :!7 Annalen. 33, 319. LECTURE XII.] HISTORY OF CHEMISTRY. 229 of acetic acid by replacing the methyl by ethyl, propyl, amyl, etc., while in benzoic acid the radical phenyl occupies the place of the methyl. In general, Kolbe employs the so-called homo- logous and isologous radicals as equivalent to one another, just as Gerhardt had done. The radical ethyl is assumed in alcohol, which Kolbe writes (C4H5)O, HO, as Liebig also did, only with different atomic weights ; on oxidation it splits into C.7H3 and C.,H.,, and the latter is then converted further into C2O0. This explanation is complicated in comparison with the one given by Williamson, but still it afterwards led to important con- clusions (see p. 237). Kolbe formulated Leblanc's monochloracetic acid,38 and Dumas' trichloracetic acid 2, 03 and HO(QC13)C>,O3. The formulae become much more complicated in the case of the products obtained by the action of sulphuric acid upon organic acids, where the mode of writing them approximates to that proposed by Dumas andPiria for the "acides conjuges."39 Sulphacetic acid, for example, becomes c " c,-< so, Co, o Kolbe is still undecided at this time as to whether he should admit the existence of dibasic acids, and he, therefore, retains the old formulae. Accordingly, oxalic acid is HO, C,O3, and succinic acid is HO, (CoH.2)C,O3. It is also worthy of mention that Kolbe assumes radicals containing oxygen in anisic and in salicylic acids, and that he. therefore, no longer agrees with Berzelius upon this point. In addition to the conjugated metallic radicals and carbon radicals, Kolbe also recognises radicals containing sulphur, and thus the analogy, already noticed, between the ordinary 38 Ann. Chim. [3] 10, 212. 39 Annalen. 44, 66. 230 HISTORY OF CHEMISTRY. [LECTURE XII. and the sulpho-acids, is preserved. Thus, we have, for example, HO(C2C18)S2, 05 HO(C2Cl3fc2, 08. Chlorkohlenunterschwefelsaure Chlorkohlenoxalsaure (Trichlormethyl-sulphonic acid). (Trichloracetic acid). The paper of Kolbe referred to here forms the complete foundation of a chemical system, from which I have only been able to select the most important parts. In it the attempt is made to maintain the radical theory, but the fundamental conception of this theory has undergone important changes. Thus the capability of radicals to undergo substitution was now necessarily admitted, and, with this admission, the radicals ceased to occupy an exceptional position. Besides this, the conjugated radicals had been tacked on to the theory, and these had not been by any means sharply defined. Kolbe tries to rescue the electro-chemical theory, but he is obliged to make very important admissions to the opponents of Berzelius. Opposite electrical conditions are still supposed to exist between the constituents of a compound ; but which is the positive and which the negative constituent remains undecided, simply because Kolbe assumes that the same element may possess different electro-chemical properties — an assumption for which justification is found in the existence of elements in allotropic conditions. But the very admission which Kolbe makes, becomes the central point of the con- troversy ; and it is only demonstrated anew that the theory of Berzelius, in the old form, is no longer tenable. In addition to Frankland, Kolbe had only a few special adherents, and when the former made important changes in the notions with respect to coupling, in 1852, Kolbe, having regard to the facts, was obliged to modify his views. The new hypotheses which he advances, now approach much more nearly to the notion of types, even although his mode of naming and of formulating substances is peculiar to himself. As regards its fundamental principles, the system of Kolbe ranks below that proposed by Gerhardt, particularly because it does not contain any distinction between molecule, atom, LECTURE XII.] HISTORY OF CHEMISTRY. 23! and equivalent ; but still it also possesses over the latter system, certain advantages which are to be found especially in the greater importance that is attached to the formulae, and in the breaking up of the radicals containing carbon into simpler ones. I have just pointed out that, in Kolbe's opinion, the radical (or element) with which a substance is conjugated has only a subordinate influence upon the nature of the compound; Frankland attacks this doctrine in i852,40 and he succeeds in convincing Kolbe that it cannot be maintained. Frankland justifies his views by reference, especially, to the radicals containing metals. In the coupling of arsenic with methyl, the former, according to Frankland, changes its saturat- ing capacity. Whilst it possesses in the free state, the capacity for uniting with five atoms of oxygen, the highest stage of oxidation of cacodyl contains only three atoms of this element. The remaining organo-metallic compounds give occasion to similar considerations, and in consequence Frankland is led to make the following important observations: "When the formulae of inorganic chemical compounds are considered, even a superficial observer is struck with the general symmetry of their construction ; the compounds of nitrogen, phosphorus, antimony and arsenic especially exhibit the tendency of these elements to form compounds containing three or five equivs. of other elements, and it is in these proportions that their affinities are best satisfied ; thus in the ternal group we have N03, NH3, NI8f NS3, P08, PH3, PC13, SbO3, SbH3, SbCl3, AsO3, AsH3, AsCl3, etc. ; and in the five-atom group, NO5, NH4O, NH4I, PO5, PH4I, etc. Without offering any hypothesis re- garding the cause of this symmetrical grouping of atoms, it is sufficiently evident, from the examples just given, that such a tendency or law prevails, and that, no matter what the character of the uniting atoms may be, the combining power of the attracting element, if I may be allowed the term, is always satisfied by the same number of these atoms. It was probably 40 Phil. Trans. 1852, 417; Annalen. 85, 329. 232 HISTORY OF CHEMISTRY. [LECTURE XII. a glimpse of the operation of this law amongst the more com- plex organic groups, which led Laurent and Dumas to the enunciation of the theory of types ; and had not those distin- guished chemists extended their views beyond the point to which they were well supported by then existing facts — had they not assumed, that the properties of an organic compound are dependent upon the position and not upon the nature of its single atoms, that theory would undoubtedly have contributed to the development of the science to a still greater extent than it has already done ; such an assumption could only have been made at a time when the data upon which it was founded were few and imperfect : and, as the study of the phenomena of substitution progressed, it gradually became untenable, and the fundamental principles of the electro-chemical theory again assumed their sway. The formation and examination of the organ o-metallic bodies promise to assist in effecting a fusion of the. two theories which have so long divided the opinions of chemists, and which have too hastily been considered irre- concilable ; for, whilst it is evident that certain types of series of compounds exist, it is equally clear that the nature of the body derived from the original type is essentially dependent upon the electro-chemical character of its single atoms, and not merely upon the relative position of those atoms."41 It is then pointed out, in conclusion, how " Stibethin furnishes us, therefore, with a remarkable example of the operation of the law of symmetrical combination above alluded to, and shows that the formation of a five-atom group from one containing three atoms, can be effected by the assimilation of two atoms, either of the same, or of opposite electro-chemical character." 42 Frankland, consequently, gives up the idea of coupling, and now regards cacodyl as sulphide of arsenic in which both sulphur atoms are replaced by methyl. He had now adopted, although in a somewhat different form, the theory of types, and if he considers that he really differed from the decided adherents of this theory, inasmuch as he did not assume, with 41 Phil. Trans. 1852, 441 ; Annalen. 85, 368. 42 Phil. Trans. 1852, 442 ; Annalen. 85, 371. LECTURE XII.] HISTORY OF CHEMISTRY. 233 them, " that the properties of an organic compound are de- pendent upon the position and not .upon the nature of its single atoms," I cannot altogether agree with him in this. Since the investigations of Hofmann 43 on substituted bases, the idea of substitution was no longer held, even by Laurent, in the absolute sense in which the latter had at one time stated it.44 The chlorinated ethers previously prepared by Malaguti 45 could not by any means be brought into harmony with com- plete invariability of the type ; and when Williamson referred ether, alcohol, and acetic acid to the water type, it was plain that he used the word type more in the sense of the mechanical than of the chemical types. By this paper of Frankland's the first step was taken in the approach towards one another of the heretofore separated schools, and the way to a mutual understanding was provided. It was destined to lead to a fusion of the different opinions, out of which the theory of valency then arose. The change of opinion on the part of Frankland was a gain to the supporters of the theory of types, since he brought with him novel ideas which were capable of being turned to excellent account. I do not assert that they might not themselves have been able independently to make the last great advance — that is, the step to the classification of the atoms according to their valence. In the way in which the development actually took place, the influence of Kolbe, and more particularly that of Frankland, upon the supporters of the Gerhardt-Williamson school (VVurtz, Kekule, and Odling) can hardly fail to be recognised. Both schools were required, in order to raise the significance of the formulae to what it subsequently became; especially as Williamson, the only one who desired, even at that time, to write anything more than decomposition formulae, withdrew from the further development of chemistry. It may now be expedient to explain at once the transition, for which the way had been prepared by Frankland, from the theory of copulae to that of types ; and then, when I enter 43 Annalen. 53, I. 44 Comptes rendus par Laurent et Gerhardt, 1845. 45 Annalen. 24, 40 ; 56, 268. 234 HISTORY OF CHEMISTRY. [LECTURE XII. upon the consideration of atomicity and of structural formulae, I shall only need to refer to Kolbe's views, and shall be the better able to point out the influence which he exercised. It was not an easy matter for Kolbe to follow Frankland in his most recent developments ; to assume that the affinity of the elements is always satisfied by the same number of atoms without regard to their chemical character amounted to giving up the electro-chemical theory altogether, and to admitting that the electro-chemical nature of the elements was without influence upon the formation of compounds. Kolbe was not, at first, able to reconcile himself to this.46 In his Text-book, while recognising the premises of Frankland's arguments, he endeavours to combine these with the electro-chemical prin- ciples by means of new hypotheses;47 and it is only in 1857 46 Kolbe, Lehrbuch der Chemie 1854, I, 20 et seq. 47 As evidence for this statement, which Kolbe attacked as erroneous (J. pr. Chem [2] 23, 365), I quote the following passage from Kolbe's Lehrbuch. I, 23 : — " Frankland felt himself justified in concluding from this that in cacodyl, stibrnethyl, stannethyl, etc., a real replacement of different oxygen atoms by the same number of atoms of methyl or ethyl takes place ; in other words, that cacodylic acid is arsenic acid which con- tains two atoms of methyl in "place of two atoms of oxygen, and that oxide of stannethyl must be regarded as composed according to the rational formula ( C* TT Sn -! i-v 5, in which the substitution of one atom of oxygen by one atom of ethyl is evident. However little it is possible to agree with this opinion, there can still be no doubt that a regularity does prevail here. The circum- stance is perhaps deserving of attention, that, as is well known, those very elements which stand next after potassium in the electro-chemical series — that is, the metals of the alkalies and of the alkaline earths — unite with oxygen in but few proportions ; whereas those upon the other side, such as chlorine, sulphur, nitrogen, phosphorus, etc., take up oxygen, on the con- trary, in very numerous proportions. Accordingly, when one of these elements by virtue of its coupling with hydrogen or with ether radicals, approaches more closely to potassium in respect to its electro-chemical character and its affinities, its capacity of now uniting with fewer atoms of oxygen than previously, in consequence of this change of position in the electro-chemical series, may probably be found. less surprising; although it may not by any means be explained how it comes that the number of atoms of the copula and of oxygen is regularly increased up to a definite number," LECTURE XII.] HISTORY OF CHEMISTRY. 235 that he adopts Frankland's views,48 which he further develops and turns to account, especially in organic chemistry. He only publishes the detailed statement of the opinions so arrived at, in 1859, in a paper "On the natural relationship of organic to inorganic compounds," 40 which contains many new ideas. Frankland had compared the radicals which contain metals with the corresponding oxides. Kolbe now says that "The chemical organic substances are wholly derivatives of inorganic compounds, and are formed from these, directly to some extent, by extremely simple substitution processes." Carrying out a suggestion made by Liebig,50 he derives the compounds of carbon from carbonic acid, and those of sulphur from sulphuric acid. The experimental bases of these opinions are the work partly of Mitscherlich,51 partly of himself (compare p. 2241/4?^.), and partly also of Wanklyn,52 who had succeeded in preparing propionic acid from sodium ethyl and carbonic acid. Kolbe employed at this time, as well as long afterwards, the atomic and equivalent weights of Gmelin, adopting at the same time molecular weights for the majority of compounds agree- able to the determinations of Gerhardt, Laurent, and William- son. Accordingly he writes carbonic acid C2O4, and from this anhydride he apparently derives the organic compounds such as acids, aldehydes, ketones, alcohols, etc. I say " apparently," for I shall afterwards show that it is not really so ; but I shall, in the first place, state Kolbe's system in the form in which he applied it. In carbonic acid, oxygen atoms are distinguished from one another according to whether they are within, or outside of the radical. The formula is therefore written (C.2O.2)Q.,, carbonic oxide being regarded as the radical of carbonic acid. When an atom of oxygen outside the radical is replaced by hydrogen or an alcohol radical, the series of the fatty acids is obtained : — HO, H (C.2O,) O Formic acid, HO, C2H3 (C2Ot) O Acetic acid, etc. When the second oxygen atom also is replaced by an alcohol 48 Annalen. 101, 257. 49 Ibid. 113, 293. 50 Ibid. 58, 337. 51 Ibid. 9, 39- 5'2Journ. Chem. Soc. II, 103; Annalen. 107, 125. 236 HISTORY OF CHEMISTRY. [LECTURE XII. radical, a ketone is formed, and when it is replaced by hydrogen an aldehyde is formed : — ?TT JGoO, Aldehyde ; &58JC8O2 Acetone. ^2 •"•3-' \-"2"-$J By the replacement of three atoms of oxygen by three of hydrogen, or by two of hydrogen and a radical, the alcohols are obtained : — HO H3 C2, O Methyl alcohol. HO ~ 54a, O Ethyl alcohol, etc. U2tigj If this mode of deriving substances is more minutely examined, we recognise that Kolbe's procedure is not altogether justified. By the replacement of an atom of oxygen in carbonic acid by hydrogen, HC.2O3 is obtained, and not formic acid. Kolbe simply adopts from the dualists, the error of sometimes adding HO to the formula, and sometimes omitting it. It is true that he states a reason for his doing so, inasmuch as the basicity of a compound (and therefore also the number of HO groups) is determined, according to him, by the number of oxygen atoms outside the radical. Thus nitric acid is mono- basic, since it is (NO4)"O; sulphuric acid is dibasic because it contains two oxygen atoms outside the radical (S.,O4) O2 ; and phosphoric acid is tribasic (PO.2)O3. Since, in accordance with this mode of regarding the matter, every atom of oxygen outside the radical carries -with it one OH group, when Kolbe speaks of the replacement of such an atom of oxygen, this means that O + OH = O.,H is substituted ; and, taken in this sense, the view can be sustained in a strictly logical manner. It is necessary, however, to start from the hypothetical hydrated carbonic acid 2 HO, (C2O2) O.2. By the replacement of O0H by H we get HO,H(C,O,)O Formic acid, O2H „ CoH8 „ HO,(C2H3)~(C2O2)O Acetic » 2O2H „ H2 „ HJC2O2 Methyl aldehyde, 20.,H „ (C.,H8).,,, &28)c.,0o Acetone, LECTURE XII.] HISTORY OF CHEMISTRY. 237 O.2H by H and of O, by H, HO,H3C2,O Methyl alcohol, 02H „ C,,H3 „ O, „ H, HO, CH2}C2,0 Ethyl alcohol. On the conversion of the alcohols into the corresponding acids, the two hydrogen atoms are replaced again by oxygen equivalents. Kolbe's view is now more definite than William- son's. Whereas the latter assumes the conversion of the radical C.,H5 into C9H3O [C=i2], according to Kolbe, C0 r2rj is " ^2"3 H produced from C, •{ H . The difference is important, and - [C.H, it leads Kolbe to foresee the existence of a new class of alcohols, which he announces as follows : — 53 " If the undernoted formulae, by means of which I pre- viously represented the rational composition of acetic acid, and of the corresponding aldehyde and alcohol, are inspected— - HO. (C2H8) [C,OJ O Acetic acid, J Aldehyde, HO.|C^3| C2,O Alcohol, it will be understood, at the first glance, how it comes that of the five hydrogen atoms in the ethyl oxide of the alcohol, only two are substituted in the oxidation of the latter and only one in that of aldehyde. It is those atoms of hydrogen in alcohol and in aldehyde which stand by themselves that are subject to the oxidising influences, and that present themselves to the oxygen as points of attack far more easily accessible than the other hydrogen atoms which are more firmly held in the methyl radical. " The above conceptions of the chemical constitution of the alcohols reveal to us the prospect of the discovery of new alcohols as yet unknown, as well as of a new class of substances which, while closely related to the alcohols in respect to their composition, will also probably share many properties with 5:1 Annalen. 113, 305-306. 238 HISTORY OF CHEMISTRY. [LECTURE XII. them, but must also behave differently from them in many essential points." These new substances can also be obtained from carbonic acid or the fatty acids by substitution. We have : — 2 HO (C./X), O, Carbonic acid, HO CgHg (C"2O2) O Acetic H 1 r1 n Alcohol with one hydrogen atom substi- (C2H8)2/ tuted by methyl (dimethyl-carbinol), HO (C0H3)3 C.,, O Alcohol with two hydrogen atoms substi- tuted by methyl (trimethyl-carbinol). Kolbe goes so far as to prophesy the chemical character of these hypothetical substances. Thus, according to him, alcohol with one hydrogen atom substituted by methyl must yield acetone on oxidation, by virtue of a reaction which is analogous to the conversion of the normal alcohols into aldehydes :— HO C-H3}c,, O gives ^^JCA Aldehyde, (~* T~T "I HO C!HJ -C,, ° Sives r2S3)c-°" Acetone. -H J ^2««J All these conjectures have been justified in the most brilliant manner, and consequently they have exercised a guiding influ- ence upon the development of the considerations regarding constitution. For this reason, I must return to them in the next lecture. Kolbe now admits the existence of polybasic acids. Dibasic acids are produced, according to him, from two "atoms" of carbonic acid by the replacement of two oxygen atoms outside the radical (and therefore of twice O.,H) by bivalent radicals such as ethylene, phenylene, etc.54 thus : — Succinic acid 2 HO (C4 H4) f £*^ j-O2 Phthalic „ 2 HO (CuHijI&S^Oy ^2^2^ 54 The idea of polyatomic alcohol radicals is not due to Kolbe, but to Williamson and Wurtz, as the accompanying development of the subject shows. LECTURE XII.] HISTORY OF CHEMISTRY. 239 Tribasic acids are derived in the same way, from three atoms of carbonic acid, by the replacement of three oxygen atoms by trivalent radicals. Of the other highly interesting matters discussed in the paper, I mention only the mode of regarding the sulpho-acids, in which the analogy with the carbon acids, mentioned already, again makes its appearance. Just as the latter may be derived from carbonic acid, so the former may be derived from sulphuric acid. We have — 2HO(S, O4) O, Sulphuric acid, HO (C2 H8) (S2O4) ° Methylsulphonic acid, HO (C12H6) (S2O4) O Phenylsulphonic „ The dibasic sulpho-acids are produced from two atoms of sul- phuric acid : — ur\ in u \ /S.,O,\ M Disulphometholic acid (methylene [C2H2) ^S-Q J O, disulphonic acid), wr» ir T-r \ /'S-jOA n Disulphobenzolic acid (phenylene tt U disulphonic acid). Besides these, Kolbe is acquainted with intermediate acids which are derived from an atom of carbonic acid and an atom of sulphuric acid ; amongst them are sulphacetic and sulpho- benzoic acids : — 2 HO (C,H,) '2 O., Sulphacetic acid, 2 HO (C^H,) (£*%?} O., Sulphobenzoic acid. \Oo ^4/ This mode of regarding them furnishes a simple explanation of the conversion, observed by Buckton and Hofmann, of sulph- acetic acid (really of acetonitrile) into disulphometholic acid by treatment with sulphuric acid.55 In this operation C2O2 is replaced by S.,O4. I cannot enter here upon the other points in this extremely important paper, but must advise that a study be made of it, as it is full of clever ideas. It is true that there are views advanced in it with which I cannot agree. Thus, for example, Kolbe did 53 Annalen. 100, 129. 240 HISTORY OF CHEMISTRY. [LECTURE XII. not regard the conception of polyatomicity in the way in which it had been advanced by Williamson, otherwise he could not have inquired why amid-acids with a monobasic radical do not also exist, if these are referred, as was done by Gerhardt and Kekule, to the type NH3 + H.2O.5G There can, consequently, be no discussion as to whether Kolbe had not recognised the quadrivalence of carbon before Kekule. Even although the former rendered unquestionable services with respect to the origination of constitutional or, as Butlerow calls them, struc- tural formulae, still his participation in the development of the notions as to the atomicity (valency) of the elements and radicals is not of importance, because, as I believe, he did not distinguish between molecule, atom, and equivalent, and also, as follows from the foregoing, because he had not then grasped the idea of the part played by the polyatomic radicals in holding the molecule together. The doctrine of valency was possible, and was bound to ensue as soon as atom and equivalent were separated from each other. If the atoms were not equivalent, the question as to the valency of one when referred to another must necessarily arise. Consequently, those who first distinguished the two ideas from each other, took the first step towards considerations of atomicity, and Dumas, Liebig, and Laurent must be men- tioned in this connection. Whilst the different valency of the elements was recognised by means of the phenomena of substi- tution, the theory of polybasic acids led to the conception of the polyatomic radicals. Both views remain side by side for a long time without exercising any important influence upon each other, until a fusion of the two took place at the hands of Kekule ; that is, until the valency of the radicals was explained by that of the elements. We have already seen in the preceding lecture57 how Williamson was led to advance the conception of the polyatomic radicals. He employed it to explain the formation of chemical compounds, inasmuch as the polybasic radicals possess the 56 Annalen. 113, 324. 57 Compare p. 214. LECTURE XII.] HISTORY OF CHEMISTRY. 241 power of holding together several atomic groups. There were but few who then understood the meaning which lay in Williamson's words, and likewise few who recognised the extension that might be given to them. Amongst these, it was Kekule in particular, whose sagacity led him to perceive at once the bearing of Williamson's ideas, and who made use of these ideas to elucidate the relations of thioacetic acid which he had discovered in i853.58 - Kekule compares the reactions of phosphorus pentachloride and of phosphorus pentasulphide upon acetic acid, writing : — [C=i2,0=i6, 6 = 32, etc.] Referring to this he remarks: — "The above diagram . . . exhibits ... the relations between the reactions obtained with the chlorine and with the sulphur cojnpounds of phosphorus. In fact, it is perceived that the decomposition is, in its essential features, the same; only, when the chlorides of phosphorus are employed, the product breaks up into chlorothyl [C2H3OC1] and hydrochloric acid, . . . whereas when the sulphur com- pounds of phosphorus are employed, both groups remain united, because the quantity of sulphur which is equivalent to the two atoms of chlorine is not divisible." These points of evidence lead Kekule to declare in favour of the accuracy of the "new atomic weights " (Gerhardt's equi- valents). These, according to Kekule, are a better expression of the facts than the mode of representing them previously in use. Even if the new formulae are adopted, and the old equivalents retained, it cannot be perceived why phosphorus sulphide produces mercaptan from alcohol whilst phosphorus chloride forms ethyl chloride and hydrochloric acid (CfL^C1 and HC1); why the latter do not remain combined just as C4H5 S and HS do, etc. It is not merely a difference in the 58 Annalen. 90, 309. Q 242 HISTORY OF CHEMISTRY. [LECTURE XII. mode of representing them, but it is a real fact that an atom of water contains two atoms of hydrogen, and only one of oxygen ; also that the quantity of chlorine equivalent to one indivisible atom of oxygen is divisible by two, whereas the sulphur, like oxygen, is dibasic, so that one atom is equivalent to two atoms of chlorine.59 The investigations by Frankland, of the radicals containing metals, and his views as to saturating capacity (compare p. 231) had an important influence upon the development of the theory of polyatomic radicals, as also had the interesting paper on salts by Odling, and the important researches of Berthelot on glycerine and of Wurtz on glycols. We shall consider these more minutely. Odling 60 makes a distinct advance by applying the idea of polybasicity to the metals also, and by reintroducing molecular formulae for all salts, even for those of the sesqui oxides, for which Gerhardt had written equivalent formulae. He not only refers the polyatomic acids to condensed types, as Williamson had done before him, but he is also acquainted with poly- atomic bases, which can be regarded in a similar way. Thus, for example, he writes ? — Oxide of Bismuth QsO ; Nitrate of Bismuth Those metals which possess, according to Gerhardt, various equivalent weights, now have several atomicities assigned to them. Odling knows, for example, monatomic and triatomic iron, and monatomic and diatomic tin, whence he obtains the following formulae :— to citric acid C*" Ferrous Fe' ) oxide Fe'J 59 Annalen. 90, 314. Gt> Journ. Chem. Soc. 7, I. 61 Odling indi- cates the atomicity of the elements by means of the dashes to the right of their symbols. LECTURE XII.] HISTORY OF CHEMISTRY. 243 His conception of the acids of phosphorus is likewise interesting. Odling writes : — Ordinary phosphoric acid ,-T J-O3 Pyrophosphoric acid - |r fO6 Jrl4-' PO'"1 Metaphosphoric acid HJ ^L> PO'^PFF "'i Phosphorous acid VT j-O5 T)TT '/' Hypophosphorous acid j| O.2. According to Odling, therefore, phosphorous acid stands to pyrophosphoric acid in the same relation as hypophosphorous acid to metaphosphoric acid. A similar relation exists between dithionic and sulphuric acid on the one hand, and oxalic acid and carbonic acid on the other : — CO Carbonic Oxalic Sulphuric Dithionic acid. acid. acid. acid. Kay, a pupil of Williamson's,6'2 almost simultaneously pub- lished a research (obviously suggested by his teacher) which deserves our notice. By the action of sodium ethylate upon chloroform, he had obtained an ether which he called tribasic formic ether, and which had been produced according to the following equation : — 4. •? 2 5 O — O 4- -7 MsP1 C13 + 3 Na ( ~(C2H5)3°* + 3N Williamson specially draws attention to the fact that the resi- dues of three molecules of alcohol are held together in the new substance by the trivalent radical CH. This was the first example of a polyatomic hydrocarbon radical, and it was soon to be shown how valuable was this mode of regarding the sub- stance. Berthelot, occupied at the time with the investigation of glycerine (which was completed in 1854 so far as its most & Proc, Roy, Soc. 7, 135, 244 HISTORY OF CHEMISTRY. [LECTURE XII. important parts were concerned),"3 found the very important result that glycerine can unite with acids in three different proportions. Thus : Monostearine = i Glycerine + i Stearic acid - 2 Water = C0HS0, + C30H3r,04 -2 HO [C = 6, 0 = 8] Distearine = i Glycerine + 2 Stearic acid - 4 Water = C6H806 + 2C3t;H3604 -4 HO (In the paper 2 HO is given.) Tristearine = i Glycerine -f 3 Stearic acid - 6 Water = C6H806 +3C36H3604 -6 HO Monochlorhydrine= i Glycerine + i Hydrochloric acid - 2 Water - CCH8O(3 + HC1 - 2 HO Dichlorhydrine =i Glycerine + 2 Hydrochloric acid - 4 Water = C6H8Oa + 2HC1 -4 HO Berthelot interprets these facts in the following manner : — " These facts show us that glycerine exhibits the same rela- tion to alcohol that phosphoric ncid does to nitric acid. In fact, whilst nitric acid "does not produce more than one series of neutral salts, phosphoric acid gives rise to three distinct series of neutral salts, the ordinary phosphates, the pyro- phosphates, and the metaphosphates. These three series of salts, when decomposed by powerful acids in presence of water, reproduce one and the same phosphoric acid. " Likewise, whilst alcohol only produces one series of neu- tral ethers, glycerine gives rise to three distinct series of neutral compounds. These three series, on complete decomposition in presence of water, reproduce one and the same substance, glycerine." This comparison between glycerine and phosphoric acid on the one hand, and alcohol and nitric acid on the other, is of great importance, even although it is unfortunately to some 63 Comptes Rendus. 38, 668 ; Ann. Chim. [3] 41, 216 ; Annalen. 88, 304 ; 92, 301, LECTURE XII.] HISTORY OF CHEMISTRY. 245 extent impaired by taking account of pyrophosphoric and meta- phosphoric acids. Odling's orthophosphoric acid 64 has neither the same composition nor the same basicity as these two latter acids, whilst it is always the same substance, glycerine, which is contained in the ethers prepared by Berthelot. Wurtz was happier in his view of these remarkable facts, looking, as he did, upon glycerine as a triatomic alcohol and representing it as c* TT "'1 S5 "OG.05 Tne compounds examined by Berthelot are W3 J produced, according to Wurtz, by the replacement of one, two, and three hydrogen atoms by acid radicals. In this connec- tion he points out how the monatomic group C0Hr passes into the trivalent residue QH5 by loss of H2. [C = 6, O = 8.] It could not escape the notice of a clever investigator like Wurtz that the existence of monatomic and of triatomic alcohols necessarily involves that of diatomic alcohols, and he at once institutes experiments which have for their aim the preparation of such substances. From his way of looking at the subject he necessarily expected a diatomic radical in the still hypo- thetical alcohol ; the univalent group C6Hr and the trivalent group C,3H5 rendered possible the formation of the monatomic and of the triatomic alcohols ; the homologues of C^H,. must correspond to the diatomic alcohols.66 The chlorides and bromides which were already well known (in part, at least) favoured the accuracy of this view, and it was now simply a question of converting them into the corresponding hydroxy- compounds and the object was attained. The action of the basic hydroxides did not realise the hopes which Wurtz had entertained, but earlier experiences now came to his aid. Four years previously, he had discovered a reaction whereby a similar conversion could be carried out ; 6T and this reaction had subsequently proved serviceable in several cases.68 It was now to appear that it deserved to be called a general method. 64 Phil. Mag. [4] 18, 368. 65 Ann. Chim. [3] 43, 492. 66 Compare Odling, Journ. Roy. Inst. 2, 62. 67 Comptes Rendus. 35, 310; 39, 335 ; Ann. Chim. [3] 42, 129. 68 Zinin, Annalen. 96, 361 ; Cahours and Hofmann, ibid. IOO, 356. 246 HISTORY OF CHEMISTRY. [LECTURE XII. Thus, by heating ethylene iodide with silver acetate, Wurtz obtained an acetic ether, which, on decomposition with potash, yielded the desired alcohol :— C4H4Ia + aC In this way Wurtz succeeded in preparing glycol, the first of the diatomic alcohols.09 He was well rewarded for the difficulties of the investigation, for it is seldom that the dis- covery of a single substance has exercised such an influence upon the development of chemistry, and seldom that a single compound has given rise to such a series of elegant and useful investigations as this glycol did. I may be permitted to justify this assertion by making some observations with respect to the compounds which stand in the closest relation to glycol. By the oxidation of glycol, Wurtz obtained glycollic acid and oxalic acid.70 The former was identical with the sub- stance that Horsford had prepared from glycocoll ten years previously,71 the nature of which had been announced by Strecker.712 In exactly the same way lactic acid is produced C" "FT O ^ from propylene glycol.73 Wurtz proposed (i^4 - -O4 as the formula of the former, regarding it, and also glycollic acid, as dibasic acids.74 The discovery of ethylene oxide and of -the polyethylene alcohols was also of great importance. By the H \ treatment of glycol-chlorhydrine C4H4 -O2,'° (obtained from Cl ) glycol by the action of hydrochloric acid) with potash solution, Wurtz obtained the ether of the diatomic alcohol, which stands 69 Comptes Renclus. 43, 199, 1856; compare also 43, 478; 45, 306; 46, 244 ; 47, 346. 70 Ibid. 44, 1306. 71 Annalen. 60, I. 72 Ibid. 68, 55 ; compare Socoloff and Strecker. 80, 38. 73 Comptes Rendus. 46, 1228. 74 Strecker (Annalen. 8l, 247) adopted a formula for lactic acid which was double the above. 73 Comptes Rendus. 48, 101 ; 49, 813 ; 50, H95; 54> 277- LECTURE XII.] HISTORY OF CHEMISTRY. 247 in the same relation to the latter as the anhydride of sulphuric acid does to the acid :— C4H40, S204A Ethylene oxide. Sulphuric anhydride. Glycol. Sulphuric acid. By heating ethylene oxide with glycol, Wurtz next prepared the polyethylene alcohols 7G which Louren^o had obtained, a short time previously, from ethylene bromide and glycol.77 The importance of these substances was enhanced by the acids obtained from them by oxidation ; 78 they furnished excellent examples of the formation of substances according to the con- densed types such as Wurtz afterwards so well understood how to employ in explaining the silicates.79 Finally, by the action of ammonia and its analogues, Wurtz obtained from ethylene oxide bases containing oxygen.80 These substances afterwards became of greater interest in consequence of the synthesis of neurine from glycol-chlorhydrine and trimethylamine.81 The adoption of ethylene as a diatomic radical furnished Hofmann 8L> with a means of correctly regarding the bases pre- pared by Cloez in i853.83 These now appeared as substances derivable from two molecules of ammonia, and bearing the same relation to glycol that ethylamine does to alcohol. By the further study of these substances,84 Hofmann succeeded in adducing new proofs of the accuracy of the theory of poly- atomic radicals, and in preparing the way to a clear conception of the complicated metal-ammonium compounds. It would be unjust if I were to close these observations without alluding to H. L. Buff's claim with respect to the recog- 76 Comptes Rendus. 49, 813. 77 Ibid. 49, 619. 78 Ann. Chim. [3] 69, 317. 79 Repert. de Chimie pure. 2, 449 ; compare also Lecons de philosophic chimique, Paris, 1864, 181. 80 Comptes Rendus. 49, 898 ; 53, 338. 81 Ibid. 65, 1015. 82 Ibid. 46, 255. 83 L'Institut. 1853, 213 ; Jahresbericht 1853, 468. 84 Proc. Roy. Soc. 10, 224 and 594 ; Comptes Rendus. 49, 781 ; 51, 234 ; Proc. Roy. Soc. II, 278 ; Comptes Rendus. 53, 18, etc. 248 HISTORY OF CHEMISTRY. [LECTURE XII. nition of the bivalence of ethylene. Buff, in a preliminary com- munication, several months prior to the first publication by Wurtz on the subject of glycol, and in a paper laid before the Royal Society one month prior to this publication, had endea- voured to prove the diatomic nature of the hydrocarbons CWH;£ [C = 6].85 By treating ethylene chloride with potassium thiocyanate, he had obtained a substance of the formula C4H4Cy2S4, and this, on oxidation with nitric acid, yielded a compound identical with Buckton and Hofmann's disul- phetholic acid80 for which Buff proposes the name ethylene H rsA sulphurous acid, representing it by the formula C4H4 -[ H t$A- The formation of ethylene sulphocyanide is represented by the equation : — Cv^i ^ (S* C4H4C1, + 2^y \ S, = C4 HA + 2KC1. Cy (S, It appears from the whole paper that Buff had recognised the diatomic nature of ethylene, and that, in this respect, he can lay claim to priority over Wurtz. As regards the experi- mental proof of this view, the two investigations scarcely admit of comparison. The examination of the glycols by Wurtz is amongst the most brilliant achievements of the period ; and by its means the hypotheses as to the different valencies of the radicals were provided with so broad a basis that nothing more could be desired in this respect. Buff's experiments conformed to the same theoretical conceptions, but they never could have led to the conclusions which were drawn from the labours of Wurtz. 85 Annalen. 96, 302; Proc. Roy. Soc. 8, 188. 8G Annalen. IOO, 129. LECTURE XIII. IDEAS REGARDING THE TYPES— ELUCIDATION OF THE NATURE OF THE RADICALS BY THE VALENCY OF THE ELEMENTS — QUADRIVALENCE OF CARBON — SPECIFIC VOLUME— CONSTITUTIONAL P'ORMUL/E — SEPARATION OF THE IDEAS OF ATOMICITY AND BASICITY — Iso- MERISM AMONGST ALCOHOLS AND AdDS — PHYSICAL ISOMERISM — UNSATURATED SUBSTANCES. IN the preceding lecture I showed how the views respecting the different valencies of the radicals and elements attained a great importance through the labours of Williamson, Frank- land, Kekule, Odling, Berthelot, Buff, and Wurtz. I shall begin this lecture by showing how the types can be explained by means of these views.1 At the period in question Kolbe had attacked Gerhardt's mode of regarding substances as an arbitrary one/2 Wurtz endeavours to show that this is not so, and that Gerhardt's four types, which, in his opinion, can be reduced to three, represent different states of condensation of matter. Besides the hydro- gen type H2, Wurtz also assumes the types H2H2 and H3H3. Water, H2O, formed by the replacement of H2 by O, corre- sponds to the former, while ammonia represents triple condensed hydrogen, in which one half of this element is replaced by trivalent nitrogen. Since all of these formulae correspond to the same volume of the substances in the gaseous state, the view of Wurtz is fully justified. One atom of hydrogen corre- sponds to one volume, to one half of a volume, or to one third of a volume according to whether it is present in compounds which belong to the type H2, H4, or H(i. Wurtz looks upon the existence of still more highly condensed states of matter 1 Ann. Chim. [3] 44, 306. 2 Lehrbuch der Chemie. I, 50. 250 HISTORY OF CHEMISTRY. [LECTURE XIII. as possible, but he does not proceed to introduce types corre- sponding to these states. It was in 1857, on the occasion of the discussions respect- ing the constitution of mercury fulminate, that Kekule 3 first stated that the constitution of this substance, as well as that of the other compounds of the methyl series, may be referred to the marsh gas type, C2H4; and to the latter type mercury fulminate is, according to Kekule's experiments, to be reckoned as belonging.4 He therefore writes : — C2H4 C2H3C1 CjHClg C,(N04)C13 Marsh gas. Methyl chloride. Chloroform. Chlorpicrin. C2(N04)2CL, C2H3(C,N) C,(C,N)(N04)Hg, Marignac's Oil.5 Acetonitrile. Mercury fulminate. A beginning had thus been made, but still the type C2H4 was only of very little use. So long as it could not be extended to all carbon compounds, there could be no question of making it the basis of a system of organic chemistry such as it afterwards actually became. The idea which rendered this possible was still wanting. Kekule had perhaps already conceived it at the time, and merely did not-dare to publish it ; or it may be that the hypothesis of the linking of carbon atoms had not yet occurred to him. In any case, the views which he then held must have approximated very closely to those which he pub- lished in 1858 concerning the nature of carbon, for, at the end of 1857, when he refers the types to the different valencies of the elements,6 he definitely mentions the quadrivalence of 3 Annalen. 101, 200. 4 [C = 6, O = 8.] Kekule here employs these atomic weights again, whereas he had discarded them as inaccurate four years previously. Kolbe (J. pr. Chem. [2] 23, 374) regards as important the fact that Kekule then pointed out that he did not employ the word type "in the sense of Gerhardt's unitary theory, but in the sense in which it was first employed by Dumas upon the occasion of his fruitful investiga- tions on the types." I consider this unimportant, more especially as Kekule proceeds as follows: — "I shall indicate the actual relations in which the substances mentioned stand to one another, by saying that, under the influence of suitable agencies, the one can be produced from, or converted into, the other." 5 Annalen. 38. 16. 6 Ibid. 104, 129. LECTURE xiii.] HISTORY OF CHEMISTRY. 251 carbon ; but this is only referred to incidentally, and does not lead him to any further conclusions. At length, in the spring of 1858, the paper appears which has become of such fundamental importance in chemistry.7 In this paper Kekule begins to direct attention to the necessity of studying the nature of the elements. This alone, in his opinion, can lead to an explanation of the valency of the radicals. As regards organic chemistry, the chief part in such considerations is played by carbon ; and, consequently, the properties of this element are subjected by Kekule to a very minute examination. " When the simplest compounds of this element are considered (marsh gas, methyl chloride, chloride of carbon, chloroform, carbonic acid, phosgene gas, sulphide of carbon, hydrocyanic acid, etc.) it is perceived that that quantity of carbon which chemists have recognised as the smallest pos- sible, that is, as an atom, always unites with four atoms of a monatomic or with two atoms of a diatomic element ; that, in general, the sum of the chemical units of the elements united with one atom of carbon is four. This leads us to the opinion that carbon is tetratomic (or tetrabasic)." The hypothesis of the linking of carbon atoms also appears now, and is dealt with in a very detailed manner. " In the cases of substances which contain several atoms of carbon, it must be assumed that at least some of the .atoms are in the same way held in the com- pound by the affinity of carbon, and that the carbon atoms attach themselves to one another, whereby a part of the affinity of the one is naturally engaged with an equal part of the affinity of the other. "The simplest, and, consequently, the most probable case of such an attachment of two carbon atoms is that in which one unit of affinity of the one atom is united with one of the other. Of the 2x4 affinities of the two carbon atoms, two are therefore employed in holding the two atoms themselves to- gether ; six thus remain over which can be held by atoms of other elements." 7 Annalen. 106, 129. 252 HISTORY OF CHEMISTRY. [LECTURE XIII. On the assumption which is here made, the number of valencies of other elements that can unite with n carbon atoms which are united to one another may be expressed by the equation : — n (4 - 2) + 2 = 2n + 2. It is true that this species of mutual union of carbon atoms does not pass with Kekule as the only one; he draws attention to the fact that in benzene and its homologues a closer union, " the next simplest," may be assumed. In order to indicate the standpoint which Kekule adopted at that time, I may add that, so far as the value of for- mulae is concerned, he is a follower of Gerhardt, and does not conceive these as representing the arrangement of the atoms, but merely as reaction formulae. Accordingly he re- tains Gerhardt's mode of writing them, and assumes, as the latter had done, that several rational formulae are possible for one substance. Kekule is well aware that, starting from the hypothesis of quadrivalent carbon, formulae may appear in a new guise, but he avoids entering more particularly into the matter. This is comprehensible when we recollect that Kekule only attaches such a limited significance to formulae ; since he believes that the physical properties of substances can alone lead to the establishment of hypotheses regarding the arrange- ment of the atoms. These views possess all the more interest from the fact that it had already been stated in a very in- fluential quarter that two formulae cannot be assigned to one substance, and that one substance cannot be referred to dif- ferent types. In collating the results of his excellent investi- gation on the specific volumes of liquids, Kopp 8 showed that these specific volumes can be calculated from the composition of the substances if a certain specific volume is assigned to each element. This volume is not the same in all cases, how- ever, but is dependent upon the part which the element plays in the compound. Thus, for example, according to Kopp, 8 Annalen. 92, I ; 95, 121 ; 96. I, 153, 303; 97, 374; and especially ioo, 19. LECTURE XIII.] HISTORY OF CHEMISTRY. 253 two different specific volumes may be assigned to oxygen, dependent upon whether it exists within or outside the radical. Hence it would not by any means be a matter of indifference, in calculating the specific volumes of aldehydes and of ketones, whether they were referred to the hydrogen or to the water type ; whereas Gerhardt had declared both to be admissible.9 Kopp's rule agreed with the first alternative only. Kopp draws attention to this, and points out that it is exactly in this way that propyl aldehyde is distinguished from the isomeric allyl alcohol : — C.H.01 C3H6\0 H J H JU Propyl aldehyde. Allyl alcohol. It may certainly be looked upon as a very important sign of the times that the chemists of Gerhardt's school are now forced, upon physical grounds, to attach greater value to their formulae and speculations than had hitherto appeared to them to be justified. It is true that no special stimulus was now required. Indeed it would appear that Couper had already tried to write real constitutional formulae. Assuming the quadri- valence of carbon (quite independently of Kekule) Couper had pointed out how the existence of a great number of organic compounds might be explained. I should like to compare this paper of Couper's 10 with that of Kekule, published a short time previously, in order to show how these two investigators, starting from different points of view, arrive at very similar results. Kekule, in recognising and explaining the real essence of the types, hit upon the quadrivalence of carbon and the mutual union of the atoms. Couper, on the other hand, rejects the types because they do not appear to him to satisfy the philosophical requirements which are essential to a theory. According to him, Gerhardt's system rests upon general statements from which individual cases are deduced ; whereas he declares the opposite method to be the 9 Gerhardt, Traite. 4, 632 and 80$. 10 Comptes Rendus. 46, 1157 Ann. Chim. [3] 53, 469. 254 HISTORY OF CHEMISTRY. [LECTURE XIII. only correct one. Couper considers it necessary to study first the properties of the elements ; and these he regards as : 1. Elective affinity, 2. Degree affinity. The latter governs the limits of combining capacity, and coincides approximately with what we now call valency or atomicity. In his further consideration of the subject Couper confines himself to the determination of the degree affinity of carbon, and believes that, by means of it, he can elucidate organic compounds. There are two essential properties of this element which serve to characterise it : *i, It combines only with an even number of hydrogen atoms ; and 2, It combines with itself. The latter assertion is justified by reference to compounds containing carbon. Hydrogen, oxygen, etc., can be removed from these compounds and their places can be taken by chlorine without interfering with the linking, consequently the cause of this linking cannot be sought for in the atoms capable of substitution. The maximum number of atoms united to one carbon atom is four, and from this Couper obtains for what we should now call saturated organic compounds the expression : — These observations are sufficient to enable us to understand Couper's formulae, of which I shall quote a few examples [C = 1 2, 0 = 8]:-" • . a | 3 3| TT /"V TT TT CO— OH CO— OH CO — O"C Alcohol. Acetic acid. Ether. 11 Couper makes curious hypotheses with respect to the properties of the oxygen atom, obviously in order that he may not be obliged to assume in the formation of salts the replacement of the hydrogen by metal (and therefore a reduction of the oxide). According to him O = 8 is bivalent, but one valence must always be satisfied by oxygen. The limit of the valency of nitrogen is adopted as 5. (In the paper in Ann. Chim., the formulae are written on the basis of C — 6.) LECTURE XIII.] HISTORY OF CHEMISTRY. 255 rO— OH rO— OH H2 C02 /H N/O-OH I Ntc N\c pH2 p^2 ^0— OH U0— OH Glycol. Oxalic acid. Hydrocyanic acid. Cyanic acid. We here meet, for the first time, with constitutional for- mulae in the present sense of the term — with symbols called forth in consequence of the recognition of the atomicity of the elements. In connection with these formulae it must be remarked that the views as to alcohol and acetic acid repre- sented by them, concide with those of Kolbe,12 and that the only differences are in the manner of writing them. These two papers of Kekule and of Couper constitute the foundations of our views respecting the structure of com- pounds. As a consequence of them, organic chemistry took an altogether new direction, and they may be regarded as the most important advances of our science, on the speculative side, in recent times. Adopting the doctrine of the quadri- valence of carbon, the endeavours put forth since that time have been largely directed towards arriving at conceptions regarding the mutual relations of the combining atoms ; and I have now to deal with this portion of the historical develop- ment of chemistry. In these considerations respecting constitution, besides the hypothesis as to the nature of carbon, numerous series of experimental data were required ; and in view of the fact that the accumulation of the latter was only very incomplete, the labours of years were frequently necessary in order to render possible any application of the principles stated above, to the determination of rational formulae in the cases of certain classes of substances. Even up to the present we have not succeeded in completely solving this problem, as there are many compounds still which we cannot arrange in the system. What is most important, however, has been accomplished. We are satisfied that the doctrine of atomicity can be employed 12 Compare pp. 235-236, 256 HISTORY OF CHEMISTRY. [LECTURE XIII. as the foundation of an edifice, and in this connection, our thanks are due to Kekule, who, in his excellent text-book, has furnished us with the proofs. Although Kekule has been reproached from many sides, that in carrying out the principles he had advanced, he does not always adhere to them faithfully — an accusation which is not altogether groundless — still I wish to point out that any such want of adherence only took place in cases where the facts were insufficient at the time for a final decision, and that a quite consistent observance of the principles was, therefore, scarcely possible. It may be pointed out here, however, that at the very time when many difficulties attended the employment of structural formulae, and when ambiguities frequently arose, Butlerow 13 and Erlenmeyer 14 constantly advocated them with much zeal. It is not the business of a historical account to follow in detail the establishment of general principles. Such an account must be confined, rather, to developing the history of the rise and decline of prominent ideas ; whereas the enume- ration of the facts, and their arrangement from one common view-point, constitute the sum and substance of the science itself, and must, consequently, be dealt with in text-books. I shall therefore content myself by adducing here what was actually of service in strengthening the system, what led to new conceptions or opinions, what appears to be irreconcilable with the principles and leads us to expect an expansion or alteration of the present theories. I shall begin with a description of the discussion regarding the constitution of lactic acid, which took place within the period 1858-60, and led to the distinction between atomicity and basicity in the case of acids. Following the lead of Gerhardt, who regarded lactic acid as a dibasic acid,15 many chemists doubled the formula for this acid and wrote it C12H12Or) [C = 6, O = 8], whereas the interesting synthesis of alanine, and the conversion of this compound into lactic acid 13 Zeitschrift fiir Chemie. 4, 549 ; 6, 500. 14 Ibid. 7, i. 15 Gerhardt, Traite. I, 689, LECTURE XIII.] HISTORY OF CHEMISTRY. 257 by Strecker,10 made the halved formula more probable. Wurtz, in the oxidation of propylene glycol, brought forward a decisive reason for the latter view.17 At the same time the dibasic character of the acid also seemed to be confirmed, so that Wurtz wrote : — C4H,02\o C4H4)0 H,r4 Glycol . Glycollic acid [C - 6, O = 8]. Propylene glycol. Lactic acid. The reaction with phosphorus pentachloride, which yielded the chloride C(;H4O2C1.2, (a substance that was converted by alcohol into chlorlactic ether, C4HJ 2' corresponding to Cl glycol-chlorhydrine) was a new argument in favour of this view, whilst the vapour density of the chlorlactic ether justified the molecular weight adopted.18 Kolbe regards lactic acid as monobasic, and calls it oxy- propionic acid, assuming the same relation between it and propionic acid as that between oxybenzoic acid and benzole acid.19 In the same way that Gerland was able to convert amido- benzoic acid into oxybenzoic acid (by means of nitrous acid 20), lactic acid can also be obtained from alanine. The latter and glycocoll were to be regarded as amido-acids — a view which found new support upon the conversion of bromacetic acid into glycocoll by Perkin and Duppa 21 —so that Kolbe was able to write : — ,(C4H5)C,(X,O HO, Propionic acid. Alanine. Lactic acid. ](5 Annalen. 75, 27. 17 Comptes Rendus. 45, 306. 18 Ibid. 46, 1228. 1!) Annalen. 109, 257. 20 Ibid. 91, 185. 21 Ibid. io8, 106. R 258 HISTORY OF CHEMISTRY. [LECTURE XIII. Kolbe tries to bring the substances prepared by Wurtz into harmony with his views, contending that lactyl chloride is chlorpropionyl chloride, which passes, by the action of alcohol, into chlorpropionic ether ; as indeed Ulrich obtains propionic ether from it by means of nascent hydrogen.— Kolbe could also have adduced, in favour of his ideas, the preparation of glycollic acid from monochloracetic acid which Kekule had succeeded in effecting,23 whereas Kekule discovers in it the conversion of a monobasic into a dibasic acid.24 Wurtz now brings forward new proofs in support of the accuracy of his view,25 finding these in the existence of the dibasic lactates which had been described by Engelhard and Madrell 26 and by Briining.27 Further, he succeeds in preparing dibasic lactic ether (by treating chlorpropionic ether with sodium ethylate), and also lactamethan and butyro-lactic ether. The reduction of lactic acid to propionic acid by means of hydriodic acid — a reaction discovered by Lautemann 28 — and the conversion of chlorpropionic ether into alanine 29 furnished Kolbe, on the other hand, with new grounds for the assumption that lactic acid is a monobasic oxy-acid. Acids of this kind he defines as monobasic acids in which a hydrogen atom within the radical is replaced by HO9, hydrogen peroxide.30 The analogy between the carbonic and the sulphonic acids, already alluded to several times, is employed in support of the views he is now defending, lactic acid being compared with isethionic acid. Thus :— HO(C4H5)C202>0 HO C4j C A,0 Propionic acid. Lactic acid. HO(C4H6)S204>0 HO(C4|jjg'jS2O4,O Ethyl sulphonic acid. Isethionic acid. 22 Annalen. 109, 268. M Ibid. 105, 286 ; compare also R. Hoffmann, ibid. 102, i. L>4 Compare also Heidelberger Jahrbiicher 1858, 339. 25 Comptes Rendus. 48, 1092. 26 Annalen. 63, 93. 27 Ibid. 104, 191. 88 Ibid. 113, 217. » Kolbe, Ibid. 113, 220. 30 Ibid. 112, 241. LECTURE XIII.] HISTORY OF CHEMISTRY. 259 In this discussion, so far as we have considered it as yet (up to 1859), Kolbe's way of regarding the matter was better adapted to explain the facts than that of Wurtz. In particular, it was possible to explain, in a very satisfactory manner, the relations between the fatty acids and the lactic acids, as well as the phenomena of isomerism amongst the ethers of the latter acids, which Wurtz discovered in the following year.31 What Kolbe misunderstands, are the relations, pointed out by Wurtz, between the glycols and these acids;3'2 and even in 1860, when he returns to the constitution of lactic acid, he still adopts the same standpoint.33 He emphasises the difference between the two hydrogen atoms, replaceable by radicals, in lactic and glycollic acids; but he does not admit that the hydrogen peroxide groups which they contain, also occur in the glycols. Wurtz, meanwhile, has gone a step further. He introduces a distinction between atomicity and basicity in the case of acids.34 Whilst the former of these is determined by the valency of the radical present, the latter is regulated by the number of hydrogen atoms replaceable by metals. According to Wurtz, "the capacity of saturation of an acid towards basic oxides depends not only upon the number of equivalents of typical hydrogen which it contains, but also upon the electro- negative nature of the oxygenated radical. In proportion as the oxygen increases in this radical, the typical hydrogen becomes more and more basic hydrogen." This is illustrated by the following example : — CTT "v (~~* TT (~\\ f~* (~\ ~\ f~* T. oil, \ r^ L/orloU I /•% ^oUfttjm ^9* Gfycol. 2 typical H atoms. ^J Glycollic acid. 2 typical H atoms. I basic. Oxalic acid. 2 typical H atoms. Both basic. Glycerine. 3 typical H atoms. Glyceric acid. 3 typical H atoms. I basic. Glyceric acid is triatomic but only monobasic ; phosphorous and cyanuric acids are triatomic and dibasic. Further, Wurtz 31 Ann. Chim. [3] 59, 161. :52 Compare especially Annalen. 109, 262 etc. 33 Ibid. 113, 306. 34 Bull, Soc. Chim, I, 33; Ann. Chim. [3] 56, 342, 260 HISTORY OF CHEMISTRY. [LECTURE XIII. regards lactic acid as dibasic, and as different in constitution from glycollic acid. He is forced to this by the existence of the lactates described by Briining and others. The first part of Kekule's text-book appeared in the same year, and it was possible to see from it how easily the nature of the lactic acids might be explained by reverting, as Kekule did, to the elements themselves. For although he too employs the mode of representation according to types, still this is elucidated by means of the so-called graphic formulae which are intended to express the relationships of the atoms. These formulae constituted a new mode by which to represent the constitution of compounds. They remained in use for some time, but they were again replaced, at a later date, by written formulae which approach to those introduced by Couper. The following are examples of these formulae : — CH3 CH3 CH,OH COOH i I I I CH2OH COOH CH2OH COOH Alcohol. Acetic acid. Glycol. Oxalic acid. The relations of the atoms in glycollic acid were furnished to Kekule by the method for its formation from chloracetic acid, which was discovered by himself. These relations can be CH2OH represented by the formula . Both glycollic acid and COOH lactic acid contain two typical hydrogen atoms ; that is, as Kekule now explains,3"* two atoms of hydrogen united to the carbon by means of the oxygen. These two atoms differ in their properties, inasmuch as the one behaves like the typical hydrogen of acetic acid, being influenced by two oxygen atoms, whilst the other plays a part resembling that of the typical hydrogen in alcohol. The solution of the difficulty was now supplied, since this view explained to the adherents of the doctrine of atomicity, all the chemical reactions of glycollic acid, 35 Kekule, Lehrbuch der Chemie. I, 130 and 174. LECTURE XIII.] HISTORY OF CHEMISTRY. 261 as well as its relation to glycol and also to acetic acid. A few years later, Kekule proved, by the action of hydrobromic acid on these acids,36 that they are converted by this reagent into the corresponding bromides just as readily as the alcohols are ; and thus the views which he had previously stated regarding the existence of " alcoholic hydrogen " in these compounds received new support. Perkin37 had already tried to confirm the alcoholic nature of glycollic and lactic acids, from the fact that sodium acts upon lactic ether with the evolution of hydrogen, and from the formation of ethereal compounds and the evolu- tion of hydrochloric acid when the acids are treated with acetyl chloride or succinyl chloride. These investigations of glycollic and of lactic acids are also highly important, because it was in connection with them that proof was furnished of the specially noteworthy circumstance that a two-fold function may be ascribed to one and the same substance, the two sets of properties in such a case being simply superadded. Kekule was also able to explain the fact that carbonic acid, which is homologous with glycollic acid, is a dibasic acid and forms salts with two atoms of metal. The formula of the OH hypothetical hydrate became CO ; both hydrogen atoms OH were equally influenced by the oxygen, and there was no reason for any difference between them.38 It must be specially pointed out here that Kolbe's formula for glycollic acid possesses a great similarity to the one that has just been given, when the signification is alone taken into account and not the form ; it was due, no doubt, to his some- what more complicated mode of writing the formula that Kolbe did not deduce from it all the consequences which it involved. On the whole, the advantages of Kolbe's way of regarding com- pounds were now about to become more manifest. In 1862, by the addition of hydrogen to acetone, Friedel obtained a 36 Annalen. 130, n. 37 Zeitschrift fiir Chemie. 4, 161. & Kekule, Lehrbuch. i, 739. 262 HISTORY OF CHEMISTRY. [LECTURE XIII. propyl alcohol 39 identical with that prepared by Berthelot from propylene.40 Kolbe 41 at once recognised this as the first re- presentative of the group of isomeric alcohols whose exist- ence he had foreseen.42 He assigned to it the formula S~* TT -\ 8 C.,O,HO, and maintained that the dissimilarity between it and Chancel's fermentation propyl alcohol 43 would be de- cided by an experiment involving its oxidation, since the new alcohol should thereby yield acetone. Friedel actually proved that it did so.44 Kolbe returns to this alcohol two years later.45 By a com- parison of the ammonia bases with the alcohols, he arrives at the conclusion that cases of isomerism must occur amongst the latter, quite similar to those amongst the former: — CTT "\ (~* TT A /"« TT "\ /"» TT ^ B**8 I V_/.>n-Q I l^i tt.r I l^.rlr I "H -N "H -C.,,O,HO HlN H -C.,,O,HO HJ H] HJ HJ Methylamine. Methyl carbinol. Ethylamine. Ethyl carbinol. CoHg-N cX~C,,O,HO CgHglN dH* ~C,,O,HO "H j "H I cinj QH:J Dimethylamine. Dimethyl carbinol. Trimethylamine. Trimethyl carbinol. Kolbe now extends his observations to the acids, and thinks that he can foresee the occurrence of isomerism in this class of substances also. Frankland, at a somewhat earlier date, prepared leucic acid by treating oxalic ether with zinc ethyl,40 and this interesting synthesis suggested the new ideas to Kolbe. He regards Frankland's acid (as Frankland himself had done) as diethyl-oxyacetic acid, and represents it by the formula C.2* C4H5 -C.,O2,O,HO. It corresponds to diethyl -acetic acid. 30 Comptes Rendus. 55, 53 ; Annalen. 124, 324. 40 Comptes Rendus. 44, 1350. 41 Zeitschrift fur Chemie. £, 687. 4- Compare p. 237. 43 Annalen. 87, 127. ^ Repert. de Chimie pure. 5, 247. ^ Zeitschrift fiir Chemie. 7, 30 ; Annalen. 132, 102. * Proc. Roy. Soc. 12, 396 ; Annalen. 126, 109. LECTURE XIII.] HISTORY OF CHEMISTRY. 263 Kolbe is also acquainted with a dimethyl-acetic acid, which he calls isobutyric acid, since it is different, according to him, from ordinary butyric acid : — / C.2H3\ C,-C,H3 V H C,.H-(C.,O.,)O,HO C,- C,H, C,O.,,O,HO V ( "H/ Butyric acid. Isobutyric acid. Kolbe assumes three isomeric substances having the formula of valerianic acid : trimethyl-acetic acid, methyl-ethyl-acetic acid, and propyl-acetic acid. Isomeric derivatives correspond to these compounds, such, for example, as oxyacids, amongst which Kolbe classes Stadeler's acetonic acid.47 These views were completely confirmed, and Kolbe's clever prediction thereby achieved a great triumph. FriedeFs acetone alcohol was the first compound representative of this class of substances that was prepared, and it, therefore, is of great importance. As the constitution of acetone had been settled by Freund's synthesis, there could scarcely be any doubt as to the formula of the new propyl alcohol ; and this formula was employed by Erlenmeyer 48 (after he had found that the alcohol was identical with the one prepared from glycerine49) in explaining the constitution of the triatomic alcohol. The discovery of the "hydrates" by Wurtz50 followed immediately after the discovery of isopropyl alcohol. Wurtz obtained these substances by treating the hydrocarbons of the ethylene series with hydriodic acid and silver oxide ; and he studied their properties particularly in the case of amylene hydrate, where he was able to recognise its difference from amyl alcohol. At first he looked upon this substance as a compound of the hydr^arbon with water, and represented it by the formula C5H]0, H.,O, a view which appeared to be warranted by its ready decomposition into these substances. At a later period he belte^l^mat the divergences from the normal alcohols could be explained on the assumption that the union of the 47 Annalen. in, 320. * Ibid. 139, ill. 49 Zeitschrift ftir Chemie. 7, 642. 50 Comptes Rendus. 55, 370 ; 56, 715, 793 5 57> 479- 264 HISTORY OF CHEMISTRY. [LECTURE XIII. hydrogen atom in the hydrates is different from (i.e., less intimate than) that in the substances isomeric with them.51 Kolbe looks on these substances as likewise belonging to the group of secondary alcohols, whose existence he had foreseen,0-' and he tries to prove this by means of an oxidation experiment, which, however, does not furnish any decisive result. Wurtz,53 who also carries out this experiment, obtains acetone as well as acetic acid. The question thus remained undecided 54 until, at a much later date (in 1878), Wischnegradzky was able to prove that amylene hydrate belongs to the group of tertiary alcohols.55 These had, however, been discovered much earlier (in 1863) by Butlerow56 as the products of a very remarkable, complex reaction ; but a large number of investigations by Butlerow and his pupils were required in order to establish conclusively the nature of these substances, and their relations to the other alcohols. With respect to isomerism in the fatty series, Erlenmeyer had obtained isobutyric ether in 1864, by the method proposed by Kolbe,57 but had not been able to discover any decided difference between it and ordinary butyric ether. Morkownikoff, however, established the difference later by a careful study of the salts ; 58 and proved, besides, that acetonic acid is identical with oxy-isobutyric acid.59 The neat syntheses by Frankland and Duppa are of consderable importance in connection with this question. These chemists succeeded in passing from oxalic acid to substances of the lactic acid series, and then they further converted these into the corresponding members of the acrylic acid series.60 Further, by means of aceto-acetic ether, which had been discovered by Geuther,01 they were able to introduce alcohol radicals into acetic acid, and thus to obtain homologues 51 Zeitschrift fiir Chemie. 7, 419. 5'2 Annalen. 132, 102. 53 Comptes Rendus: 58, 971. 54 Ibid. 66, 1179. 55 Annalen. 190, 328. 56 Repert. de Chimie pure. 5, 582 ; Bull. Soc. Chim. [2] 2, 106 ; Annalen. 144, I. 57 Zeitschrift fiir Chemie. 7, 642. M Annalen. 138, 361. 59 Zeitschrift fiir Chemie. 10, 434. w Annalen. 133, 80 ; 135, 25 ; 136, I ; 142, I. 61 Jahresbericht 1863, 323. LECTURE XIII.] HISTORY OF CHEMISTRY. 265 of the latter.02 Wislicenus afterwards followed up these reac- tions, and elucidated them more fully. Numerous syntheses were carried out by Wislicenus and his pupils according to this method, and thus our knowledge of the constitution of acids containing a number of carbon atoms was very considerably advanced. The fact established by Schorlemmer,03 that dimethyl is identical with ethyl hydride, had a distinct value in all con- siderations with respect to constitution ; and so also had the recognition of the identity of carbonic ethers containing two different alcohol radicals 64 (a point that was doubted at first °5). It was only after these matters had been settled that the similarity of the four valencies of carbon — the first thing necessary in order to inspire confidence in the " structural for- mulae " now so commonly employed — could be assumed. It will be apparent from the researches already mentioned that it is really a part of the business of scientific chemistry to explain the phenomena of isomerism. Cases of isomerism occur so frequently that even the clearest head would not be in a position to survey the facts, if these were simply enume- rated without any theoretical assumptions. Experience has shown, however, that graphic or structural formulae are ex- tremely valuable in explaining known cases of isomerism and in helping us to foresee new ones ; and it can thus be under- stood why the efforts of chemists were more and more directed towards establishing such formulae It is obvious that I cannot here take notice of all these efforts, but I must give an exposi- tion of the principle that is adopted in drawing conclusions from the reactions of substances with respect to their constitutions. This principle states that the mutual relations of the atoms remain unchanged during transformations, with the exception of those which are severed, and that with respect to the latter the atoms or groups that enter into the new combinations re- tw Annalen. 135, 217; 138, 204 and 328. ra Ibid. 131, 76; com- pare also Carius, ibid. 131, 173 ; and Schoyen, ibid. 130, 233. °4 Rose, Annalen. 205, 227. tir' J. pr. Chem. [2] 22, 353. 266 HISTORY OF CHEMISTRY. [LECTURE XIII. establish similar relations. I do not think I am mistaken in regarding this fundamental principle as a new form of Laurent's law of substitution.66 It is a generalising of this law, but the law has acquired, at the same time, another meaning, in so far that chemists now no longer desire to determine the arrangement of the atoms in space but only their relations to one another. Unfortunately, no general proof of the principle has been fur- nished, and experiments have not even been carried out which aim at this. Its accuracy, which is certainly not beyond doubt, is merely assumed because the conclusions drawn from it have repeatedly yielded concordant results; that is, because these conclusions led to identical formulae for the same substance whichever mode of formation was considered. Nevertheless, this concordance is not always met with. There are many cases known where the constitution deduced from one mode of formation does not correspond to the formula which may be derived by starting from another mode of formation, or from the decomposition products.67 In such cases we are obliged to assume the conversion of the substance into an isomeric modification, in one of the reactions that have taken place ; that is, we must hold that the principle stated above does not here apply, and that the atoms remaining in the molecule have changed their mutual relations during one of the reactions. Cases of this kind deserve attention. They are well calculated to shake our faith in the accuracy of the fundamental principle, even although an attempt has been made to regard them as two-fold reactions, and in agreement with this principle. A special interest attaches to those investigations that define the conditions under which such isomeric changes — the so-called migration or wandering of the atoms — take place within the molecule, and to those actually designed to elucidate this particular matter. As examples of these, Hofmann's investigation of the conversion 66 Compare pp. 143-144. 67 Compare, for example, Carius, Annalen. 131, 172 ; Tollens, ibid. 137, 311 ; P'riedel and Ladenburg, ibid. 145, 190; Linnemann and Siersch, ibid. 144, 137; Butlerow and Ossokin, ibid. 145, 257 ; Simpson, ibid. 145, 373 ; Erlenmeyer, ibid. 145, 365, etc. LECTURE xin.] HISTORY OF CHEMISTRY. 267 of the methyl anilines into homologues of aniline 6S and Demole's examination of the spontaneous oxidation of ethylene derivatives,69 may be mentioned here. Those phenomena of isomerism, however, which cannot be expressed or represented by means of the ordinary formulae, are of still greater importance. Examples of this nature have long been known, and some of them were minutely studied at an early period. More recently (after the recognition of the importance of the matter) an attempt has been made to introduce a special method of explaining them, which is, no doubt, connected with the theory of valency, but is also a further development and an extension of that theory. We must here enter upon a more particular account of this matter. The discovery of racemic acid, isomeric with tartaric acid, has already been mentioned in Lecture VII. (p. 118). The recognition of the relations between these two acids forms the subject of an investigation by Pasteur which is of fundamental importance for the subject now to be considered.70 Pasteur showed that there are four isomeric tartaric acids, viz. : racemic acid, inactive tartaric acid, and right and left rotating tartaric acids. He showed, moreover, that the two latter acids crystallise in similar, but in oppositely built-up (enantiomorph) forms j that they both deviate a ray of polarised light through equal angles, but in opposite senses ; and that when mixed in equal quantities they yield optically inactive racemic acid. Further, he succeeded in decomposing racemic acid again into the two optically active tartaric acids, by three different methods :— 1. By preparing and crystallising the sodium-ammonium salt ; when two enantiomorph varieties were obtained. These, after having been separated and decomposed, yielded the two tartaric acids. 2. By preparing the cinchonicine and quinicine salts. In 68 Berichte. 4, 742 ; 5, 704, etc. 69 Ibid. II, 315, 1302 and 1307. 70 Ann. Chim. [3] 24, 442 ; 28, 56 ; 38, 437 ; compare also Pasteur, Recherches sur la dissymetrie moleculaire des produits organiques natu- rels. Le9ons de chimie, Paris 1861 ; Alembic Club Reprints, No. 14. 268 HISTORY OF CHEMISTRY. [LECTURE XIII. the case of the former the salt of left tartaric acid crystallised first ; in that of the latter, the salt of right tartaric acid crystallised first. 3. By treating a solution of acid ammonium racemate with spores of Penicillium Glaucum, by which means the salt of left tartaric acid alone remains in the solution after the develop- ment of the fungus. Pasteur prepared inactive tartaric acid by heating cinchonine tartrate ; and Dessaignes was able to show that this acid was partially reconverted into racemic acid by heating it to 2oo°.T1 Facts of a similar kind have been observed in the cases of various other substances, such, for example, as the glucoses, the terpenes, amyl alcohol, aspartic acid, etc. Exactly the same relations that are observed in the case of tartaric acid occur also in that of mandelic acid, according to experiments by Lewkowitsch.72 In all these cases of isomerism it is in their physical properties that the corresponding substances differ from each other ; and on this account Carius 73 introduced for such cases the designation physical isomerism. In 1874 Le Bel74 and, shortly after him, van 't Hoff75 en- deavoured to explain 'these facts also upon the theory of atomicity, having first adopted the view that a substance only possesses optical activity when its molecule contains an asymmetric carbon atom ; that is, when a carbon atom is present in it which has its four valencies satisfied by union with four atoms or groups all different from one another. This view is warranted by the facts, in so far that all optically active substances known up to the present contain at least one asymmetric carbon atom. It must be stated, however, that by no means all substances which contain asymmetric carbon atoms possess rotating power, and that the view stated above cannot, therefore, be turned about and generalised as if this 71 Annalen. 136, 212. 7- Berichte. 16, 1565 and 2721. 73 Annalen. 126, 214 ; 133, 130. 74 Bull. Soc. Chim. [2] 22, 337. 75 Ibid. [2] 23, 295 ; compare also La chimie dans 1'espace, Rotterdam, 1875 ; Chemistry in Space, translated and edited by Marsh. Oxford 1891. LECTURE XIII.] HISTORY OF CHEMISTRY. 269 were so. Van 't Hoff has endeavoured to render the matter clear by means of a geometrical conception as to the arrange- ment of the atoms in space. This cannot, however, be more fully entered into here. A considerable time ago, Rochleder 7G pointed out that substances belonging to one particular class, very easily undergo isomeric changes ; and he called these substances defective because they are produced by the separation of certain atoms from saturated substances. They are now commonly called unsaturated compounds, and we shall consider them particu- larly, as their study is a subject of great interest. In his paper on the theory of organic compounds, Couper 77 ascribed to carbon the capacity of bringing sometimes two, and sometimes all four of its units of affinity into play ; hence it was not difficult for him to explain the existence of such compounds as carbonic oxide, ethylene, etc. Amongst others, Wurtz 7S and Kolbe 79 fell in with this view. The latter derives the unsaturated hydrocarbons from the type of carbonic oxide, assuming in all of these substances one or several carbon atoms which are active with two affinities. He writes C2O2 TT TT TT Carbonic oxide, C2p TT Ethylene, C2p TT Propylene, C2p H or C2 . C2H2 Acetylene. Kekule at first tried to explain the unsaturated substances by the assumption of a denser arrangement of the carbon atoms,80 but afterwards, in his admirable and important investi- gations of organic acids 81 he appears to have been of opinion that, in these substances, the affinities of the carbon are not fully satisfied, and that they contain free affinities or blanks. This assumption became more probable both from Kekule's own experiments and from those of Carius,82 in accordance 76 Ber. Wein. Akad. II, 852 ; 12, 727 ; also ibid. 49 (second part), 115. 77 Ann. Chim, [3] 53, 469 ; compare p. 254. 78 Le5ons de philosophic chimique. 136. 79 Kolhe, Lehrbuch der organischen Chemie. I, 738 ; 2, 576. 80 Kekule, Lehrbuch der organischen Chemie. I, 166. 81 Annalen. 117, 120; Supplementband I, 129, 338; Supplementband 2, 85 ; 130, i. 8- Ibid. 124, 265 ; 126, 195 ; 129, 167, 270 HISTORY OF CHEMISTRY. [LECTURE XIII. with which the substances can combine with hydrogen, chlorine, hypochlorous acid, etc. The capacity for entering into direct additions thus became a characteristic of the group ; but it cannot be said to be really distinctive, since some substances which are classed as saturated also possess this capacity. As examples of the latter substances the aldehydes and ketones in particular may be instanced, and these are substances which contain oxygen wholly united to carbon. In explanation of the facts, the assumption is made regarding these compounds that, by addition, the group (C = O)" passes into (C — O)"" ; that is to say, a diatomic radical becomes a tetratomic one. Later experiments of a very detailed char- acter on the unsaturated acids, by Fittig, have led to the confirmation of the view mentioned above ; 83 that is, they have shown that the facts are best accounted for when blanks, or bivalent carbon atoms, are assumed in some compounds at least. That it is not possible to avoid some assumption of this kind is shown by carbonic oxide and by the group of isonitriles or carbylamines, discovered almost simultaneously by Hofmann84 and by Gamier.85 The latter interesting substances are obtained by treatment of the amines with chloroform and by the action of the alkyl "iodides on silver cyanide. They are isomeric with the nitriles, and their constitution cannot be f R represented otherwise than by the formula N p, first proposed by Gautier,80 in which R stands for a monatomic alcohol radical. If the nitrogen is assumed to be trivalent,87 the carbon then appears as bivalent or unsaturated. There is, besides, a class of unsaturated substances, in which, following Kekule's lead, a more intimate union of the carbon atoms is quite generally assumed. I refer to the aromatic com- pounds. Under this heading a number of substances which stand in a close chemical relationship to certain strongly smell- ing oils were formerly grouped together. 83 Annalen. 188, 95. 84 Ibid. 144, 114 ; 146, 107. 8r> Comptes Rendus. 65. 468 ; Annalen. 146, 119. 86 Comptes Rendus. 65, 901. 87 Quinquivalent nitrogen is referred to in Lecture 15. LECTURE XIII.] HISTORY OF CHEMISTRY. 271 Kekule showed that all these substances may be regarded as derivatives of benzene, and that their chemical nature is dominated by the constitution of this hydrocarbon.88 A large number of earlier observations told in favour of this view, but some synthetical investigations which had been carried out a short time previously by Fittig in conjunction with Tollens so and others-10 were also of importance. These chemists em- ployed a method which originated with Wurtz ; 91 that is, they treated mixtures of the alkyl iodides and the bromine substi- tution products of aromatic hydrocarbons with sodium, whereby they succeeded in preparing homologues of the hydrocarbons in question. They were thus able to show that methyl-benzene, obtained from bromo-benzene and methyl iodide, is identical with toluene, but that ethyl-benzene is different from xylene, which, however, approaches very closely in its character to methyl-toluene or dimethyl-benzene. I do not need to enter more fully here into the further results of these interesting researches, as they were only obtained subsequent to the publi- cation of Kekule's paper, and in this paper they were partially foreseen. On the other hand, some of the results of Beilstein's researches were of fundamental importance with respect to the theoretical investigations now to be discussed. Of this char- acter was the proof, carried out in conjunction with Reichen- bach,92 that the so-called salylic acid, which was regarded as a benzene-carbonic acid, isomeric with benzoic acid,93 was simply impure benzoic acid ; and so was the fact that the chloro- benzoic acids prepared up to that period, could be reduced in number to three.94 Benzene, as the fundamental substance in the aromatic group attains to quite a special significance in consequence of the views of Kekule, and the latter therefore makes a special study of its constitution. I shall deal with this matter in the next lecture. 88 Annalen. 137, 129. 89 Ibid. 131, 303. 90 Ibid. 136, 303, etc. 91 Ann. Chim. [3] 44, 275. *2 Annalen. 132, 309. 93 Kolbe and Lautemann, ibid. 115, 183; Kekule, ibid. 117, 158; Griess, ibid. 117, 34 ; Cannizzaro, ibid.'Supplementband I, 274. 94 Beilstein and Schlun, ibid. 133, 239. LECTURE XIV. THEORY OF THE AROMATIC COMPOUNDS— DETERMINATION OF POSITION OF SUBSTITUTED ATOMS OR GROUPS — QUINONES — ARTIFICIAL DYES— RING COMPOUNDS— CONSTITUTION OF THE ALKALOIDS- SYNTHESES — CONDENSATION PROCESSES. TAKING the quadrivalence of carbon as his starting-point, Kekule points out that in the fatty compounds the carbon atoms are linked together by one valency of each.1 In the case of benzene, the next simplest assumption is made, in accordance with which the carbon atoms are linked together by one and by two valencies alternately, so as to form a closed chain or ring. Of the twenty-four affinities of the six carbon atoms, eighteen are employed in linking carbon to carbon, thus :— 6 6 —.4+ —.2 - 18 2* 2 Six valencies then remain which are satisfied by the six hydrogen atoms of the benzene. Hence, according to Kekule, benzene may be represented by means of a regular hexagon whose sides are composed of single and of double lines alter- nately, the CH groups occupying the corners. This conception is designed to illustrate, in the first place, the relatively great stability of benzene as compared with the hydrocarbons of the fatty series, which consist of open carbon chains with, for the most part, singly linked carbon atoms. It further illustrates the fact, which is of such great importance with respect to the aromatic compounds, that the six hydrogen atoms of benzene are symmetrically disposed in the molecule —that is, that they are identical in function. 1 Bull. Soc. Chim. [2] 3, 104 ; Annalen. 137, 129 ; Lehrbuch der organischen Chemie, 2, 493. LECTURE XIV.] HISTORY OF CHEMISTRY. 273 The aromatic compounds are obtained by the replacement of these hydrogen atoms in benzene. But it follows from the equivalence of all six hydrogen atoms that when only one of them is replaced, it must be a matter of indifference which of them it is ; or, in other words, only one variety of any of the mono-substitution products of benzene can exist. A view of this kind was only possible after it had been shown that methyl-benzene is identical with toluene, and benzoic acid with salylic acid (compare p. 271). When two or more hydrogen atoms in benzene are replaced, Kekule's hypothesis predicts the existence of numerous isomers, occasioned by differences in the relative positions of the atoms or groups that enter the molecule ; and the number of these isomers can be determined. Thus there are three isomers possible when two hydrogen atoms in benzene are replaced by other atoms or by radicals ; and it is immaterial whether the atoms or radicals which enter are identical or different. Of tri-substitution derivatives of benzene there are three possible isomeric forms when the three substituting atoms or groups are the same, but six when two of them are different from the third. Further the hypothesis foretells the existence of three isomeric tetra-substituted benzene derivatives but only one penta- and one hexa-substituted derivative when all the sub- stituting atoms or groups, in each case, are the same. In accordance with the hypothesis, by the replacement of one or more of the hydrogen atoms in benzene by any given element or group of atoms, twelve substances can be obtained, and this has been actually accomplished in at least one case. Thus Beilstein was able to show that exactly twelve chlorinated ben- zenes exist,2 after it had been proved that the alleged existence of two isomeric pentachlorobenzenes 3 was a mistake. It follows, moreover, from the constitution of benzene, that ethylbenzene must be different from the three possible dimethyl- benzenes ; and, further, that by the action of chlorine or of 2 Beilstein and Kurbatow, Annalen. 192, 228. u Ladenburg, ibid. 172, 331. S 274 HISTORY OF CHEMISTRY. [LECTURE xiv. bromine two different classes of substitution products should be obtainable from toluene, these classes being characterised by the fact that in the one the halogen replaces a hydrogen atom of the benzene (nucleus), and in the other a hydrogen atom of the methyl (side-chain). Differences of this kind were actually observed,4 and Beilstein5 showed that substances belong- ing either to the one class or to the other are produced according to whether chlorine acts in the cold or at the boiling temperature. The members of the first of these groups of chlorine compounds (of which, as di-substitution products of benzene, three isomers exist) do not permit any exchange of their chlorine for iodine, for cyanogen, or for hydroxyl or other groups containing oxygen ; whereas the chlorine derivative of the other group (which, as a mono-substitution product is the only chlorine representative of the group) behaves like the chloride of an alcohol radical, and can be converted, just as easily as chlorides of this kind can, into an alcohol, an ether, etc. The two formulae— CtiH4Cl(CH3) C0H5CH,C1 Chlorotoluenes. Benzyl chloride. indicate these differences, which arise, according to Kekule, from the fact that the chlorine atom of the chlorotoluenes stands in intimate relation to the carbon (being almost entirely surrounded by it), whereas in benzyl chloride it is combined in a manner similar to the halogen of the alkyl chlorides. An explanation of an exactly similar kind is now furnished for the essentially different behaviour of the phenols and of the aromatic alcohols. Whereas in the former the hydroxyl group replaces a hydrogen atom of benzene, in benzyl alcohol the replaced hydrogen belongs to the methyl group : — C6H4(OH)CH3 CCH5CH2OH Cresols. Benzyl alcohol. On oxidation, the latter alone behaves as a primary alcohol 4 Fittig, Annalen. 136, 301 ; Kekule, ibi-L 137, 192. 5 Beilstein and Gcitner, ibid. 139, 331. LECTURE XIV.] HISTORY OF CHEMISTRY. 275 and yields an aldehyde and an acid, whilst the ethers of the former are converted into alkyl-oxybenzoic acids C(5H4(OR) CO2H.6 Kekule's views concerning the oxidation of aromatic hydro- carbons into acids are very important. " It may be said in general that the alcohol residues (methyl, ethyl, etc.) attached as side-chains to the nucleus C0 are converted, by sufficiently vigorous oxidation, into the group CO.2H. The oxidation products always contain, therefore, just as many side-chains as the substances from which they have been produced ..... When the reactions are more moderate it is possible, in the case of those derivatives of benzene which contain two or more alcohol radicals, to restrict the action to the formation of intermediate products ; thus, one alcohol radical only is oxidised in the first place, while the other remains unchanged. Dimethylbenzene (xylene) in this way yields toluylic acid. . . . On more vigorous oxidation the toluylic acid is then converted into terephthalic acid." * ~ ** /~» T Xylene. Toluylic acid. Terephthalic acid. It is worth while pointing out, lastly, that Kekule in the further elaboration of his views, cleared up the constitution of the azo-compounds 7 discovered by Mitscherlich, and, more particularly, that of the diazo-compounds discovered and minutely investigated by Griess ; 8 besides showing the con- nection existing between these groups.9 These researches upon the aromatic compounds exercised an immense influence upon chemistry. The investigation of these substances, which, up to that period, had been rather neglected, was by many chemists almost exclusively worked at during the succeeding ten years. The countless examples of isomerism which previously rendered this branch so difficult to 6 Korner, Zeitschrift fiir Chemie. II, 326. 7 Annalen. 12, 311. 8 Ibid. 106, 123 ; 109, 286 ; 113, 334 ; 117, i ; Supplementband I, 100 ; 121, 257, etc. ; compare also Phil. Trans. 1864, 667, etc. 9 Lehrbuch, 2, 703- 276 HISTORY OF CHEMISTRY. [LECTURE XIV. investigate (since it was only possible for a few persons to obtain a real grasp of the facts) increased the attractiveness of the investigations, now that a simple explanation of these phenomena was forthcoming. And, what is of the greatest importance, Kekule's views were confirmed by them in the most complete manner, and did not require alteration in any essential particulars — isolated statements at variance with them always proving capable of very early refutation as incorrect. Moreover, these hypotheses were widened to a considerable extent and perfected, as a consequence of the immense number of facts afterwards discovered. The problem of determining the positions of the substitut- ing atoms and groups deserves to be mentioned here first. Merely referred to by Kekule,10 it was fully solved afterwards. By determining the position in the aromatic series, we understand the ascertaining of the relations to one another and to the carbon nucleus of the atoms or groups which replace the hydrogen in benzene. Obviously the question can, at the earliest, only possess any significance in the case of the di-sub- stitution products. The three isomers here possible, according to Kekule, have had the distinguishing prefixes ortho-, meta- and para- attached to their names, and the question at once arises how these are to be conceived as regards their constitu- tion. The first step in this connection was taken by Baeyer,11 after it had been proved by Fittig12 that mesitylene is a trimethylbenzene. From the mode of its formation, Baeyer draws the conclusion that the three methyl groups are symmetri- cally arranged with respect to the benzene nucleus ; that is to say, that mesitylene and isophthalic acid are meta-compounds. This hypothesis was afterwards proved 13 by an accurate investi- gation of the substitution products of mesitylene. Grabe was then able to show by detailed discussions and experiments as to the nature of naphthalene u that this substance, and consequently, phthalic acid also, must be regarded as ortho- 10 Annalen. 137, 174. u Ibid. 140, 306. 12 Zeitschrift flir Chemic. 9, 518. ly Ladenburg, Annalen. 179, 163. 14 Ibid. 149, 22. LECTURE XIV.] HISTORY OF CHEMISTRY. 277 compounds. Finally, it was pointed out by Ladenburg,15 taking into account the experiments of Hiibner and Petermann,16 that terephthalic acid and para-oxybenzoic acid belong to the para-series. A very neat and original idea with respect to the solution of this problem originated with Korner,17 who showed that, by the introduction of a third atomic group into the di-substitution products containing two similar substituting atoms or groups, three isomeric tri-substitution products are possible when the original substance belongs to the meta-series, or two when it is an ortho-compound, while in the case of a para-derivative, only a single tri-substitution product is possible. By employing this method he determined the constitution of the dibromo-benzenes, and Griess18 determined that of the phenylene diamines. After the constitution had thus been determined in some compounds, it was still necessary to establish the relations between these and other compounds by means of simple reac- tions, so as to have the problem solved in the cases of all the doubly substituted benzenes. Not only has this been quite possible, but the position of the substituting groups has also been ascertained in the higher substitution products. In the whole of these often very extensive investigations, which were only practicable by the co-operation of many hands, the reactions of Griess (see above) rendered very important services. It is likewise of considerable importance for the theory of aromatic compounds that, starting from the quadrivalence of carbon, and a series of accurately determined facts, it has proved possible to establish the two fundamental principles as to the constitution of benzene, as follow: — i. The equivalence of the hydrogen atoms of benzene ; and 2. The symmetry of two pairs of hydrogen atoms in benzene, with respect to the third pair of hydrogen atoms.19 It must further be pointed out here that a prolonged con- 15 Berichte. 2, 140. ]G Annalen. 149, 129. 17 Gazzetta Chimica Italiana. 4, 305 ; Journ. Chem. Soc. 29, 204. ]S Berichte. 7, 1226. 19 Ladenburg, Theorie der aromatischen Verbindungen, Braunschweig 1876; Berichte. 10, 1224; Wroblewsky, Annalen. 192, 196. 278 HISTORY OF CHEMISTRY. [LECTURE XIV. troversy arose concerning the formula of benzene (that is, as to the mutual linkings of the carbon atoms which it contains) after attention had been drawn to the fact that Kekule's for- mula does not altogether take account of the requirements involved in the two principles stated above.'20 From this con- troversy it appeared that only the so-called prism formula can give a clear idea of the bearings of isomerism in the aromatic series, since it furnishes likewise an accurate expression for the thermal relations, according to Thomsen, and for the mole- cular volume of benzene and its derivatives, according to R. Schiff.21 Nevertheless Kekule's hexagon formula has been generally retained, because it is superior to the other formula in many respects. Amongst the notable researches which were instigated by Kekule's investigations, I only enter into detail here respecting a single one which may probably be looked upon as the most important amongst them. I refer to Grabe's examination of the quinones. Kekule propounded a peculiar view respecting quinone,22 a substance which had been discovered by Woskresensky.-3 This substance was supposed to consist of an open chain of six carbon atoms, which were joined to one another by single and double linking alternately. In opposition to this view, Grabe 24 advanced another, in accordance with which quinone is a benzene derivative in which two hydrogen atoms are re- placed by two oxygen atoms ; and these latter are further united to each other. He bases this view especially upon the already well-known relations of quinone to hydroquinone, and upon the conversion of chloranil into hexachlorobenzene by means of phosphorus pentachloride. These grounds were so convincing that Grabe's view was generally adopted, even 20 Ladenburg, Berichte. 2, 140. 21 Ibid. 13, 1808; see also Thomsen, Thermochemische Untersuchungen. 4; also Schiff, Annalen. 220, 303. According to Schroder (Wiedem. Ann. 15, 667) this also holds for the molecular refraction ; whereas, according to Briihl (Annalen. 200, 229), the opposite is the case. ~ Annalen. 137, 134. 23 Ibid. 27, 268. >24 Ibid. 146, i. LECTURE XIV.] HISTORY OF CHEMISTRY. 279 although it appeared soon after that quinone did not belong to the ortho-compounds, as Grabe supposed, but to the para-compounds.25 Grabe afterwards studied other quinones also, and so arrived at the investigation of alizarine, the nature of which as a quinone he desired to establish. In conjunction with Liebermann, and by making use of a method discovered by Baeyer,26 he showed that alizarine was not, as was then supposed, a naphthalene derivative, but that it was derived from anthracene ; 2T that it was a quinone ; and that, in par- ticular, it was a dioxy-anthraquinone. These chemists after- wards accomplished the synthesis of this valuable colouring matter,28 which was at .once prepared technically according to a method elaborated by Grabe, Liebermann, and Caro ; 29 thus leading to one of the most extensive industries of the present time. It may be stated generally that the theory of the aromatic compounds had a great influence in technology and especially in that of dyes. Although the aniline colour industry was called into existence quite independently of these investigations (especially by Hofmann's comprehensive researches on aniline, and the bases homologous with it), and although the first aniline colours had been discovered and turned to account long prior to the publication of Kekule's celebrated paper — mauveine by Perkin 30 as early as 1 856, and fuchsine by Verguin31 in 1859, after it had been previously observed by Natanson,32 Hof- mann,33 and others — still its further development is intimately connected with the more accurate insight into the constitution of the aromatic compounds. With respect to this, it is only necessary to recall the discovery of orthotoluidine by Rosen- stiehl,34 and the explanation of the chemical nature of rosaniline, 23 Petersen, Berichte. 6, 368 and 400. 2G Annalen. 140, 295. 27 Berichte. I, 49. 28 Annalen. Supplementband 7, 257 ; Berichte. 2, 14. 29 Ibid. 3, 359. 30 Perkin, Zeitschrift fitr Chemie. 4, 700; Annalen. 131, 201. 31 Repert. de Chimie appliquee. 2, 114, 299; compare also Dingl. Polyt. Journ. 154, 235, 397. 3'2 Annalen. 98, 297. s3 Jahresbericht 1858, 351. :u Zeitschrift fur Chemie. u, 557; 12, 189-190. 280 HISTORY OF CHEMISTRY. [LECTURE XIV. which was eventually furnished by E. and O. Fischer,35 the way having been prepared by Hofmann.30 The manufacture of other classes of dyes has also arisen independently of this theory, although no doubt advanced by means of it. Examples of these substances are the phenol dyes, of which the first representative is rosolic acid, discovered by Kolbe and Schmitt 3" and simultaneously J. Persoz ; 3S and this group was greatly enlarged by the phthale'ines, discovered and studied by Baeyer.39 Other examples are the azo-dyes, which are, almost without exception, connected with the im- portant researches ot Griess. The influence which Kekule's conception of the aromatic compounds exercised upon the views concerning the more complicated hydrocarbons is much more direct. Erlenmeyer,40 in an interesting paper on aromatic acids, which contains a criticism of Kekule's views, assigns to naph- thalene, C10HS, the formula : — H H H H I I ! = c—c=c— c = c '- I i -H— C C— H II II H— C— C— H In accordance with this formula, naphthalene could be con- ceived as composed of two benzene hexagons with two carbon atoms common to both of them. Grabe rendered this concep- tion very probable by means of experimental investigations and theoretical considerations.41 Aronheim's synthesis of naphtha- lene from phenyl butylene 42 also tells in support of it, and so, especially, does Fittig's synthesis of a-naphthol 43 (the hexagon 35 Annalen. 194, 242. M J. pr. Chem. 87, 226 ; Jahresbericht 1863, 417; 1864, 819; Annalen. 132, 160 and 289. ^ Ibid. 119, 169. 38 French Patent, 2ist July 1862. <«. V^GIT.Oy~( -WJ Hence, it can scarcely be doubted that the other hydro- 49 Berichte. 5, 908. r>0 Ibid. 7, 972; 8, 152. 51 Ibid. 7, 1785. 52 Annalen. Supplementband 5, 371. 53 Berichte. 6, 187; Annalen. 193, 134. 54 Berichte. 10, 2022. 55 Fittig and Gebhard, ibid. IO, 2143; Annalen. 193, 142; Fittig and Liepmann, ibid. 2OO, I. 56 Berichte. 12, 1078. r'7 Ibid. 18, 1024 and 1750; Annalen. 229, 102, LECTURE XIV.] HISTORY OF CHEMISTRY. 283 carbons with high molecular weights, which are not so fully examined as yet (such as pyrene, picene, etc.), may be derived from benzene in an analogous manner. Some other investigations in which a relationship can be recognised between certain compounds containing nitrogen — especially the alkaloids — and benzene, appear to be probably still more important than the foregoing. This branch of the subject, only opened up a few years ago, already presents so many remarkable results that it cannot be omitted here. The analogy of the formulae of benzene, C0H0, and naph- thalene, C10H8, on the one hand, with those of pyridine, C5Hf)N, and quinoline, C9H7N, on the other, admitted of the hypothesis that the latter compounds might be derived from the former by the replacement in each of a CH group by N. and consequently the following formulae were advanced for pyridine and quinoline. CH N Pyridine. Quinoline. This view was made known by means of private communi- cations by Korner, and is usually known as Korner's hypothesis. It was first published by Dewar/"s A large number of facts can now be adduced in support of the view, and the more important of these may be mentioned here. Anderson, the discoverer of pyridine, had already found in animal oil, besides pyridine, a number of homologous bases.59 The further examination of bone tar has yielded, as yet, only methyl pyridines,00 just as methyl benzenes only are contained 58 Chem. News. 23, 38; Zeitschrift fiir Chemie. 14, 117. r>9 Ann- alen. 60, 86 ; 70, 32 ; 75, 80 ; 80, 44 ; 94, 358 ; see also Unverdorben, Pogg. Ann. II, 59. 60 Weidel, Berichte. 12, 1989; Ladenburg and Roth, ibid. 18, 47 and 913. 284 HISTORY OF CHEMISTRY. [LECTURE XIV. in coal tar. Ethyl and propyl pyridines are already known, however.01 On oxidation, these bases behave exactly like the alkyl derivatives of benzene ; that is to say, every side-chain, by sufficiently energetic oxidation, yields a CO.2H group ; so that, in this case also, conclusions may be drawn as to the number of side-chains in the base oxidised, from the basicity of the acid produced. The isomerisms amongst the derivatives of pyridine are far more complicated than in those of benzene, since the hydrogen atoms are not similarly related to the pyridine nucleus and three different mono-substitution products must exist, as was probably first pointed out by Weidel.0'2 This conclusion is like- wise confirmed by experiment, as three mono-carbonic acids,03 three methyl-pyridines °4 and three ethyl-pyridines °5 are known. Determinations of the positions of the substituting atoms or groups in the pyridine series have been accomplished with tolerable certainty by Skraup.66 As additional supports for the pyridine formula, there may also be adduced the synthesis of pyridine by Ramsay °7 (which is upon the same lines as the famous synthesis of benzene from acetylene by Berthelot 6S) as well as the synthesis of pyridine derivatives.09 Finally the conversion of pyridine derivatives into benzene derivatives 70 is of importance in this connection. The formula for quinoline is also based upon numerous syntheses, of which that by Konigs 71 may be mentioned here as the first. This was followed by that of Baeyer 7- and then by 61 Williams, Jahresbericht 1855, 549; 1864, 437; Cahours and Etard, Comptes Rendus. 92, 1079; Ladenburg, Berichte. 16, 2059 ; 17, 772 and 1 121 ; 18, 1587 ; Hofmann, ibid. 17, 825. 62 Ibid. 12, 2012. Ki Huber, Annalen. 141, 271 ; Berichte. 3, 849 ; Weidel, ibid. 12, 1989 ; Skraup, ibid. 12, 2331. w Weidel, ibid. 12, 1989; Behrmann and Hofmann, ibid. 17, 2681. ^ Wischnegradsky, ibid. 12, 1480; Ladenburg, ibid. 16, 2059. 6(! Skraup and Cobenzl, Monatshefte. 4, 450 ; compare also Ladenburg, Berichte. 18, 2967. 67 Ibid. 10, 736. tis Ann. Chim. [4] 9, 469. 0!) Hantzsch, Annalen. 215, I ; Pechmann and Welsh, Ber- ichte. 17, 2384 ; Behrmann and Hofmann, ibid; 17, 2681. 7() Ladenburg, ibid. 16, 2059. n Ibid. 12, 453. ~'2 Ibid. 12, 460. LECTURE XIV.] HISTORY OF CHEMISTRY. 285 the one of Skraup78 which has become so important for the whole group. This latter synthesis depends upon the carrying out of one of the ideas indicated by Grabe.74 Quinoline, also, is the starting-point for a large number of compounds, which are formed from it and can be converted into it just as benzene passes into aromatic compounds and can be obtained from them. The relations between pyridine and quinoline, which are quite analogous to those between benzene and naphthalene, are also worthy of mention. In the same way that the latter is converted by oxidation into benzene-ortho-dicarbonic acid (phthalic acid75), so quinoline, according to Hoogewerff and Van Dorp,70 is converted by oxidation into an ortho- (a/?-) pyridine-dicarbonic acid. But what is of the greatest significance is the fact that the most important alkaloids are derivatives of pyridine and of quinoline (or of their hydrogenised derivatives) in the same way that the aromatic oils are derivatives of benzene. The first fact bearing upon this relationship was found out by Gerhardt in 1842, when he discovered quinoline as a product of the decomposition of quinine, of cinchonine, and of strych- nine.77 Huber obtained in 1867, by the. oxidation of nicotine,78 an acid C(;H5NO., which he recognised, three years later, as pyridine-carbonic acid ~9 — a fact which was at first disputed and then confirmed.80 Piperidine, which was discovered by Wert- heim and Rochleder by the decomposition of piperine,81 and the correct formula of which was established by Cahours 8'2 and by Anderson,83 was regarded by Hofmann as a hydrogen addition product of pyridine,84 a view which was proved to be correct by Konigs and others.85 73 Monatshefte. I, 317 ; 2, 141. 74 Annalen. 201, 333. 73 Laurent, ibid. 19, 38; 41, 98. 76 Berichte. 12, 747. 77 Annalen. 42, 310; 44, 279. 78 Ibid. 141, 271. 79 Berichte. 3, 849. 80 Weidel, Annalen. 145, 328; and Laiblin, ibid. 196, 129. 81 Ibid. 54, 254 ; 70, 58. 8'2 Ibid. 84, 342. 83 Ibid. 84, 345. 84 Berichte. 12, 984. 8r' Ibid. 12, 2341; Schotlen, ibid. 15, 421; Hofmann, ibid. 16, 586; Ladenburg, ibid. 17, 156, 388; Ladenburg and Roth, 17, 513. 286 HISTORY OF CHEMISTRY. [LECTURE XIV. These and other facts, which seemed to place beyond doubt the relations of several natural bases to pyridine,86 led Wischnegradsky to the opinion stated above regarding the con- stitution of the alkaloids ; 8T and this was more fully discussed and established, a year later, by Konigs.88 Since then, this view has gained ground more and more ; especially as a series of facts have been discovered in support of it. Thus Weidel obtained a pyridine-tricarbonic acid by the oxidation of ber- berine ; s9 Gerichten was able to prepare pyridine-dicarbonic acid from narcotine,90 and Ladenburg dibromo-pyridine from atropine;91 while Hofmann converted coniine into propyl- pyridine.92 The results which attended this view of pyridine and of quinoline, and the recognition which they met with, led to the introduction of a similar view concerning many other sub- stances. In the first place it is necessary to consider the formula which was assigned as early as 1869, by Baeyer and Emmerling,93 to indol,'14 the starting-point for most of the indigo derivatives : — According to this formula indol is represented as a double nucleus resembling naphthalene and quinoline. This mode of representing it acquired greater significance when Baeyer and Emmerling,95 somewhat later, regarded pyrrol also as a "ring." The same relation was now assumed between pyrrol and 86 Weidel, Annalen. 173, 76; Ramsay and Dobbie, Berichte. II, 324. 87 Ibid. 12, 1506; compare also Ladenburg, ibid. 12, 947. ^ Studien uber die Alkaloide, Munich 1880. 8r (discovered more recently by V. Meyer), which is looked upon as thiofurfuran, and which has already become of great importance on account of its numerous derivatives. In the case of thiophen the resemblance exhibited by it and its derivatives to benzene and the benzene derivatives is particularly noteworthy. Carbazol,98 discovered by Fritzsche, may also be mentioned here — a substance which Grabe91' regards as fluorene, in which CH., is replaced by NH, thus: j NH. Further, there is C0H acridine,100 found in crude anthracene by Gra'be and Caro, which is regarded as a derivative of anthracene 1G1 or of phen- 96 Berichte. 3, 90. 97 Ibid. 16, 1465. 98 J. pr. Chem. 73, 286; 101, 342. " Annalen. 167, 125 ; 174, iSo. 10° Ibid. 158, 265. Berichte. 16, 1609; Bernthsen and Bender, ibid. 16, 1803. Riedel, 288 HISTORY OF CHEMISTRY. [LECTURE XIV. anthrene 1()- in which an atom of nitrogen has taken the place of CH: I have already stated (p. 1 1 7) that all these investigations as to the constitution of organic compounds were occasioned by the numerous cases of isomerism which meet the chemist at almost every step, and the existence of which seems to require some explanation. It must be acknowledged, that the theory of the valency or atomicity of the elements fulfilled this requirement to a large extent, and in this lies the great import- ance of that theory ; whereas, on the other hand, it cannot be denied that the principles of the theory are far from being clearly and precisely worked out — a matter into which I intend to enter more fully in the next lecture. But attention may here be called to the fact that not merely is the possibility of an explanation of these isomerisms supplied to us by the advancement of our theoretical knowledge, but this explanation chiefly depends upon the much more extensive experimental material at our disposal. And this material has, in great part, been obtained by the application of a method which, even although it has been recognised for a long time as a possible one, has only attained to pre-eminent importance within com- paratively recent times.. I refer to the method of synthesis, which is, moreover, in many cases, not merely a means to an end, but is itself the aim of the experiments. In an earlier lecture (p. 116) the synthesis by Wohler of an organic compound (urea) was mentioned, and also the importance of this synthesis in regard to our whole conception of nature. Similar results were only obtained in the cases of other substances long afterwards, and the value of this method was shown in a proper light by Berthelot's comprehensive work.103 The syntheses of some specially important substances — marsh gas, ethylene, alcohol, formic acid, benzene, etc. — also originated with Berthelot. It has been found in many cases that the earlier analytical 102 Ladenburg, Berichte, 16, 2063; Grabe, ibid. 17, 1370. ™3 Chimie organique fondee sur la Synthese, Paris 1860. LECTURE XIV.] HISTORY OF CHEMISTRY. 289 method is not sufficient for establishing the chemical nature of a compound, and that the synthetical method constitutes a necessary complement. The first method usually precedes the second ; but, in the history of a substance, its synthesis, with rare exceptions, marks a period, and, with it, the interest which the scientific investigation of the substance presents, is usually at an end. From this point of view, the syntheses of specially important substances are worthy of mention here. Thus alanine was prepared, in 1850, by Strecker, from aldehyde ammonia, hydro- cyanic acid, and hydrochloric acid.104 Five years later, Zinin obtained mustard oil from allyl iodide and potassium thio- cyanate,105 its connection with garlic oil having already been established much earlier by Wertheim.106 Glycocoll was pre- pared synthetically by Perkin and Duppa107 from bromacetic acid and ammonia, and Hiifner afterwards obtained leucine108 in an analogous manner. Racemic acid was prepared syntheti- cally by Perkin and Duppa109 from dibromsuccinic acid, and malic acid by Kekule no from monobromsuccinic acid. We are indebted to Kolbe for the synthesis of taurine,111 a substance which he prepared from isethionic acid. Anthracene was first prepared artificially by Limpricht, by boiling benzyl chloride with water;112 and guanidine was prepared by Hofmann,113 who obtained it from chlorpicrin, and by Erlenmeyer,114 who obtained it from cyanamide by the action of ammonia. Volhard prepared creatine synthetically from chloracetic acid,115 by converting the latter into sarcosine by the action of methyl- amine and then converting the sarcosine into creatine by means of cyanamide. Picoline and collidine were prepared syntheti- cally by Baeyer116 from aldehyde ammonias; crotonic acid, by Kekule, from aldehyde;117 and glycerine by Friedel and 104 Annalen. 75, 29. 10S Ibid. 95, 128. 106 Ibid. 55, 297- 107 Ibid. 108, 112. 108 Hiifner, J. pr. Chem. [2] I, 6. 109 Journ. Chem. Soc. 13, 102 ; Annalen. 117, 130. no Ibid. 117, 120. ln Kolbe, ibid. 122, 33. 112 Ibid. 139, 308. 113 Berichte. I, 145. m Annalen. 146, 259. 115 Zeitschrift fur Chemie. 12, 318. 116 Annalen. 155, 283. 117 Ibid. 162, 92. T 2QO HISTORY OF CHEMISTRY. [LECTURE XI V. Silvn, starting from acetone.118 Wurtz 119 converted glycol chlorhydrine into choline (neurine) by means of trimethylamine, whilst Reimer and Tiemann obtained vanilline from guaiacol.1'20 Grimaux synthetically prepared allantoi'n,121 alloxantine,1'22 and citric acid ; 123 and, more recently, Erlenmeyer prepared tyrosine,124 Ladenburg, piperidine 125 and coniine,1'26 and Hor- baczewsky, uric acid.127 The synthesis of indigo blue by Baeyer 128 also deserves to be mentioned, since not only did it familiarise us with the preparation and furnish us with an explanation of the constitution of an important colouring matter, but it was accomplished, besides, by means of new and peculiar reactions. Special attention is due to the general methods which permit the synthesis of whole groups of substances. The most important of these methods will be specified here. Frankland was the first who succeeded in building up hydrocarbons.1'29 He obtained dimethyl (ethane) from zinc and methyl iodide, and diethyl (butane) from zinc and ethyl iodide. This reaction was extended by Wurtz, who treated mixtures of alkyl iodides with sodium 136 — a method which Fittig and Tollens turned to account in the synthesis of aro- matic hydrocarbons. 13t It had already been found possible to obtain hydrocarbons, according to a reaction discovered by Berthelot, by the distillation of benzoates with salts of the fatty acids.132 A synthetical method was elaborated by Zincke, which permits of the preparation of hydrocarbons with two phenyl groups, and depends upon the action of benzyl chloride upon aromatic hydrocarbons in presence of zinc dust.133 These compounds can also be obtained, according to Baeyer, from aldehydes and aromatic hydrocarbons, by the aid of substances 118 Bull. Soc. Chim. [2] 20, 98. 119 Annalen. Supplementband 6, 116. 120 Berichte. 9, 424. 121 Ann. Chim. [5] II, 389. ia3 Jahresbericht 1878, 361. 123 Comptes Rendus. 90, 1252. 124 Annalen. 219, 161. ia5 Berichte. 18, 2956 and 3100. 126 Ibid. 19, 439 and 2578. 127 Mon- atsheftc. 3, 796 ; 6, 356. 128 Berichte. 13, 2254. 12B Annalen. 71, 171 ; 74, 41; -77, 221. 13° Ibid. 96, 364. 15J1 Ibid. 131, 303. 13>- Ann. Chim. [4] 12, 81. 133 Annalen. 155, 59, etc. LECTURE XIV.] HISTORY OF CHEMISTRY. 2QI which remove the elements of water.134 A method of very general applicability is that discovered by Friedel and Crafts,135 which renders it possible, by the help of aluminium chloride, to introduce groups of very different kinds into an aromatic substance, with simultaneous separation of hydrochloric acid or of water, and thus permits the synthesis of hydrocarbons, ketones, acids, etc. The possibility of ascending in the series of the primary alcohols, from one term to the next higher term, was shown by the investigations of Pelouze,136 Kolbe and Frankland,137 Piria,138 and Wurtz.139 The desired end is attained by the conversion of the alcohol into cyanide, acid, aldehyde, and alcohol, in accordance with the following equations : — CNK f C2H5S04K = C,H5CN + SO4K9 C,H6CN -I- KOH + H,O = C2H5CO9K + NH3 C.H-COH {- H2 = C2H5CH,OH. Lieben and Rossi ascertained the general applicability of the methods.140 There is also a second mode for obtaining, from one alcohol, the next term in the homologous series, viz., by converting the cyanide (nitrile) into an amine by means of nascent hydrogen (Mendius 141), and then decomposing this by means of nitrous acid (Hunt 142). A statement has already been made about the synthesis of secondary and tertiary alcohols (pp. 262 and 264). The preparation of the phenols from the hydrocarbons is accomplished by a process which Dusart, Kekule, and Wurtz 143 announced simultaneously. Aceto-acetic ether has become of great importance in the synthesis of acids, as aleady stated (p. 264). Malonic ether144 and benzoyl-acetic ether 145 have also been made use of in a 134 Berichte, 5, 1094. 135 Comptes Rendus. 84, 1392, 1450; 85, 74, etc. 136 Annalen. 10, 249. 137 Ibid. 65, 288 ; see also Fehling, ibid. 49, 95. 138 Ibid. IOO, 104 ; compare also Limpricht, ibid, 97, 368. 139 Ibid. 123, 140. 14° Ibid. 165, 109. 141 Mendius, ibid. 121, 129. 142 Liebig and Kopp's Jahresbericht 1849, 391. 143 Comptes Rendus. 64. I44 Conrad and Bischoff, Annalen. 204, 121. 145 Baeyer, Berichte. 15, 2705; Baeyer and Perkin, ibid. 16, 2128. 2Q2 HISTORY OF CHEMISTRY. [LECTURE XIV. similar manner ; whilst it has been possible, on the other hand, to obtain synthetically some interesting nitrogen compounds by the aid of aceto-acetic and malonic ethers.146 Perkin's reaction,147 which is connected with observations made by Bertagnini,148 and depends upon the action of aldehydes on the salts of organic acids in presence of agents which remove the elements of water, has led to the preparation of a large number of acids. It was first employed in a some- what more complicated form, however, in the synthesis of cumarine.149 The conversion of nitriles into acids, already referred to above, has also been employed in the preparation of polybasic acids ; and, for this purpose, it is possible to start, as Simpson did,150 from the cyanogen compounds of polyatomic radicals, or, as was shown by Kolbe 151 and by H. Miiller,152 from the cyanogen derivatives of acids. The first conversion of a nitrile into an acid was carried out, however, by Pelouze,153 who, in 1831, converted hydrocyanic acid into formic acid, and who reconverted the ammonium salt of the latter into hydrocyanic acid by the action of heat. Winkler,154a few years afterwards, converted oil of bitter almonds containing hydro- cyanic acid, into mandelic acid — a reaction which was correctly interpreted by Liebig.*65 Polybasic acids can also be obtained by a process published by Wislicenus,156 whilst Kolbe's reaction, which consists in treating phenates with carbonic anhydride, is of great importance in the synthesis of phenol acids.157 Re- lated to this, there is Reimer's synthesis of phenol aldehydes from phenates and chloroform.158 Finally Hofmann's method for the formation of alkyl 146 Compare especially Hantzsch, Annalen. 215, I ; Knorr, Berichte. 17, Referate. 148, 540, 1635 etc. ; Riigheimer, ibid. 17, 736. 147 Ibid. 8, 1599; compare also Fittig, Annalen. 216, 115; 227, 48. 148 Ibid. 100, 126. 149 Ibid. 147, 229. 15° Ibid. 118, 373; 121, 153. 151 Ibid. I3I5 348. lsa Ibid. 131, 350. 153 Ann. Chim. [2] 48, 395. 154 Annalen. 18, 310. 155 Ibid. 18, 319. 156 Ibid. 149, 215. 157 Kolbe and Lautemann, ibid. 115, 201 ; Kolbe, J. pr. Chem. [2] 10. 93. 158 Ber- ichte. 9, 42? LECTURE XIV.] HISTORY OF CHEMISTRY. 293 bases 159 may here be referred to. This method he afterwards altered and considerably improved.160 In many of these investigations an idea was turned to account which has already borne much fruit, and which will, no doubt, also be of great service in future. I refer to the so- called condensation processes ; that is, to the very frequently occurring formations, both in nature and in artificial reactions, of complex substances from simple ones, where several iden- tical or similar molecules unite to form one molecule, usually with the simultaneous elimination of hydrogen, water, ammonia, etc. Gerhardt drew attention to reactions of this kind when he formulated his theory of residues (p. 1 80), but it is only in comparatively recent times — within the last thirty years or thereabouts — that the necessary attention has been bestowed upon these processes. Berthelot was probably the first who closely studied such reactions, and he obtained valuable results by doing so. Amongst these results are the syntheses, dis- covered by him, of benzene C6H6 from acetylene, of diphenyl from benzene, of anthracene from toluene,161 etc. In these experiments he established, amongst other things, the fact, which has since been frequently confirmed, that at a high temperature several molecules of a hydrocarbon may unite to form a new molecule with the elimination of hydrogen. Some time afterwards, Baeyer began to work at this subject. He regards the difference between condensation and poly- merisation as consisting in the fact that in the former the molecules combine by virtue of union with carbon atoms, and in the latter of union with oxygen or with nitrogen atoms.162 It is already clear to him that for purposes of synthesis, condensation is alone of importance. He draws attention, besides, to very important syntheses which have already been carried out, such as the formation of mesitylene from acetone by Kane,163 and Chiozza's synthesis of cinnamic aldehyde from bitter almond 159 Annalen. 66, 129 ; 67, 6 1 and 129 ; 70, 129; 73, 180 ; 74, I, 33, 117 ; 75> 356; 78, 253; 79, ii. 16° Berichte. 14, 2725; 15, 407, 752., 762. 161 Bull. Soc. Chim. [2] 6, 268. 162 Annalen. Supplementband 5, 79. 163 Ibid. 22, 278, 294 HISTORY OF CHEMISTRY. [LECTURE XIV. oil and aldehyde by the action of hydrochloric acid.104 He then turns the theoretical views advanced at that time to immediate account in the synthesis of picoline and collidine, which are obtained by the condensation of acrolein-ammonia and of aldehyde-ammonia : — 165 2C3H4ONH3 = CGH7N + 2 4C2H4ONH3 = CsHnN + 4 Kekule, a few years later, condensed two molecules of alde- hyde so as to form crotonic aldehyde,160 thereby throwing light upon the chemical nature of the so-called acrylic aldehyde already examined by Lieben.107 This reaction was afterwards studied by Wurtz,168 who showed that the two aldehyde mole- cules unite in the first place, without the elimination of water, to form aldol, the aldehyde of /3-hydroxybutyric acid, and that crotonic aldehyde is then formed from the latter by the loss of water. The general character of this interesting reaction was established, subsequently, by various investigations, and espe- cially by the researches of Claisen.169 The idea of condensation has been greatly extended in recent times, every reaction in which union of carbon to carbon occurs amongst the molecules that act upon one another being designated a condensation. The word thus became synony- mous with synthesis, and lost all independent meaning and all meaning corresponding to its etymology. It is due to this that Baeyer's reaction for the formation of hydrocarbons from aldehydes and benzene and its derivatives, and likewise Perkin's method of forming unsaturated acids from aldehydes and the salts of fatty acids, came to be designated as condensations. The idea of condensation has also undergone change in another direction inasmuch as internal condensations have been contrasted with the processes just mentioned, which have in turn been called external condensations. By internal condensations we now understand reactions in which a single 164 Ibid. 97, 350. 165 Ibid. 155, 283 and 297. 166 Ibid. 162, 77. 167 Ibid. 106, 336; Supplementband. i, 114. 168 Jahresbericht 1872, 449; 1873, 474; 1876, 483; 1878, 612. 169 Annalen. i8o? i; ibid. 218, 121. LECTURE XIV.] HISTORY OF CHEMISTRY. 295 molecule of a substance becomes converted into a new mole- cule by parting with some of its atoms, which unite to form such a molecule as H.>, HC1, H.,O, NH3, etc. ; />., reactions which occur within a molecule. As applied to such reactions, the word condensation so far retains its meaning, that the atoms are related more intimately (that is, by a greater number of valencies) to one another. Amongst these reactions there are many processes which have long been known, such as the formation of ethylene from alcohol, of C0C14 from C2C16, of aldehydes or of ketones from alcohols, of ethylene oxide from glycol, of anhydrides from polybasic acids, etc. But the for- mation of the anhydrides of monobasic acids, of the lactones and of the lactone acids, which have been minutely studied by Fittig, comparatively recently, must also be regarded as internal condensations. To the same class of reactions belong, further, the formation of cumarine, and that of the oxycumarines (umbelliferone, daphnetine, etc.), of isatine, indol, rosaniline, of rosolic acid, of the phthale'ines, of the aldehydines, of quinoline, naphthalene, anthracene, etc. Consequently these processes have played an important part in more recent investigations, and they will engage our attention here a little longer. The formation of ethenyl-xylene-diamine and of ethenyl- toluylene-diamine by the reduction of nitro-acet-xylid and of nitro-acet-toluid, observed by Hobrecker,170 first attracted Pliibner's attention to this matter. The latter chemist prepared a large number of analogous compounds, and was able to show that this abnormal course of the reduction only occurred with the ortho-benzene derivatives, and not with the meta- or para- derivatives.171 This was entirely confirmed by Ladenburg's investigations.17- The latter chemist discovered quite a number of reactions which proceed altogether differently in the ortho- series from the way in which they proceed in the other isomeric series. In the case of the diamines, he showed, by means of 170 Berichte. 5, 920. 171 Ibid. 8, 471 ; Annalen. 208, 278 ; 209, 339; 2IO, 328. 172 Berichte. 8, 677 ; 9, 219 and 1524; 10, 1123, 1260, etc. 296 HISTORY OF CHEMISTRY. [LECTURE XIV. reactions of this very kind, how the ortho-compounds may be distinguished from their isomers ; and he was the first to point out that the formation of the above-mentioned substances depends upon " ortho-condensation." Baeyer then turned his attention to this subject, and the syntheses of quinoline and of oxindol, already referred to, constitute the valuable fruits of his studies. Closely related to this internal condensation, is internal oxidation. Reactions involving the latter change are those in which oxygen atoms already present in the molecule, and generally belonging to NO.2 groups, oxidise, by the dissolution of existing unions, other groups belonging to the same molecule. The first reaction of this kind was observed by Wachendorff,173 but Greiff 1T4 was the first to explain it. The matter involved was the action of bromine on ortho-nitrotoluene, which the latter of the above-named chemists represented in the following manner : — This reaction furnishes the explanation of the important method, discovered by Baeyer,175 for preparing isatine from ortho-nitro- phenylpropiolic acid, by boiling this acid with alkalies : — /CO— CO = C6H4<^ / The formation of indigo from ortho-nitro-phenylpropiolic acid, depends upon similar rearrangements. "3 Berichte. 9, 1345. 174 Ibid. 13, 288. 175 Ibid. 13, 2259. LECTURE XV. THE FUNDAMENTAL CONCEPTIONS OF CHEMISTRY — PHENOMENA OF DISSOCIATION — ABNORMAL VAPOUR DENSITIES -CONSTANT OR VARIABLE VALENCY — THE DOCTRINE OF VALENCY IN INORGANIC CHEMISTRY— THE PERIODIC LAW— LATER DEVELOPMENT OF THE DOCTRINE OF AFFINITY — SPECTRUM ANALYSIS— SYNTHESIS OF MINERALS— CONTINUITY OF MATTER IN THE LIQUID AND GASEOUS STATES — LIQUEFACTION OF THE SO-CALLED PERMANENT GASES — TIIERMO-CHEMISTR-Y — ELECTRO-CHEMISTRY — PHOTO-CHEMISTRY — MOLECULAR PHYSICS— MORPHOTROPY. HAVING now followed organic chemistry in some of its more recent discoveries, and having obtained a knowledge of its re- markable progress under the influence of the theory of valency, it is appropriate to suggest and to discuss the question whether this theory is capable of serving as a fundamental principle in mineral chemistry ; and also to recount some of the most important results of investigations in general chemistry. Before passing on, however, to this part of our task, the theories themselves must be subjected to a more minute con- sideration and scrutiny. In describing how they have come into existence we have not always been able to enter into the exact significance of their fundamental conceptions. We shall now turn our attention to this matter, although, naturally, it is only possible to bring forward the most important points. For the remainder, the reader is referred to the standard text-books of theoretical and general chemistry. Our views rest essentially on the precise formulation and distinction of the conceptions of atom, molecule, and equivalent. An atom is denned as the smallest indivisible quantity of an element which exists under any circumstances ; and most generally it only exists in combination with other atoms. 298 HISTORY OF CHEMISTRY. [LECTURE XV. A molecule is defined as the smallest quantity of a chemical substance that occurs in the free state, whether the substance be elementary or compound. The determination of the mole- cular weight depends essentially upon our combining the con- ceptions of the physical and of the chemical molecule ; that is to say, we apply the word molecule to the smallest quantity of a substance which occurs free in the gaseous state, as well as to the smallest quantity that enters into a reaction. With respect to determinations of atomic weights, it is to be remarked here that the numbers proposed by Gerhardt1 were subjected to an important alteration in so far that the atomic weights of all the metals were doubled, except those of the monatomic ones (i.e., the alkali metals and silver). As early as 1840, when the atomic weights of Berzelius were still in use, Regnault had proposed to halve the atomic weight of silver, and, in accordance with this proposal, to assume two atoms of metal in silver oxide for one atom of oxygen.2 He afterwards made a similar proposal with respect to the atomic weights of potassium, sodium, and lithium.3 The reason was, that his classical experiments on specific heat had shown him that Dulong and Petit's law only applied to these metals when this assumption was made. .Had this proposal of Regnault's been adopted at that time, our present atomic weights would (with few exceptions) have been obtained. But since, following Ger- hardt's lead, the atomic weights of all the metals were halved, it was afterwards necessary (when the desirability of Regnault's proposal had been shown upon new grounds, especially by H. Rose4 and by Cannizzaro5) to double them again, with the exception of those of the metals mentioned above. Cannizzaro in particular, showed, in his pamphlet referred to below, that the law of Dulong and Petit was a guide in the determination 1 Compare, for example, Gerhardt, Introduction a 1'Etude de la Chimie, 1848, 29. 2 Ann. Chim. [2] 73, 5 ; Annalen. 36, no. 3 Ann. Chim. [3] 26, 261. 4 Pogg. Ann. 100, 270. 5 Nuovo Cimento. 7, 321 ; also Repert. de Chimie pure. I, 201 ; compare Suhto dj un Corso di Filosofia CWmica 1858, 35. LECTURE XV.] HISTORY OF CHEMISTRY. 299 of the atomic weights, just as the hypothesis of Avogadro was in that of the molecular weights. Even at this time the carrying out of these principles was still confronted by great difficulties. It is true that Deville and Troost6 showed, at this date, that the vapour density of sulphur, at about 1000°, was only one- third of the number previously found by Dumas and Mitscher- lich (compare p. 105) at lower temperatures; so that it was possible to adopt the molecular formula S2 for sulphur. The anomalies previously observed in the cases of mercury, phos- phorus, and arsenic, remained, however; but these were not an obstacle to Cannizzaro. The chemical relations had to stand aside in order to procure acceptance of the principle. He assumed that only one-fourth of a molecule of phosphorus and of arsenic, respectively, was contained in two volumes of phosphuretted and of arseniuretted hydrogen ; whilst half a molecule of nitrogen is present in two volumes of ammonia, and a molecule of mercury in two volumes of mercuric chloride. Consequently the divisibility of the molecule is different, according to Cannizzaro, even in chemically analogous sub- stances. Even although this appeared to be a bold view, still no decisive reasons could be established against it. The assumption of differences of constitution amongst the elementary molecules, although striking at first, seemed after a t/ time to be fully justified. Why should not a state of matters be met with in the case of the elements similar to that observed amongst compound substances, the molecules of which are known to present the greatest variety with respect to the num- bers of their atoms ? Cannizzaro very aptly compares the elements with the hydrocarbons— the molecules of hydrogen, oxygen, etc., with the so-called alcohol radicals, methyl, ethyl, etc., and the molecules of mercury, zinc, and cadmium, with the olefines, a view which may also be extended to the derivatives of both classes of substances : — 7 6 Comptes Rendus. 49, 239 ; Annalen. 113, 42. 7 Wurtz, Le9ons de philosophic chimique, 172. 300 HISTORY OF CHEMISTRY. [LECTURE XV. H0, Oo, N0 corresponding to (CH3).,, (C.,H5)., Hg,Zn" , C2H;, C3H6 " K.,O, CaO, ZnO Bi.203, Sb.20j SnO0, SiOo KOH Ca(OH)o Bi(OH)"3 Sn(OH)4 (CH3)oO, (CoH6),0 C2H4O, C,H6O (C3H5)203 COo CH3OH, C,H6OH CoH4(OH)2, C3H,(OH)o C3H5(OH)3 C4H6(OH)4 Of decisive importance, however, as regards the nature of the molecule of mercury, are the experiments of Kundt and Warburg,8 which may be adopted as a direct proof of Canniz- zaro's view. These physicists, by observing the velocity of sound in mercury vapour, determined the ratio of the specific heats at constant pressure and at constant volume to be 1.67 — a number furnished by the mechanical theory of heat, on the assumption that the total energy of the gas consists of the translatory motion of the molecules. The demonstration, furnished by Victor Meyer, of the variable vapour density of iodine,9 which, as Crafts in particular has shown,10 eventually sinks to one-half of the original density and then remains constant, can only point to the fact that. the molecule of iodine, at high tem- peratures, consists of a single atom. A great deal more trouble was experienced in fixing the molecular weights of compounds, in those cases where the numbers calculated from the vapour densities did not agree with those deduced from the chemical relationships of the sub- stances. In his determinations of the relative densities of vapours, Bineau obtained such remarkable numbers that he considered decomposition to be the cause of the peculiar volume relations.11 Thus he found the density of ammonium carbamate (anhydrous carbonate of ammonia, as he calls it) to correspond to six volumes, whence he assumes a decomposition into four volumes of ammonia and two of carbonic anhydride. 8 Berichte. 8, 945. Pogg. Ann. 157, 353. 9 Berichte. 13, 394. 10 Comptes Rendus. 92, 39. n Ann. Chim. [2] 68, 434 ; 70, 272. LECTURE XV.] HISTORY OF CHEMISTRY. 30! Mitscherlich made a similar assumption in the case of antimony pentachloride,12 and so did Gladstone 13 in that of phosphorus pentabromide. In both of these cases it was assumed that, besides the halogens, the trichloride and the tribromide of the respective elements had been produced. Cahours14 also ex- pressed the same view in 1847, in explanation of the low vapour density of phosphorus pentachloride. In the same year, Grove 15 made the remarkable observation that water is decomposed into its elements by contact with brightly glowing platinum, — a circumstance which he sought to explain as a result of the high temperature. This view met, however, with little acceptance, the fact that platinum can be melted by means of the oxy-hydrogen flame being looked upon as opposed to it. Consequently, Grove's experiment was re- garded as a result of the action of affinity ; and it was explained as exactly similar to the decomposition of water (observed by Regnault) by means of melting silver, where silver oxide and hydrogen were supposed to be formed.16 Grove's way of regarding the matter was first definitely proved by Henry St Claire Deville as the result of a very detailed investigation which constitutes the basis of the theory of dissociation. Before passing on to describe more minutely these pheno- mena, which are highly important for chemistry, I must here point out that the views with respect to them have been affected by the advances which have meanwhile been made in our knowledge of heat. These advances have been called forth by the law of the conservation of energy, which, as is well known, was first clearly formulated by J. R. Mayer ; ir and they find their expression especially in the mechanical theory of heat and in the kinetic theory of gases, developed chiefly by Clausius, Joule, Rankine, Thomson, Helmholtz, Maxwell, and others. After drawing attention to the fact that the affinity of silver 12 Pogg. Ann. 29, 227. 13 Phil. Mag. [3] 35, 345. 14 Ann. Chim. [3] 20, 369. 15 Phil. Trans 1847, I ; Annalen. 63, I. 16 Ann. Chim. [2] 62, 367. 17 Annalen. 42, 233. 302 HISTORY OF CHEMISTRY. [LECTURE XV. for oxygen could not come into play in Regnault's experiment (since silver oxide breaks up into its constituents at much lower temperatures, and the same thing must certainly take place in presence of hydrogen), Deville shows that the decomposition of water by means of strongly heated lead oxide (at 1200° to 1300°) is also observed. He succeeds in effecting the same decomposition by means of ingeniously contrived apparatus, without the action of a foreign substance ; and in this way his opinion that the decomposition is a result of the high tem- perature is confirmed in an elegant manner.18 The difficulty in these investigations arises from the fact that the constituents separated during the decomposition, com- bine again at lower temperatures, so that the decomposition which has occurred is not recognisable under ordinary circumstances. The proof that decomposition has occurred may be furnished, as Deville shows, (i) by diluting the products of the decom- position by means of a rapid current of an indifferent gas, so that complete recombination is prevented ; (2) by diffusion, whereby the composition of the gaseous mixture is altered ; or (3) by means of the so-called tube chaud et frbid , i.e., by sudden cooling of the products of decomposition. In the forms of ap'paratus constructed to carry out these methods, Deville succeeded in proving not only the decom- position of water into hydrogen and oxygen, but also that of carbonic anhydride into carbonic oxide and oxygen, of carbonic oxide into carbon and carbonic anhydride, of hydrochloric acid into chlorine and hydrogen, of sulphurous anhydride into sulphuric anhydride and sulphur, etc. Supported by these experiments, Deville compares the formation of compounds with the condensation of vapours. According to him, both changes begin at definite temperatures and both proceed gradually. Certain quantities of heat are given out during the condensation of vapours, and the same thing takes place (frequently to a much greater extent) in the 18 Comptes Rendus. 56, 195, 322, 729 ; compare also Deville, Lecons sur la dissociation, 1864. LECTURE XV.] HISTORY OF CHEMISTRY. 303 combination of two substances. But further, in exactly the same way that evaporation begins below the condensing point, the decomposition of substances can be observed below the true combining temperature. As every degree of the thermo- metric scale corresponds to a definite vapour pressure, so, in certain cases at least, the pressures of the products of decom- position can be stated. Deville distinguishes between decom- position by the action of heat and decomposition by chemical means. He applies the name dissociation to the former only.19 It is characterised by the facts that its different phases can be observed ; that it begins at one definite temperature and is com- pleted at another ; and that between these limits the pressures increase from o to 760 mm. and more of mercury, so that a definite pressure, due to the gaseous products of the decom- position, corresponds to every temperature. Subsequent experiments on this subject (of a very detailed character) have confirmed Deville's views, in general at least. Only those decompositions are now regarded as examples of dissociation which take place in opposition to the chemical forces and are accompanied by the absorption of heat.20 The comparison of these phenomena with evaporation, even if it is not quite generally applicable, still holds in the decomposition of solid substances with the formation of gaseous constituents, as was shown by Debray in the case of calcium carbonate,21 by Naumann in that of ammonium carbonate,22 of Isambert in that of ammonium hydrosulphide,23 and by others. Investigations of the compounds of silver chloride with ammonia24 and of compounds containing water of crystallisation, came to be of special importance, because in the cases of these substances the different compounds with ammonia and the different stages of 19 Annalen. 105, 383. 20 Compare Horstmann, Theoretische Chemie. 666. 21 Comptes Rendus. 64, 603 ; Bull. Soc. Chim. 7, 194. ^ Berichte. 4, 779. '23 Comptes Rendus. 92, 919; 93, 731. "4 Isambert, Laden- burg's Handworterbuch der Chemie. 3, 400 ; Ilo-stmann, Berichte, 9> 749- 304 HISTORY OF CHEMISTRY. [LECTURE XV. hydration of the salts, respectively, were indicated by the abrupt variations of the pressure.25 Pfaundler 2(5 endeavoured to explain the, at first, surprising fact of a partial decomposition which gradually increased with rise of temperature, involving, as it did, the different behaviour of similar molecules under the same conditions. Naumann 27 further developed these views, and they were more definitely formulated by Horstmann,28 who made use of Maxwell's pro- bability theory 29 of the distribution of the velocities.30 A close agreement between this theory and the observations was noted in various cases. Horstmann was the first who tried to establish a general theory of dissociation,31 starting from the principles of the mechanical theory of heat — especially from the so-called second law. This was found to be in complete accord with the results of experiment in one case.32 The researches of Gibbs 33 and of Helmholtz,34 which were based upon similar principles, were more comprehensive and highly productive. These investigations, which belong, in part, to the domain of physics, have become of great importance with respect to the question we have here to consider. Shortly after the first researches of Deville, the opinion was stated by three different chemists — Cannizzaro,35 Kopp,36 and Kekule37 — that the so- called abnormal vapour densities were to be explained as due to the substances concerned breaking up into two or more con- stituents. The latter were supposed to re-combine on cooling, so that no decomposition was perceptible upon distillation. The difficulties which thus stood in the way of a direct 25 Debray, Comptes Rendus. 66, 194 ; G. Wiedemann, Pogg. Ann. Jubelband 1874, 474. w Pogg. Ann. 131, 60. <27 Annalen. Supple- mentband 5, 341. w Berichte. I, 210. <29 Phil. Mag. [4] 19, 22 ; 35> l%5- 30 Compare also Boltzmann, Wiedem. Ann. 22, 31. 31 Ann- alen. Supplementband 8, 112 ; 170, 192. ** Ibid. 187, 48. 33 Silli- man's Journal. 16, 441 ; 18, 277. a4 Berlin. Akad. Ber. 1882, 22, 825 ; 1883, 647. 35 Nuovo Cimento. 6, 428 ; 7, 375 ; 8, 71. Compare also Repert. de Chimie pure. I, 201. 36 Annalen. 105, 390. :!7 Ibid. 106, 142. LECTURE XV.] HISTORY OF CHEMISTRY. 305 proof of decomposition, were only overcome some years after- wards by Pebal,38 who based his experiments on the state- ment, first made by Bunsen,39 that it was only possible to distinguish mixtures of gases from homogeneous gases by physical methods (diffusion or absorption). On causing the mixture of gases obtained by heating ammonium chloride to diffuse through an asbestos plug, Pebal was able to show, by the colours imparted to litmus, that the gas in one part of the apparatus possesses an alkaline and in another part an acid reaction. In a similar manner, by means of diffusion, Wanklyn and Robinson 40 endeavoured to show the breaking up of sulphuric acid into sulphuric anhydride and water, and of phosphorus pentachloride into phosphorus trichloride and chlorine. Deville attacked the conclusions which these chemists drew from their experiments.41 According to him, complete decom- position was not necessary in order to accomplish a separation of the constituents by means of diffusion, a dissociation involv- ing a slight increase of pressure being quite sufficient. As the products of decomposition are carried forward, further quan- tities are formed, so that, given a sufficiently long duration of the experiment, a complete separation of the constituents is attained at a temperature which only corresponds to a very slight decomposition. Deville points out that the vapour den- sity of water is still normal at 1000°, while at this temperature it can be shown by diffusion that dissociation has already taken place ; 42 and hence he considers that the abnormal vapour density must be ascribed to the undecomposed vapour of ammonium chloride. He finds what he regards as a positive proof of this, in the considerable rise of temperature which he believes he can recognise upon the intermixture of ammonia and hydrochloric acid gases in a vessel previously heated to 350°. 43 Robinson and Wanklyn having raised the objection a8 Annalen. 123, 199. 39 Bunsen, Gasometrische Methoden, 1857, 242. 40 Comptes Rendus. 56, 547. 41 Ibid. 56, 729. 4a Deville, Le$ons sur la dissociation, 365. 43 Comptes Rendus. 56, 729. V 306 HISTORY OF CHEMISTRY. [LECTURE XV. that the ga^es had not been sufficiently heated prior to their inter-mixture,44 Deville afterwards repeats the experiment in a manner which no longer permits of this objection being raised, and again observes a rise of temperature, the amount of which, however, he does not state.45 He finds a further argument in his favour in the fact that ammonia, when heated to 1100°, breaks up into nitrogen and hydrogen. In his opinion ammo- nium chloride, after having been heated to this temperature, ought to yield these two gases on cooling as evidence of the formation of ammonia ; but this is not the case. In opposition to this argument, Than adduces the fact that a gaseous mixture is much more difficult to decompose than a pure gas,46 and this is in complete agreement with Deville's views regarding dissociation.47 By diminishing the partial pres- sure, the temperature at which dissociation begins is raised ; 48 or, the temperature remaining the same, the pressure due to decomposition is diminished. Further, Than observed no rise in temperature on mixing hydrochloric acid and ammonia at 360°. Even if the errors were greater in his arrangement of the experiment, and assuming that he was unable to measure very small differences of temperature, still it is placed beyond doubt by his statements that only inconsiderable quantities of heat are liberated by the intermixture of ammonia and hydro- chloric acid at 360°. This is confirmed by an experiment by Marignac,49 who was able to • show that just as much heat is evolved in the formation of ammonium chloride from ammonia and hydrochloric acid as is required for its volatilisation. Hence it may be looked upon as fully proved that ammonium chloride does not exist in the gaseous state, but that it breaks up, on volatilisation, into its components. Similar facts, even if not always so convincing, have also been observed in the cases of many other compounds whose molecules in the gaseous state correspond to four volumes ; as, 44 Comptes Rendus. 56, 1237. *» Ibid. 59, 1057. * Annalen. 131, 129. 47 Deville, Le9ons. 364. 48 Compare Naumann, Annalen, Supplementband 5, 341. 49 Comptes Rendus. 67, 877. LECTURE XV.] HISTORY OF CHP:MISTRY. 307 for example, phosphorus pentachloride,50 ammonium sulphide,51 ammonium carbamate,52 etc. A lengthy discussion took place between Wurtz on the one hand.53 and Troost,54 Deville,55 and Berthelot 50 on the other, regarding the nature of the vapour ob- tained from chloral hydrate. This discussion ended in favour of the former, and led to the proof of the decomposition of chloral hydrate upon vaporisation. Returning now to the definitions of our fundamental con- ceptions (compare p. 297), we designate as the equivalent, or better, the equivalent weight, that quantity of an element or of a radical which can replace or combine with one atom of hydrogen. This conception, however, no longer plays any essen- tial part ; another one, which stands in close relationship to it, having been introduced instead of it, — that, namely, of valency or atomicity. By this term we understand the quotient obtained by dividing the atomic weight by the equivalent, and it was discussed at length in the preceding lecture. The question as to whether the valency of any given element is constant or variable is one of particular importance. So long as we are satisfied to formulate and to make use of our conception of valency in harmony with the above definition, constant valency may, of course, be assumed. As soon, however, as we compare (as it is necessary that we should do) the valencies of the multi- valent elements with one another, we 'can no longer assert the absolute constancy of the valency of any element. Even in the case of carbon, where the assumption of a uniform quadri- valence encounters relatively few exceptions, the existence of carbonic oxide is at variance with its universal accuracy. We find a similar thing, only to a greater extent, in the cases of the other elements, and are therefore obliged to admit the pos- sibility of exceptions in every case. Two different methods 50 Cahours, Ann. Chini. [3] 20, 369; Deville, Comptes Rendus. 62, 1157. 51 Horstmann, Annalen. Supplementband 6, 74. 52 Naumann, ibid. 160, i. 5:? Comptes Rendus. 84, 977, 1183, 1262, 1347 ; 85, 49 ; 86, 1170 ; 89, 190, 337, 429, 1062 ; 90, 24, 118, 337, 572. « Ibid. 84, 708 ; 85, 32, 144, 400; 86, 331, 1394. 55 Ibid. 84, 711, 1108, 1256. 56 Ibid, 84, 1189, 1269; 85, 8; 90, 112, 491. 308 HISTORY OF CHEMISTRY. [LECTURE XV. have been proposed in order to bring these exceptions as far as possible into harmony with the system, but neither of them wholly gets rid of the difficulty. One party, under the leadership of Kekule,57 adheres to the definition of valency given above, but admits that there is a large class of substances to which it is not applicable. This class is composed of the molecular compounds, the smallest particles of which consist of aggregates of molecules held to- gether by means of molecular forces. Examples of the class are the compounds containing water of crystallisation (also alcohol, benzene, etc., of crystallisation), the majority of double salts, the ammonium salts, phosphorus pentachloride, iodine trichloride, etc. No precise definition of them can be given, but they are characterised generally by the facts that they can- not pass, undecomposed, into the state of vapour (although Thorpe found phosphorus pentafluoride to be an exception to this rule 58), and that they are easily formed from and decom- posed into their molecular constituents. The adherents of constant valency are further obliged to recognise the un'saturated compounds as exceptions. Even although there are not a very great many of these compounds, still their existence constitutes a serious objection to the doctrine; and fruitless endeavours have been made to weaken this objec- tion on the ground of the tendency exhibited by such substances to become saturated.59 The opponents of these views, whose first representatives are Frankland and Couper,60 define the valency of an element as its maximum saturating capacity, and under this definition the unsaturated compounds cease to occupy an exceptional posi- tion. In view of the fact that they further assume the valency in the case of many elements to be considerably higher than had previously been assumed — nitrogen and phosphorus, for example, as quinquivalent, sulphur as sexivalent, iodine as 57 Lehrbuch der Chemie. I, 142, 443 ; Comptes Rendus. 58, 510. 58 Annalen. 182, 204. °9 Compare p. 269 and Horstmann, Theoretische Chemie. 295. M Compare pp. 231 and 254. LECTURE XV.] HISTORY OF CHEMISTRY. 309 quinquivalent or septivalent — it is possible for them to include in the system a large number of molecular compounds. But it also becomes necessary for the adherents of this view to explain the change in the saturating capacity, or at least to establish the conditions which bring about this change in the properties of the elements, if their hypotheses are to deserve the name of a theory. Very little has yet been done in this direction, how- ever, and the little that has been done is scarcely capable of any general formulation.01 On the other hand, a number of facts have become known which can only be explained with difficulty on the assumption of constant valency ; such, for example, as the identity of the naphthyl-phenyl sulphones and of the tolyl-phenyl sulphones which can be prepared in dif- ferent ways ; 62 or such as the isomerism of the two triphenyl- phosphine oxides, one of which, P(C(5H5)3O, is supposed to correspond to phosphorus pentachloride ; and the other, P(C6H5)2OC6H5, to phosphorus oxychloride.63 It is clear from these few observations that the subject of valency, quite apart from any mathematical basis (which is at present altogether wanting 64), must still be called a very anoma- lous and uncertain one, and that there is no existing conception of it which is capable of dealing in a logical manner with the whole domain of chemistry. That the idea is still retained, in spite of this, and that it is even yet regarded as one of the most important principles, is explicable, in organic chemistry at least, on account of the almost marvellous consequences which the latter branch of the subject is able to show as the result of its assistance during the last forty years. In inorganic chemistry, however, the state of matters is very different. No doubt a favourable and helpful influence may be ob- served in inorganic chemistry also ; and, in particular, classifi- 61 Compare, however, Horstmann, Theoretische Chemie. 327 et seq. ; Van 't Hoff, Ansichten liber die Organische Chemie. I, 3. 62 Michael and Adair, Berichte. 10, 583; II, 116. 63 Michaelis and Lacoste, ibid. 18, 2118. 64 Compare, however, Kekule, Annalen. 162, 86; Baeyer, Berichte. 18, 2277. 316 HISTORY OF CHEMISTRY. [LECTURE XV. cation has become essentially clearer, as I wish to illustrate in individual cases. The possibility of classifying the elements themselves according to their valencies indicates a step in advance, since analogies were thereby brought out which had only been partially recognised previously. The analogy of carbon with silicon had already been pointed out, but boron had also been placed along with these two. The analogy of the two former was now established much more clearly, whilst boron was recognised as belonging to another series altogether. On the other hand, titanium, zirconium, and tin were classed along with carbon and silicon. Similarly arsenic, antimony, and bis- muth took their places beside nitrogen and phosphorus, then vanadium also, as a consequence of Roscoe's careful investiga- tion,65 and, finally, niobium and tantalum when the results of Marignac's researches were published.66 A similar thing took place with the metals, which had hitherto been arranged either according to their relative densities or to their analytical be- haviour. The theory of valency exercised, further, a decided influence upon the views respecting many classes of compounds. This was the case with the silicates in particular. Wurtz showed how the facts ascertained by him with respect to the condensa- tions of glycol might be extended to the derivatives of silicic acid,67 and, by so doing, he brought sudden light into a hitherto obscure region. Soon afterwards this region was further illumi- nated by Tschermak's 68 important research on the felspars, in accordance with which these substances must be regarded as isomorphous mixtures of orthoclase, albite, and anorthite. The numerous metal-ammonia and metal-ammonium compounds now found a place in the system also, being looked upon as ammonia or as ammonium chloride with hydrogen atoms re- placed by metal or metallic oxide. Hofmann was the first to attempt this classification,69 and in doing so, he turned to 65 Annalen. Supplementband 6, 77. ^ Ann. Chim. [4] 8, 5 and 49 (abstract); Annalen. 135, 49; Ann. Chim. [4] 9, 249; Annalen. Supple- mentband 4, 350. 67 Repert. de Chimie pure. 2, 449 ; Lemons de philosophic chimique. 181. m Pogg. Ann. 125, 139. 69 Annalen. 78, 253; 79, ii. LECTURE XV.] HISTORY OF CHEMISTRY. 31 1 account the results of his researches on organic bases. The same idea was more fully carried out by Weltzien,70 H. Schiff,71 Cleve,72 and many others. Despite all this, it cannot be said that the theory of valency has proved very productive in inorganic chemistry. In the first place, the number of researches which it has occasioned is by no means very considerable ; and, further, a systematic treatment of the subject, based upon valency, is not capable of being uniformly and logically carried out. Another hypo- thesis has had a far more important and lasting effect, and is now, some thirty years after its first promulgation, able to show the most brilliant and undreamt-of results. I must now discuss the relations which have been found to subsist between the atomic weights and the properties of the elements. The more recent investigations on this subject are related to Prout's hypothesis, which has already been considered.73 It is true that this hypothesis never became generally accepted, but nevertheless it occasioned speculations from time to time in the same direction. I only mention here Dobereiner, who, iri 1829, first drew attention to what he called the triads74 (that is, to groups of three analogous elements possessing atomic weights such that one of them might be regarded as the arith- metical mean of the other two). Gmelin,75 Dumas,76 and Lenssen77 further elaborated these ideas, without arriving at any results specially worth mentioning. A valuable result was, however, attained by the proof that the properties of the elements are periodic functions of their atomic weights. For this, we are indebted to the investigations of Newlands,78 Lothar Meyer,79 and Mendelejeff.80 The chief merit unquestionably belongs to the latter, who first gave pro- minence to the existing relations in a quite general form, and 70 Ibid. 97, 19. 71 Ibid. 123, i. r2 Bull. Soc. Chim. [2] 7, 12 ; 15, 161 ; 16, 203 ; 17, 100, 294. 73 Compare p. 102. 74 Pogg. Ann. 15, 301. 75 Handbuch. Third Edition, I, 35. 76 See p. 102. 77 Annalen. 103, 121 ; 104, 177. 78 Chem. News. 10, 59, 94; 13, 113. 79 Moderne Theorien. First Edition, 136 ; Annalen. Supplementband 7, 354. 80 Zeit- schrift fiir Chemie. 12, 405 ; Annalen. Supplementband 8, 133. 312 HISTORY OF CHEMISTRY. [LECTURE XV. (what must be regarded as specially important) pointed out clearly the advantages of considerations of the kind. It is for this reason that his paper at once made a great sensation, whereas that of Newlands remained quite unnoticed.81 In the tabulation adopted by Mendelejeff, the elements are arranged according to their atomic weights ; but they are further arranged into divisions in such a way that the elements which are analogous to one another fall into vertical columns and form groups, whilst each set of from seven to ten elements succeeding one another in a horizontal line in the order of their atomic weights, constitutes a short period within which the properties (physical as well as chemical) progressively vary. Two successive horizontal series form a long period, in con- nection with which it is to be noted that, in the groups, the analogies between elements of the even, and also of the uneven, horizontal series, are greater amongst themselves than those between elements which belong partly to the even and partly to the uneven series. Of the applications of the "periodic law," the two following have attained special importance: — (T) The determination or correction of the atomic weights of insufficiently investigated elements, and (2) the prediction of the properties of unknown elements. With respect to the first of these applications the following facts must be mentioned here. In conformity with the pro- posal of Awdejeff,82 Mendelejeff assumed the atomic weight of beryllium to be 9, and placed this element in a group along with magnesium ; whereas it had hitherto been regarded by many chemists as a metal akin to aluminium, and of atomic weight 13.5. This new view called forth a prolonged dis- cussion, which terminated, however, with the complete triumph of Mendelejeff s opinion.83 81 With respect to the question of priority, compare Newlands, Chem. News. 32, 21, 192; L. Meyer, Berichte. 13, 259; Mendelejeff, ibid. 13, 1796. f2 Fogg- Ann. 56, 101 ; compare also KJatzo, J. pr. Chem. 106, 227. 83 Nilson and Fettersson, Berichte. II, 381 ; 13, 1451 ; 17, 987 ; L. Meyer, ibid. 13, 1780; Reynolds, ibid. 13, 2412 ; Nilson, ibid. 13, 2035. LECTURE XV.] HISTORY OF CHEMISTRY. 313 The atomic weight of indium was assumed to be 113, or one and a half times as great as previously, and this number was very soon confirmed by the determination of the specific heat of the metal by Bunsen84 and by Mendelejeff.85 The atomic weight of uranium was doubled — a proceeding which was found by the excellent and detailed investigations of Zimmermann8*5 to be in complete agreement with the facts. Finally, it may be pointed out that Mendelejeff adopted 125 as the atomic weight of tellurium, in opposition to the previous determinations which had furnished the number 128. This was also apparently confirmed by the redetermi nation of the atomic weight in a specially purified sample,87 but Brauner's most recent experiments gave essentially different results. But Mendelejeff's predictions respecting new elements have been followed by results which are simply marvellous. In order to render possible the arrangement into groups and series, and to attain approximately equal differences in successive members, blanks had to be left, which, according to Mendelejeff, would be filled up by existing, but at that time unknown, elements. He was able to foretell the atomic weights and other properties of these elements from their position in the system, with the aid of the properties observed in the groups and series, which, like a system of co-ordinates, could be called in to assist. Three such blanks occurred in the first five series, and these he indicated as representing the positions of eka-boron (at. wt. 44), eka-aluminium (at. wt. 68), and eka-silicon (at. wt. 72). Since that time, these three elements have been dis- covered, and they have been found to possess, approximately, the properties predicted by Mendelejeff. They are: scandium, discovered by Nilson,88 with atomic weight 44.1 ; gallium, dis- covered by Lecoq de Boisbaudran,89 with atomic weight 70; and germanium, discovered by Winkler,90 with atomic weight 72. But these are not the only results which render this theory 84 P°gg- Ann- I4I> I' 85 Bull, de 1'Acad. Imp. de St. Petersbourg. 16, 45. 86 Annalen. 213, 285 ; 216, I. 87 Brauner, Berichte. 16, 3055 ; compare p. 346. 88 Uerichte. 12, 550, 554 ; 13, 1439. 8 i77. LECTURE XV.] HISTORY OF CHEMISTRY. 327 remarkable that the numerous experiments designed to ascer- tain the maximum effect of the different parts of the spectrum in the process of assimilation in plants have not led to uniform results. Some find this maximum in the yellow, and others in the red. The question is one of considerable importance. The fact recognised by Ingenhousz that the decomposition of carbonic anhydride takes place in the green parts of plants, soon led to the supposition that a connection existed between the chlorophyll colouring matter and the chemical process of assimilation; and Dumas, as early as 1844, stated the view that the violet rays, which are the principal ones absorbed by that colouring matter, must be the most active in effecting assimilation.176 Lommel, on the other hand, advocated the opinion 177 that the rays lying between the lines B and C, of Fraunhofer, might play the chief part in assimilation, because they possess the greatest intensity, and also because they correspond to a maximum of absorption by chlorophyll. Since the facts actually observed were not favourable to either of these views, Pringsheim, in connection with his investigations into the effect which light exercises upon the processes of oxidation within the plant organism, advanced the hypothesis and endeavoured to establish it that the chlorophyll colouring matter is not the chemically active substance, but that it merely serves as a screen in moderating the breathing in the plant which would otherwise become excessive.178 Draper endeavoured to prove that, during its action, the light must be absorbed (loc. at., Note 173). Bunsen and Roscoe instituted quantitative experiments on this point, from which it appears that, in the case of chlorine and hydrogen, about one-third of the rays absorbed are used up in effecting chemical work. But there are two kinds of cases which must be distinguished : namely, those in which the light must supply the energy necessary for the chemical process (which proceeds 176 Essai de statique chimique ties etres organises, 24. 177 Pogg. Ann. 143, 581. 178 Lichtwirkung und Chlorophyllfunction in der Pflanze. 1879. 328 HISTORY OF CHEMISTRY. [LECTURE XV. with absorption of heat), as in the case of the assimilation by the green parts of plants, and those in which the chemical process takes place with the evolution of heat, as in the case of chlorine and hydrogen. The light appears, however, to do work in both cases, although, in the latter case, it is merely preparatory work, by which the obstacles to combination are overcome. For this effect a certain time is required, and Bunsen and Roscoe proposed to indicate this by the term photo-chemical induction. Finally, there still remains a large department to be dealt with, namely, that of molecular physics. This department has to do with the determination of the physical constants of chemical substances, and with seeking the relations between these and the chemical composition and constitution. Her- mann Kopp may be looked upon as the founder of this branch of science. From the year 1842 onwards, he occupied himself with the determination of the boiling points and of the specific or molecular volumes of liquids.179 In order that the numbers obtained might be compared with one another, they had to be determined under comparable conditions — -the boiling-points under the same pressure, and all the specific volumes at the boiling-points, so that the corresponding vapours should be under the same pressure. The guiding idea in the comparison was that the same difference in composition corresponds to the same variation in the property under investigation, or that the particular property of a compound is the sum of the properties of its elementary constituents. The values pertaining to the atoms of the elements, with respect to this property, were cal- culated empirically, and, by means of these numbers and of the composition, the theoretical value of the property was determined for the compound. This value was then compared with that obtained by observation. Investigations of this kind were carried out in the case of molecular volumes, in particular, and harmonious results were frequently obtained. Deviations 179 Annalen. 41, 86, 169; 50, 71 ; Pogg. Ann. 63, 283; Annalen. 94» 257 J 95» "I. 307 5 9^, I, I53» 3°3 J Supplementband 5, 323 ; etc. LECTURE XV.] HISTORY OF CHEMISTRY. 329 were afterwards observed, however, and it proved necessary to take the constitutions of the compounds into consideration also ; so that the value for the atomic volume, pertaining to an atom, was assumed to be different according to the way in which the atom was combined. A means was thus furnished, in certain cases, of checking the constitution which had been deduced, in the first place, by chemical methods only. (Com- pare p. 252.) This branch, which was investigated by Kopp with great skill and success, was then followed up further by many inves- tigators. Researches into molecular volumes continue up to the present day, and the results obtained are discussed and turned to account in the same way that they were by Kopp.180 But other properties of substances were also examined, and were considered in the same way in connection with com- position and constitution. This was the case especially as regards the refraction of light by liquids and gases. Since the refractive index of a sub- stance is dependent upon the wave-length of the light as well as upon the temperature, it is not itself employed for the purpose of comparison. It is true that an endeavour was at first made to render the refractive indices independent of dispersion, by adopting as basis those for a particular wave-length. Thus Landolt at first employed, in his investigations, the indices for the C line of incandescent hydrogen. Briihl, on the other hand, employing Cauchy's formula, and after determining the refractive index for several wave-lengths, calculated a coefficient which was independent of wave-length and held for waves of infinite length.181 The next endeavour was directed towards obtaining results 180 Compare, amongst others, Pierre, Annalen. $6, 139 ; 64, 158 ; 80, 125 ; 92, 6 ; Buff, ibid. Supplementband 4, 129; Ramsay, Berichte. 12, 1024; Thorpe, Journ. Chem. Soc. 37, 141, 327; Lossen, Annalen. 214, 138 ; Elsasser, ibid. 218, 302 ; R. Schiff, ibid. 220, 71 ; etc. 181 Anna- len. 200, 1 66. In a subsequent paper (Annalen. 235, i) Briihl discards the refraction coefficient derived from Cauchy's formula and readopts the former one. 33° HISTORY OF CHEMISTRY. [LECTURE XV. independent of the temperature, by employing, for the re- fractive power, the expression discovered by Laplace 182 — 2 [# = refractive index, d= density]. It soon appeared, however, that this does not satisfy the required condition of being independent of temperature ; and besides, on the aban- donment of the emission theory of light, it had lost all physical importance. Gladstone and Dale 183 now showed empirically that the expression ^—^- fulfilled this condition, in many cases a at least. Landolt adopts the product of this value and the molecular weight (i.e., the refraction equivalent) as the basis of his extensive investigations,184 and finds that it is dependent on the constitution (the influence of chemical constitution is ascer- tained, but is not followed up). He thus succeeds in calculat- ing the refraction equivalents of the elementary atoms of carbon, hydrogen, and oxygen, and in deducing from these, again, the values pertaining to the individual compounds. These fre- quently showed close agreement with the observed values. Landolt, however, confined his observations to the fatty organic compounds. These observations were further extended, first by Haagen,185 and then by Gladstone,186 who determined the refrac- tion of many inorganic compounds and the refraction equivalents of almost all the elements. In the meantime another value, .-—-HL-, was theoretically (tf -f 2)d deduced as refraction constant by H. A. Lorentz 187 and by L. Lorenz 188 in two ways that were independent of each other ; and this value was employed especially by Landolt 189 and by his pupil Briihl. They give the name molecular refraction to the product obtained by multiplying this value by the molecular weight ; and Briihl investigated this property in the cases of strongly refracting substances, and of aromatic compounds in 182 Mecanique celeste. 4, 232. 1S3 Phil. Trans. 1858, 887 ; 1863, 317. 184 Pogg. Ann. 117, 353; 122, 545; 123, 595. 185 Ibid. 131, 117. 186 Proc. Roy. Soc. 16, 439 ; 18, 49 ; 31, 327. 187 Wiedem. Ann. 9, 641. 188 Ibid, n, 70. 189 Berichte. 15, 1031. LECTURE xv.] HISTORY OF CHEMISTRY. 331 particular.190 He arrives at the conclusion that the atomic refraction of multivalent elements is variable, and that that of carbon, for instance, is distinctly greater when double or triple carbon linkings (or unsaturated carbon valencies, as he calls them) occur in the compound. He determines the amount of the increase for one ethylene linking and for one acetylene linking, and then again calculates the molecular refractions ; and in this way he frequently arrives at numbers which coin- cide with the observed values. Later investigations of Nasini and Bernheimer191 and of Kanonnikoff 192 have only partially confirmed the conclusions of Briihl ; but the latter still hopes to be able to get rid of the exceptions.193 J. Thomsen has shown, however, that many of the values found by Brtihl can also be calculated without the assumption of double or triple carbon linking.194 These investigations attain a special im- portance from the fact that, according to the conclusions of Exner,195 the molecular refractions furnish, at the same time, the " true molecular volumes." I cannot here enter more particularly into a discussion of other investigations which are designed to show, in a similar manner, a connection between physical and chemical pro- perties ; and I shall content myself by drawing attention to individual ones. Thus there are the investigations which demonstrate a relation between the lowering of the freezing points of solutions and the molecular weights of the substances in solution (Coppet 1% and Raoult 197), and which are connected with similar earlier experiments;198 the research of G. Wiede- mann on molecular magnetism ; 199 and the investigations on the transpiration of gases by Graham,200 by O. E. Meyer,201 190 Annalen. 200, 139 ; 203, I, 255, 363 ; 211, 121, 371. m Beiblatter zu Wiedem. Ann. 7, 528 ; Accad. del Lincei [3] 18, 19, etc. 192 Berichte. 14, 1697 ; 16, 3047 ; J. pr. Chem. [2] 31, 321 ; 32, 497. 193 Annalen. 235, I. 194 Berichte. 19, 2837. 195 Monatshefte. 6, 249. 196 Ann. Chim. [4] 23, 366 ; 25, 502 ; 26, 98. 197 Comptes Rendus. 94, 1517 ; 95, 187, 1030; Ann. Chim. [5] 28, 133 ; [6] 2, 99, 115 ; 4, 401; 8, 289, 317. 198 Blagden, Phil. Trans. 1788, 277 ; Riidorff, Pogg. Ann. 114, 63 ; 116, 55 ; 145, 599. 199 Pogg. Ann. 126, i ; 135, 177. 'm Phil. Trans. 1846, 573 ; 1849, 349- 201 Pogg. Ann. 125, 586 ; 127, 253, 353. 332 HISTORY OF CHEMISTRY. [LECTURE XV. and by Maxwell,202 and on the transpiration of vapours by Lothar Meyer.203 There still remain to be mentioned, the funda- mental investigations of Biot upon the rotation of the plane of polarisation,204 and the researches connected with them, by Landolt 205 and others ; and also the experiments of Perkin on the electro-magnetic rotation of the plane of polarisation.200 Finally, I must refer in a few words to relations which have been discovered between crystalline form and chemical composition, a consequence of which may be a considerable expansion of the idea of isomorphism. The credit of having discovered these relations belongs to Groth ; 20T and his views have been extensively confirmed by means of the numerous researches by himself and his pupils. Groth follows out the changes of the axial ratios which take place upon the entrance of substituting groups, and in this way arrives at definite laws. He gave the name morphotropy to the phenomena, and caused experiments to be made in order to determine the morpho- tropic influence of definite substitutions. The morphotropic effect of chlorine, bromine, and iodine, for example, proved to be analogous to that of hydrogen ; and hence these elements have been designated isomorphotropic.208 It was then an- nounced by Hintze 209 that isomorphism might be regarded as a special case of morphotropy ; a point to which Groth had, however, already directed attention. 302 Phil. Trans. 1866, 249. '-203 Wieclem. Ann. 7, 497 ; 13, i. 'M Ann Chim. [3] 59, 206. %205 Das optische Drehungsvermogen organischer Sub- stanzen, 1879. 'm J. pr. Chem. [2] 31, 481 ; 32, 523. i207 Pogg. Ann. 141, 31 ; Berichte. 3, 449 ; compare, however, Laurent, Comptes Rendus. J5> 35°l 2°» 357 ; Methode de chimie. 156; E., 129. -208 Hintze, Pogg. Ann. Erganzungsband 6, 195. 209 Habilitationsschrift. Bonn 1884. LECTURE XVI. THE DOCTRINE OF PHASES — VAN DER WAALS'S EQUATION — THEORY OF SOLUTION — ELECTROLYTIC DISSOCIATION — ELECTRO-CHEMISTRY — ATTAINMENT OF HIGH TEMPERATURES — Low TEMPERATURES — THE NEW ELEMENTS IN THE ATMOSPHERE — THE CHEMISTRY OF NITROGEN— TRANSITION TEMPERATURE— STEREO -CHEMISTRY — RACEMISM— SYNTHESES IN THE SUGAR AND URIC ACID GROUPS — IODOSO-COM POUNDS— TERPENES AND PERFUMES— NEW NOMEN- CLATURE. WHEN we look back upon the development of chemistry during the last fifteen or twenty years, we find that it is distinguished by the constantly increasing prominence of physical or, as many call it, general chemistry, which from small beginnings has advanced to the position of a science of the first rank. Contributions to this end have, naturally, been made in particular by eminent scientists such as Horstmann, Gibbs, van der Waals, and van 't Hoff, who have devoted themselves to this department exclusively and, by their ideas and discoveries, have brought about its advancement. On the other hand, however, it can- not be denied that this advancement does not coincide for- tuitously with the appearance of Ostwald's great Text-book of General Chemistry, but that the latter, in which the attempt is for the first time successfully made to give a complete repre- sentation of what has been accomplished up to the present in this department, aroused and stimulated the tendency towards investigation in an altogether exceptional manner. Further, the establishment by Ostwald and van 't Hoff of the Zeitschrift fur Physikalische Chemie, in which all the more important investigators in this department are active as collaborators, has done a great deal to advance the subject;- so that this publication must be placed side by side with the best journals 334 HISTORY OF CHEMISTRY. [LECTURE XVI. representing our science and be looked upon as of equal value with them. In now passing on to the subject itself, I begin, in the first place, with the law of mass action, already mentioned on p. 315, which is apparently destined to play a constantly increasing part. These chemical studies received a new direction and a fresh stimulus from the theory of phases, due to Gibbs.1 The phase rule developed by him, and proved both by him and afterwards by van der Waals,2 is to the following effect : — Complete equilibrium can only exist when the number of phases present exceeds the number of components by one. By phases, are understood homogeneous portions of a system. Each state of aggregation represents at least one phase. In the solid or the liquid state, two or more different phases may exist ; a gas, however complex, can only form one phase. Complete equilibrium is a condition which depends only on the tempera- ture, and is mostly definable by a certain value of the pressure. By components are understood all those chemical elements taking part in the equilibrium, whose quantities are subject to independent variation.3 Ammonium chloride, for example, has only one component, it being a matter of indifference whether we choose nitrogen, hydrogen, or chlorine. If excess of ammonia or of hydrochloric acid is added, there are then two independent components. Calcium carbonate, above its dissociation temperature, has two independent components, calcium and carbon ; for the composition of the solid phases — calcium carbonate and calcium oxide — cannot be determined by the amount of calcium alone. Hence complete hetero- geneous equilibrium is established in the case of ammonium chloride with two phases, and in the case of calcium carbonate with three phases. 1 Trans. Connecticut Acad. 3, 108 and 343 (1876) ; German Translation by W. Ostwald, Leipzig 1892. 2 Rec. Trav. Chim. 6, 265, communicated by Roozeboom. a I follow here the exposition by Planck (see Article Thermochemie, in Ladenburg's Handworterbuch der Chemie. II, 636). LECTURE XVI. ] HISTORY OF CHEMISTRY. 335 If there are n + 2 phases and only n components, equilibrium is only possible at singular points ; that is to say, at some definite temperature (multiple point, transition or transforma- tion temperature). If there are just as many phases as there are components, the equilibrium is incomplete ; that is, to each temperature there corresponds a series of pressures. This phase rule has received numerous practical applica- tions, the work of Roozeboom 4 deserving especial mention. Roozeboom studied the connection of the states of aggregation, the equilibrium between water and sulphurous anhydride, the hydrates of ferric chloride, etc. The phase rule can also be applied to dissociation phenomena, to the reciprocal trans- formation of allotropic modifications of elements, and so forth.5 More important perhaps than the phase rule (the significance of which is exaggerated by many, seeing that it merely furnishes a scheme for the representation of heterogeneous equilibrium) are van der Waals's theories of corresponding conditions,5* and van 't Hoff's theory of solution.6 Van der Waals makes a distinct advance by substituting for the gas equation, PV=RT, deduced from the laws of Boyle-Mariotte and of Henry-Gay- Lussac, the expression, in which a and b are constants which depend upon the co- hesion of the gases and the not altogether negligible volume of the molecules. (According to van der Waals b is to be con- sidered as representing four times the volume of the molecules.) This equation not only represents the behaviour of gases 4 Z. physik. Chem. 2, 449, 513 ; 4, 31 ; 5, 198 ; 10, 477 ; Rec. Trav. Chim. 4 et seq. 5 Compare the summaries by Meyerhoffer, Leipzig 1893, and by Bancroft, Ithaca, New York, 1897. 5a Die Continuitat des gas- formigen und fliissigen Zustandes, Leipzig 1881. 6 Lois de 1'equilibre chimique dans 1'etat dilone ou dissous. Stockholm iS86. Abstracted, Z. physik. Chem. I, 481, 33 6 HISTORY OF CHEMISTRY. [LECTURE XVI. (and especially of compressed gases) more satisfactorily than the original equation, but it is also capable of application to liquids. Moreover, since the constants a and b can be deter- mined in a simple manner from the critical data (volume, pressure, and temperature) or from the behaviour of the gases under high pressure, van der Waals's equation furnishes a mode of giving expression to the entire behaviour of all homogeneous liquid and gaseous substances with respect to changes of pres- sure, temperature, and volume ; and, on this account, it may be regarded as of fundamental significance. Its accuracy has been proved by Young Ga in particular. The theory of solution is based upon conceptions that have arisen from the well-known experiments of Pfeffer,7 which latter only became possible after the discovery by Traube 8 of semi- permeable membranes. In explaining osmotic pressure as the result of the impacts of the dissolved molecules upon the walls of the vessel, van 't Hoff arrives at a comparison between substances in the dis- solved condition and in the state of gas. The laws of Boyle- Mariotte, and of Henry-Gay-Lussac, as well as the fundamental hypothesis of Avogadro, can now be applied directly to solu- tions ; so that this branch, which has hitherto been one of the most obscure in the whole subject of chemistry, at once becomes fully accessible to investigation. As a consequence, important results, which are capable of being turned to account throughout the whole range of chemistry, are immediately obtained. The important relations subsisting between the depression of freezing point, the diminution of vapour pressure, and the elevation of boiling point on the one hand, and the molecular weight of the dissolved substance on the other (which were ascertained experimentally and formulated by Raoult 9 in 6a Phil. Mag. [5] 33, 153 ; 34, 505. " Osmotische Untersuchungen, Leipzig, 1877. 8 Archiv. f. Anat. u. Phys. 1867, 87. 9 Ann. Chim. [6] 2, 66, 99: 8, 289, 317; 2O, 297; Comptes Rendus. 87, 167; Z. physik. Chem. 9, 343, etc. The literature of the predecessors of Raoult is very fully given in Ostwald's Lehrbuch der Allgemeinen Chemie, Second Edition, I, 705 and 741. LECTURE XVI.] HISTORY OF CHEMISTRY. 337 particular), now attain their theoretical significance for the first time. As the outcome of this, and also in consequence of improvements and simplifications that Raoult's methods of molecular weight determination underwent,10 these methods very soon obtained a footing; and their results, especially those from the depression of the freezing point, are considered to be just as accurate as those from the vapour density. Raoult had already pointed out, however, that aqueous solutions of salts, of bases, and of acids, in particular, did not agree with his rules ; but always yielded results that were too low, and only attained to a value from one half to one-third of that which had to be regarded as the normal number. All explanation of this anomaly was at first wanting, so that the general applicability of van 't Hoff's theory appeared to be placed in doubt. The difficulty was got rid of in the same way as in the case of the abnormal vapour densities (compare p. 304). Arrhenius dealt with this matter by exactly the same method that Cannizzaro, Kekule, and Kopp had adopted in solving the other difficulty. His theory, advanced in iSSy,11 adopts as actually existent that condition which must be assumed to exist in order to arrive at an agreement between the theory of van 't Hoff and the numbers furnished by Raoult's rules. He draws attention to the fact that it is in the cases of solutions of those substances which are electrolytes and break up, under the influence of the electrical current, into their ions, that numbers are obtained which do not agree with theory. He now assumes that the ionisation does not merely take place as a result of the passage of the current, but that it occurs during the dissolution; and that the latter is thus accompanied by a more or less com- plete (electrolytic) dissociation, the extent of which depends 10 Compare especially, Beckmann, Z. physik. Chem. 2, 638 ; 4, 532 ; 8,223; 18,473; etc. u Z. physik. Chem. I, 631. Clausius (Pogg. Ann. 101, 338) and Helmholtz (Weidem. Ann. II, 737) must be mentioned as predecessors of Arrhenius. Planck (Z. physik. Chem. I, 577) also clearly stated the idea of the dissociation of salts in aqueous solution simultaneously with Arrhenius. Y 33$ HISTORY OF CHEMISTRY. [LECTURE XVI. principally upon the degree of dilution. A number of methods for determining the extent of this dissociation very soon pre- sented themselves, as was pointed out by Arrhenius himself,12 and also by Planck,13 Ostwald,14 and others ; and (what is very important) these methods give results that agree with one another. The hypothesis of Arrhenius found a great many opponents — indeed it could hardly have been expected that it would be otherwise. The assumption that an aqueous solution of common salt contains free sodium and chlorine ions (which, however, are nothing but electrically charged atoms that be- have like free molecules) was certain to meet with opposition from chemists, since it stood in contradiction to observation and thus included something of a metaphysical nature. Be- sides, the explanation of many reactions that had formerly appeared simple Wias rendered much more difficult; as, for example, the decomposition of water by the alkali metals,15 since in this reaction no combination with oxygen and, on the other hand, no displacement of hydrogen ions by sodium ions could be assumed. But of what consequence are con- siderations of this kind in face of the great advantages which the theory of electrolytic dissociation affords ? A large number of otherwise inexplicable facts are satisfactorily explained by means of it. The so-called law of thermo-neutrality, of Hess,16 which has been confirmed, in part at least, by the well-known investigations of Thomsen ir and of Berthelot,18 is in complete accord with the theory of ionisation, and so are the exceptions to this law which must necessarily exist in cases of incomplete dissociation ; whereas, without this theory the facts concerned constitute an incomprehensible puzzle.19 It is similar with the identity of the heat of neutralisation of one and the same acid by means of different bases, and 12 Z. physik. Chem. 2, 491. ia Wiedem. Ann. 34, 139. 14 Z. physik. Chem. 2, 36 and 270. 15 Compare however Ostwald, Lehrbuch. Second Edition, 2, 989. 16 Pogg. Ann. 52, 97. . 17 Thermochemische Unter- suchnngen. I," 63. 18 Ann. Chim. [5] 6, 325. 19 Compare L. Meyer, Z. physik. Chem. I, 134. LECTURE XVI.] HISTORY OF CHEMISTRY. 339 of that of one and the same base by means of different acids ; also with the law of Oudemans 20 and Landolt 21 (in accordance with which the salts of optically active alkaloids and of optically active acids exhibit the same rotation in solu- tions of equivalent concentration), with the magnetic rotatory power,22 and with the atomic magnetism.23 Further, the principle in accordance with which the spectra of dilute solutions of different salts with similarly coloured ions are identical,24 and that according to which the molecular re- fractive power of the salts present in aqueous solution is an additive property,25 are explained in the same way. But probably the most important fact of this kind is that of the proportionality that exists between electrolytic conductivity and avidity in the case of acids,20 with which may be coupled the proof, furnished by Arrhenius,27 that the extent of the dissociation calculated from the electrolytic conductivity leads to very nearly the same results as that calculated from the depression of the freezing point. In these circumstances we cannot be in doubt as to whether the hypothesis of Arrhenius is warranted. This ionisation theory, as it is now commonly called, leads us directly to electro-chemistry, which has made advances that were undreamt of twenty years ago, and has now developed into a separate branch of science that constantly leads to new scientific and practical results. The enthusiasm with which the discovery of the galvanic current and of the voltaic pile was welcomed, as sketched in Lecture V., was, as we now know, perfectly justified. And even although disillu- sionment followed the great discoveries of Ritter, Davy, Berzelius, and Faraday, and although this branch remained unproductive for decades, still the opinion has been verified 20 Wiedem. Beibl. 9, 635. 2l Berichte. 6, 1073. ^ Jahn, Wiedem. Ann. 43, 280. > Annalen. 82, 137; Pogg. Ann. 91, 619; 92, 648 ; Annalen. 94, 107 etc. 342 HISTORY OF CHEMISTRY. [LECTURE XVI. became possible after the discovery, in 1872, of the dynamo- electrical machine, which was employed immediately thereafter (in the North-German Refinery at Hamburg) to remove copper from solutions. Other metals, such as zinc, magnesium, lead, silver, gold, etc., were also produced electrically afterwards. An operation of especial importance was the electrolytic pro- duction of aluminium, a metal which Bunsen first prepared by this method.44 The technical process of Heroult 4r> is different, however, from that of Bunsen, inasmuch as it is not a fused double chloride of the metal that is electrolysed, but aluminium oxide. This is the place to refer to the great scientific and prac- tical results that Moissan obtained as the outcome of his experiments with the electric furnace.46 Specially worthy of mention in this connection are the preparation of artificial diamonds ; the production of calcium carbide (which had, however, been discovered long before by Wohler4") and of many other carbides ; the preparation, in a state of purity, of chromium and of other difficultly fusible metals, etc. The first preparation of carborundum, which is also frequently attributed to Moissan, is due rather to Acheson,48 an American. Atten- tion must be drawn to tne facts that in many of these experi- ments electricity is only employed as a means of attaining to high temperatures (3,000° to 4,000°), and that the results can also be obtained in other ways, since the same high tempera- tures can, of recent years, be reached by means of chemical reactions. An entirely new branch of thermo-industry has thus arisen, by means of which great advances have already been made, and are still to be expected, in metallurgy. Of an earlier date is the employment of the oxy-hydrogen blowpipe in the melting and working of platinum,49 and so is the com- bustion of carbon and other elements (such as silicon, sulphur, 44 ?°gg- Ann- 92> 648- 45 German Patent, December, 1887. 4(i Le Four Electrique, Paris 1897 ; Comptes Rendus. 115, 1031 ; Il6, 218, 1429; 117, 425, 679 ; 118, 320, 501 ; etc. 47 Annalen. 124, 220. 48 Compare also Schutzenberger, Comptes Rendus. 114, 1089. 49 Hare, Phil. Mag. [3] 3J> 356 ; further Deville and Debray, Ann. Chira. [3] 56, 385. LECTURE XVI.] HISTORY OF CHEMISTRY. 343 phosphorus, etc.) in air or oxygen at high temperatures, for the purpose of attaining still higher temperatures ; as, for example, in the blast furnace, or in the ingenious Bessemer process. The development of these methods by Goldschmidt,50 and their application to the production of metals such as chromium, manganese, iron, and nickel, free from carbon, and of a large number of alloys, are new however. 1 must here recall the interesting results obtained, partly by Victor Meyer51 and partly by Crafts,52 by the application of the method of vapour density determination devised by the former.53 I regard as worthy of mention the proof that the molecule of iodine, I2, breaks up at high temperatures into single atoms ;54 and also the facts that the beginning, at least, of a similar dissociation has been ascertained in the case of bromine;55 that the molecule of arsenic, As4, similarly splits into two; that potassium iodide even at high temperatures corresponds to the formula KI, and cuprous chloride to the formula Cu.2Cl.2, etc. If the attainment of high temperatures has thus been of service for the purposes of our science and of technology, so likewise the endeavours, on the other hand, to obtain low temperatures have led to great advances, and to results of altogether unforeseen importance. In the preceding lecture (p. 322), where the inter-relations of the states of physical aggregation and the significance of the critical temperature are referred to, the results of Pictet, of Cailletet, and of Wroblewsky on the liquefaction of the so-called permanent gases are stated. Of especial importance were the detailed investigations of Wroblewsky and Olszewsky, who first obtained quantities of oxygen and nitrogen in the liquid state, and particularly de- scribed many of their properties.5" The mode of measuring 50 Annalen. 301, 19; Z. f. Elektrochemie, 1897-98, Heft 21. B1 Be- richte. 13, 1010. 52 Comptes Rendus. 90, 183 ; 92, 39 ; Berichte. 13, 851. 5:i Berichte. II, 1867 and 1946 ; 12, 609 and 681 ; etc. 54 Comptes Rendus. 90, 183 ; 92, 39 ; Berichte. 13, 851. 55 Langes and Victor Meyer, Fyrochemische Untersuchungen, Braunschweig 1885. r'6 Wiedem. Ann. 20, 243 and 860; Wien. Akad. Ber. 1885, 91 (2), 667 ; Monatshefte. 9, 1067. 344 HISTORY OF CHEMISTRY. [LECTURE XVI. temperatures by determining the potential of thermo-electric currents, which is now largely employed, also originated with them.57 In the experiments that have been carried out latterly, however, on the liquefaction of air and of other gases, Pictet's method has been abandoned again, and recourse has been taken to that of Cailletet (the latter method having been con- verted into a dynamical or continuous one) ; that is to say, the expansion of highly compressed gases has been employed in order that the necessary lowering of temperature may be effected. Thus Dewar,58 in his experiments upon the produc- tion of liquid air, liquefied, by its own expansion, air which was under a pressure of 100 atmospheres and was cooled by solid carbonic anhydride ; whereas the recent technical method con- sists in cooling exclusively by expansion, and the effect of the latter is turned to account in a very ingenious manner by the employment of a self-intensive apparatus. Linde 59 in Germany, and Hampson60 in England almost at the same time con- structed technically efficient forms of apparatus, based upon this method, for the production of liquid air. Liquid air has not as yet, however, found any technical application upon the large scale. Nearly pure oxygen is ob- tained from it very cheaply, and the attempt has been made to apply it in the technology of explosives, or to the production of high temperatures, but no ultimate pronouncement can be made with respect to this. Of far greater importance are the results that liquid air has achieved in scientific investigation, In the first place, it must be mentioned that Dewar, by its aid, has succeeded in liquefying helium,01 and in obtaining air, oxygen, and hydrogen in the solid state ; and that in doing so he has achieved almost everything that can be done in this direction. Dewar is at present engaged in trying to reach still 57 Compare Holborn and Wien, Wiedem. Ann. 59, 220 ; and Lad en burg and Kriigel, Berichte, 32, 1818. r>8 Journ. Royal Institution, 1878 ; 1883-1885; 1892-1899. 59 Z. d. Vereines deutscher Ingenieure, 39, 1157. 60 British Patent, April 1896. 61 The Times, nth May 1898; Phil. Mag. [5] 45, 543 ; Comptes Rendus. 126, 1408 ; Ann. Chim. [7] 14, 145 ; Proc. Chem. Soc. 14, 129, 146. LECTURE XVI.] HISTORY OF CHEMISTRY. 345 lower temperatures by the aid of liquid hydrogen boiling under low pressure, in order to approach as nearly as possible to the absolute zero.02 It is also noteworthy that ozone, which was obtained in the liquid state by Hautefeuille and Chappuis in 1882 by the aid of liquid ethylene,63 can easily be prepared in an approxi- mately pure condition by the use of liquid air, so that Troost was able to determine its boiling point64 and Ladenburg its density.'55 The latter determination is of especial importance, since the molecular formula O3, deduced from it, constitutes one of the most emphatic arguments in favour of the whole molecular theory ; and this formula, which till then had only been supported by Soret's experiments,66 could not be regarded as finally settled. But the results that have been furnished by this agency with respect to the discovery of new elements are almost of greater consequence. When Lord Rayleigh compared the relative density of atmospheric nitrogen with that of nitrogen prepared from ammonia and other nitrogen compounds, he found a differ- ence (in the third decimal place) which could not possibly be ascribed to an experimental error.67 He therefore resolved upon a minute investigation in order to find out the substance that was mixed with atmospheric nitrogen. This investigation he then carried out along with Ramsay, and it led to the discovery of argon, an element of which it is very difficult to obtain any compounds.68 The molecular weight, deduced from the density, gave the number 39-92,69 and since by Kundt's method (compare p. 300) the monatomic character of the gaseous molecules was indicated, its atomic weight would be represented by the same number. The question as to the position of this element in the periodic system is thereby 62 Proc. Roy. Soc. 64, 227 ; Ann. Chim. [7] 17, 5. w Comptes Rendus. 94, 1249. 64 Ibid. 126, 1751. 6r> Berichte. 31, 2508, 2830; 32, 221. fic Annalen. 138, 45 ; Supplementband 5, 148. 67 Nature, 46, 512. {58 Rayleigh and Ramsay, Proc. Roy. Soc. 57, 265 ; Z. physik. Chem. 16, 344; Phil. Trans. 1895 (A), 187. 69 Berichte. 31, 3121. 346 HISTORY OF CHEMISTRY. [LECTURE XVI. rendered an extremely difficult one, since it falls near that of potassium and yet is beyond it. Ramsay took up the problem from a very general point of view. It appeared to him highly probable that argon was a member of a whole group of elements, of which group he hoped to find additional members associated with nitrogen. It was thus that he came to investigate, amongst other things, the gases evolved from cleveite by heating with sulphuric acid, which Hillebrandt had considered to be nitrogen,70 and this led him to the discovery of helium. The brightest line in the spectrum of this gas, D3 (Dj and D.2 are the sodium lines), had been observed a long time previously by Lockyer in the spectrum of the sun's photosphere.71 Helium, whose atomic weight 4 was deduced from the density of the gas and from the rate of propagation of sound in it, was an analogue of argon in every respect ; and it was thus clear to Ramsay that there must be another element which, with atomic weight about 20, should be placed before sodium, in the same way that helium comes before lithium, and argon probably before potassium, although the atomic weight of argon has been found, in the meantime, somewhat higher than that of potas- sium.72 A similar thing applies to tellurium, the atomic weight of which, according to the most recent determinations, is greater than that of iodine.73 Ramsay now represents the further development of the subject74 as if the investigation, carried out with his utmost energy and effort, had remained unproductive, and as if an accident only had led him on to his further discoveries. There is in reality, however, no such accident in question, for the investigation of the residue from the evaporation of liquid air was only a link in the chain which, although perhaps unknown to himself, represented the course of his ideas. In this way he discovered crypton, the molecular weight of which was 70 Bull. U.S. Geological Survey, 78, 43- 71 Nature, 53, 319. 72 Berichte. 31, 3111. 7:$ Brauner. Journ. Chem. Soc. 67, 549. 74 Berichte, LECTURE XVI.] HISTORY OF CHEMISTRY. 347 ascertained in a preliminary manner to be 45. It should pro- bably be much higher, however, as the gas was still mixed with lighter gases, especially with argon. In the case of crypton, the ratio of the specific heats has also been ascertained to be 1.66, so that this gas is also a monatomic element, the position of which in the periodic system is still undetermined. As regards other discoveries, Ramsay found, by the syste- matic fractionation of argon 7r' (which he condensed by means of liquid air), three new substances which he considers to be elements. These are, — neon, with atomic weight 19.3 to 19.5, which is clearly to be placed therefore between helium and argon and before sodium; xenon, with density 65 (H = 2) which might, as Ramsay supposes, be raised to 81 by further purification, so that it would be placed beyond bromine ; and finally, metargon, an easily condensable and even solidifiable gas, which shows the spectrum of carbonic oxide even after it has been mixed with oxygen and exposed for a long time to the passage of electric sparks.76 Even although all doubt as to the individuality and the elementary nature of these gases is not yet removed,77 still these investigations are unquestionably amongst the most suc- cessful that have been carried out during the last twenty years. Liquid air served not merely as starting material for the investigations, but Ramsay also employed it, or at least the liquid oxygen obtained by its aid, in an ingenious manner for the purpose of separating the various new elements. The question as to the position of these " elements " in the periodic system has been much discussed, and up to the present it is not finally solved. On the other hand, we may now say that even if our views respecting the connection between the properties of the elements and their atomic weights should be modified on account of these newly discovered facts, still the periodic law has rendered excellent service as an invaluable guide in this obscure region. 7f) Berichte. 31, 3117. 76 Ibid. 31, 3119. 77 Compare Brauner, Ibid, 32, 708. 34-8 HISTORY OF CHEMISTRY. [LECTURE XVI. Although such unexpected discoveries were thus made, still they will not exercise any considerable influence upon chemistry as a whole, since all these "elements" apparently resemble argon, and probably do not enter into many com- pounds. Hence it may be said that these interesting investiga- tions will probably not prove of great significance, as regards their consequences, and that in this respect they will fall short of other researches which have not excited the interest of such wide circles. I merely recall here the isolation of fluorine by Moissan in i886,78 and the discovery of nickel carbonyl and analogous compounds by Mond in 1890, and pass on to consider more particularly the investigation of the chemistry of nitrogen, which has made great advances in recent years. The discovery of hydroxylamine, by Lessen, falls under review here, although, of course, it took place at a much earlier date (in i865).79 It has not been referred to previously, how- ever, since its importance only came to be recognised gradually, a result to which Victor Meyer's researches on the oximes 80 and their stereo-isomerism 81 materially contributed. The preparation of phenylhydrazine, by Emil Fischer,82 also deserves mention here. It must be looked upon as of par- ticular importance, on account of its leading to the clearing up of the sugar group.83 Following upon this there are the valuable researches of Curtius, who discovered hydrazine in iSSg,84 and hydrazoic acid in iSgo.85 The utilisation of these two substances has already led to numerous investigations, and will lead to others. As worthy of mention, I also refer to the 78 Comptes Rendus. 103, 202 and 256. 79 Zeitschrift fur Chemie. 8, 551 ; Annalen, Supplementband 6, 220 ; 160, 242 ; 161, 347 ; etc. 80 Meyer and Janny, Berichte. 15, 1324; Janny, Ibid. 15, 2778; 16, 170; Meyer, Ibid. 16, 822; Petraczek, Ibid. 16, 823; etc. 81 H. Goldschmidt, Berichte. 16, 2176 ; Auwers and Meyer, Ibid. 21, 784, 3510 ; 22, 537 ; etc. ^ Berichte. 8, 589 ; compare also Strecker and Roemer, Ibid. 4, 784 ; and Zeitschrift fiir Chemie, 14, 481. 8:>> Berichte. 17, 579. 84 Curtius and Jay, J. pr. Chem. [2] 39, 27. *r> Curtius, Berichte. 23, LECTURE XVI.] HISTORY OF CHEMISTRY. 349 researches of Thiele,80 who (amongst others) found out a con- venient and technically practicable method for the manufacture of hydrazine ; and to those of Raschig,87 who cleared up the nitrogen-sulphonic acids, and in doing so discovered the method now employed for the production of hydroxylamine. It does not seem to me that this is the place to enter more fully into this subject, since I am really giving a historical sketch, in which only those things that are of general importance can be prominently brought forward. I may thus recall here a discovery of HellriegePs which marks an epoch in chemistry and agriculture.88 According to Hellriegel, leguminous plants, and lupins in particular, possess the power of assimilating, with the aid of lower organisms, the nitrogen of the air. In this connection the fact must not be passed by without mention that Berthelot had previously asserted the assimilation of free nitrogen.89 An observation which is to a certain extent of an opposite character is the proof furnished by Buchner that fermentation is possible even without living organisms, by means of the liquid expressed from yeast (zymase).90 More particular consideration may be given to a research by van 't Hoff, in which the idea and the significance of the transition temperature are clearly stated.91 Van 't Hoff is led to the idea by the comparison of chemical reactions with the transitions from one of the states of physical aggregation to the others ; but the same conception may be arrived at by the aid of the phase rule. Since the observations of St Claire Deville (see p. 302), the phenomena of dissociation have been regarded and treated as analogous to those of evaporation. Van 't Hoff now shows that there are reactions which are comparable with the process of fusion, and in which a fixed temperature marks the line of 86 Annalen. 270, I ; 273, 133 ; Berichte. 26, 2598 and 2645 ; etc. 87 Annalen. 241, 161. 88 Hellriegel and Wilfahrt, Biederm. Centr. 18, 179. 89 Comptes Rendus. 106, 569. ** Berichte. 30, 117, mo, 2668, etc. 91 Van 't Hoff and Deventer, Ibid. 19, 2142. 35° HISTORY OF CHEMISTRY. [LECTURE XVI. separation between two chemically different conditions. This fixed temperature he designates the transition temperature ; and he demonstrates the accuracy of his idea in the cases of the formation of double salts (astrakanite), of the preparation of elements in allotropic modifications (sulphur), and of the splitting of racemic substances (sodium ammonium racemate). He afterwards treated this subject in a much more detailed manner in an important monograph02 "On the Formation and Decomposition of Double Salts," in which he explains the theory of the matter, and describes the methods for experi- mentally determining the transition temperature. These investigations have found very important applications in relation to the deposition of salts from ocean water,93 and in explaining the splitting of racemic compounds by Pasteur's methods. This leads us directly to the subject of stereo-chemistry, which has already been discussed in Lecture XIII. (see p. 268), but which has acquired so much importance of late that I must return to it here.04 After the propounding of the theory by van 't Hoff and Le Bel, it was only isolated investigations that were, in the first place, carried out with a view to. testing it — such, for example, as the splitting by means of fungi of a series of alcohols, which Le Bel succeeded in doing ; {r° and the splitting of synthetic coniine,96 which was of importance inasmuch as it was the preparation, for the first time, of an active base. The theory was subjected to a systematic examination by Emil Fischer, in carrying out his well-known syntheses in the sugar group.9" It is simply astonishing that the theory stood the test of this experimentum cruets, and that the sagacity of Fischer enabled him to fix the configurations of the individual hexoses 9S with- 92 German Edition by Dr Paul, Leipzig 1897. 93 Berlin. Akad. Ber. 1897, 1898, 1899. 94 Compare van 't Hoff: Die Lagerung der Atome im Raume, Second Edition, Braunschweig 1894. 95 Comptes Rendus. 87, 213 ; 89, 312 ; Bull. Soc. Chim. [3] 7, 551 ; Comptes Rendus. 92, 532. 9(5 Ladenburg, Berichte. 19, 2578 ; Annalen. 247, 83. 97 Berichte. 23, 2114; 27, 3189. 98 Ibid. 24, 1836 and 2683. LECTURE XVI.] HISTORY OF CHEMISTRY. 351 out encountering any contradictions in doing so, especially when we take into consideration the recent experiments of Walden," in accordance with which it is possible, by means of simple chemical reactions, to pass at ordinary temperatures from an active substance to its enantiomorph. As regards the application of the theory of asymmetric carbon atoms to molecules with doubly linked carbon atoms,— a matter that van 't Hoff had already mentioned, but one which had met with less attention, — very special notice was called to it by Wislicenus,100 who, moreover, had himself given the first impulse to stereo-chemical conceptions by his earlier and extended investigations of lactic acid.101 The researches of Wislicenus and his pupils 102 have certainly supplied most valuable contributions towards the clearing up of these remarkable cases of isomerism — an end towards which (after the discovery of fumaric 103 and maleic 104 acids) a great many chemists, and even Kekule himself,105 had aspired in vain. In this domain, however, there are still many unexplained contradictions, as Michael 10(5 and Anschiitz 107 in particular have shown. On the other hand it must be admitted, that by van 't Hoff's theory an extremely plausible explanation of the products arising from the oxidation of fumaric and of maleic acids is rendered possible.108 The applications of the doctrine of asymmetric carbon atoms to substances containing rings are also important and interesting. The first principles were laid down by van 't Hoff; but their significance was only fully recognised when Baeyer published his extended investigations upon hydrogenised 99 Berichte. 28, 2766; 29, 133; 30, 2795 and 3 146. 10° Uber die raumliche Anordnung der Atome in organischen Molekulen und ihre Bestimmung in geometrisch - isomeren ungesattigten Verbindungen, Leipzig 1887. 101 Annalen. 125, 41 ; 128, I ; 133, 257 ; 146, 145 ; 166, 3 ; and especially 167, 345- 102 Ibid. 246, 53 ; 248, i, 281 ; 250, 224 ; 272, i ; 274, 99. 10:5 Pfaff, in Berzelius' Jahresbericht. 1828, 216. 104 Pelouze, Annalen. II, 263. 105 Annalen. Supplementband 2, in ; Zeitschrift fiir Chemie. 10, 654. 106 J. pr. Chem. [2] 38 ; 43 ; 46 ; 52; etc. lo7 Annalen. 254, 168. 108 Kekule and Anschiitz, Berichte. 13, 2150 ; 14, 713. 35 2 HISTORY OF CHEMISTRY. [LECTURE XVI. aromatic compounds, and, in particular, upon the hydro- phthalic acids.109 While Baeyer's intention in these investigations was to discover weaknesses in the theory, and even to modify it, his labours led instead to a further confirmation of it. Besides this, the credit is due to him of having advanced the so-called tension theory,110 which has already proved of service in some cases. Emphasis must be laid upon the fact that the important consequences which the theory of the asymmetric carbon atom brought forth, gave a spur to the more and more complete application of stereo-chemical considerations. In this con- nection the numerous researches may be mentioned which deal with the non-occurrence of certain reactions, and explain this on stereo-chemical grounds.111 Amongst these investi- gations the best known are those of Victor Meyer on the formation of esters.112 The asymmetry of the nitrogen atom may also be mentioned in this connection. The researches of Hantzsch and Werner 113 were of funda- mental significance with respect to the last-named subject, and they were capable of explaining the isomerism amongst oximes, which was already familiar at that time. Hantzsch afterwards extended the views respecting this matter, and turned them to account in explaining the isomeric hydrazones114 and diazo- compounds.115 It is true that it was only geometrical isomerism 109 Annalen. 245, 103 ; 251, 257 ; 156, I ; 258, I and 145 ; 266, 169 ; 269, 145 ; 276, 255. no Berichte. 18, 2278. m Hofmann, Ibid. 17, 1915; and 18, 1825; Jacobson, Ibid. 22, 1219; 2$, 992; 26, 681 and 699; etc. ; Pinner, Ibid. 23, 2917; Kuster and Stallberg, Annaien. 278, 207. 113 Berichte. 27,^510, 1580, 3143; 28 Kef. 301 and 916; 29, 830 ; etc. 113 Ibid. 23, 1 1 ; Werner, Raumliche Anordnung der Atome in stickstoffhaltigen Molekiilen, 1890. Compare further, the previously published researches of Willgerodt, J. pr. Chem. 37, 449 ; Burch and Marsh, Journ. Chem. Soc. 55, 656 ; and especially van 't Hoff, Ansichten liber die organische Chemie, Braunschweig 1878-81. 114 Fehrlin, Berichte. 23, 1574 ; Krause, Ibid. 23, 3617 ; Hantzsch and Kraft, Ibid. 24, 3511 ; Marckwald, Ibid. 25, 3100. 115 Ibid. 27, 1702, 1726, 1857, 2099, 2968, 3527 ; 28, 741, "24, 1734; etc. LECTURE XVI.] HISTORY OF CHEMISTRY. 353 in the case of nitrogenous organic compounds that was proved by these investigations. Le Bel116 and Ladenburg117 en- deavoured to prove that asymmetric nitrogen can further pro- duce or influence optical activity. The investigations of both have, however, been attacked,118 but they have been able to establish the accuracy of their results.119 Another subject that was much discussed was the signi- ficance of racemism, about which a clear understanding only became possible upon the introduction of the conception of the transition temperature, and upon the recognition of the analogy between racemic substances and double salts. The most important method of splitting racemic substances— that by means of optically active substances— remained a standing enigma as long as the existence of partially racemic substances was denied.120 Every difficulty was removed, however, after Ladenburg had shown that such substances do without doubt exist,121 and after a transition temperature had been recognised in their case also.122 Furthermore, the much debated question as to how a truly racemic substance (inactive by intra-molecular compensation) can be distinguished from the mixture of the active components, may now be looked upon as practically settled.123 It is beyond doubt that the founding and development of stereo-chemistry (a name which originated with Victor Meyer124) is the most important thing that has been accom- plished in organic chemistry during the last two decades. Stereo-chemistry possesses a significance for this period similar to that which the foundation and introduction of the theory of 116 Comptes Rendus. 112, 724. m Berlin. Akad. Ber. 1892, 1067; Eerichte. 26, 854 ; 27, 853 and 859. 118 Marckwald and Droste- Huelshoff, Ibid. 32, 560 ; Wolffenstein, Ibid. 29, 1956. 119 Comptes Rendus. 129, 548 ; Berichte. 29, 2706. 1>2° E. Fischer, Berichte. 27, 3226 ; Landolt, Das Optische Drehungsvermogen, Second Edition, Braunschweig, 1898, 85. 121 Ladenburg and Herz, Berichte. 31, 937 ; Ladenburg and Doctor, Ibid. 31, 1969. l~2 Ladenburg and Doctor, Ibid. 32, 50. 123 Roozeboom, Z, physik. Chem. 28, 494 ; Ladenburg, Journ. Chem. Soc. 75, 465. m Berichte. 23, 568. Z 354 HISTORY OF CHEMISTRY. {LECTURE XVI. aromatic compounds possessed for the twenty years preceding. There are besides, however, other important investigations in organic chemistry which require to be mentioned. A matter of general importance was the introduction of the idea of tautomerism or desmotropy, which was brought forward by Laar,125 in 1885, on the strength of some experimental obser- vations and remarks by Zincke.126 Laar applies the term tauto- meric to a compound when two or more structural formulae can be advanced in explanation of its interactions. A very well-known example is furnished by aceto-acetic ether, which reacts sometimes as if it should be represented by the ketone formula CH8.CO.CH.2.COOC2H5, and Sometimes as if it should be represented by the enol formula CH3.C(OH) :CH.COOC2H5. There are numerous investigations dealing with substances of this kind, of which there are a large number. Some of the best known of these investigations are those of Claisen,127 of W. Wislicenus,128 and of Knorr.129 Opinions are still widely divergent, with regard to the questions as to whether a desmo- tropic substance is to be considered as a mixture of two or more compounds (Laar), or whether the forms are continuously passing into one another by means of oscillations (Kekule) or shifting Unkings (Knorr 13-°), or finally whether one form is stable under certain conditions while another is stable under different conditions. The systematic and, theoretically, almost completed nation of the sugar group has already been referred to The uric acid group, which so long resisted elucidation and synthesis, is now completely cleared up,131 and this is chiefly due to Emil Fischer's synthetical investigations.132 The hydrogenised aromatic compounds have likewise been referred to already (p. 351), but the terpenes have not been 125 Berichte. 18, 648 ; 19, 730. 12G Ibid. 17, 3030. ™ Annalen. 291, 25. 128 Ibid. 291, 147. 129 Ibid. 293, 70. 13° Ibid. 279, 1 88. 1:J1 Compare further, Grimaux, Ann. Chim. [5] u, 356; and 17, 276; Horbaczewsky, Monatshefte. 3, 796 ; 6, 356 ; 8, 201 ; Behrend and Rosen, Annalen. 251, 235. 132 Berichte. 30, 549, 559, 1839, 1846, 2220, 2226, 2400, 3009 ; 31, 104, 431, 542, 1980, 2546, 2550, 2619, 2622 ; 32, 435. LECTURE XVI.] HISTORY OF CHEMISTRY. 355 mentioned. The latter formerly constituted one of the most confused sections of organic chemistry, whereas Wallach has now succeeded in systematising them by his extended and careful researches.133 But the most important thing about them, the elucidation of their constitution, is still wanting ; for, in spite of some fortunate attempts by Baeyer,134 which led to the synthesis of substances resembling terpenes, no one has yet succeeded in making this clear. The discovery of the iodo-, iodoso-, and iodonium-com- pounds, for which we are indebted to Willgerodt 13rj and to Victor Meyer,1315 is also important, and these compounds supply new knowledge concerning the nature of iodine. The dis- covery of antipyrine by Knorr 13T was of great importance in medicine, and through it the pyrazol group 138 came to be simultaneously explored. The preparation of the so-called substantive azo-dyes has become of technical importance;139 and that of synthetic indigo, according to a method discovered by Heumann,140 promises to become so. Great advances have likewise been made in the preparation of artificial perfumes. Vanilline has already been referred to. The manufacture of piperonal U1 (heliotropine), and especially the synthesis of ionone by Tiemann and Kriiger,142 must be mentioned in this connection. This account of the most recent phases in the develop- ment of our science must not be concluded, however, without reference being made to the valuable, although unfinished, re- searches which were carried out under the direction of Friedel, and which aimed at the introduction of a new nomenclature into organic chemistry.143 Although it has not yet been 133 Annalen. 225-306 (46 papers). 134 Ibid. 278, 288 ; Berichte. 26, 232. 1:w J. pr.Chem. [2] 33, 154 ; Berichte. 25, 3495 ; and 26, 1802. 136 Ibid. 25, 2632 ; 26, 1354 ; 27, 1592 ; 28, Ref. 80. 1:f7 Ibid. 17, Ref. 148 and 149. Also Ibid. 17, 2032, etc. 1:M Annalen. 279, 188 ; 293, I. 13 IO3» IO5> IIO> *33> 192, I95> 201, 211, 299, 336 Awdejeff, 312 BAEYER, 276, 279, 280, 284, 286, 289, 290, 293, 296, 35i, 352, 355 Bamberger, 282 Bayen, 13 Beaume, 36 Becher, 5, 7 Becquerel, 319 Beilstein, 271, 273, 274 Bergman, 10, 32, 36, 37, 40, 50, 86 Bernheimer, 331 Bertagnini, 292 Berthelot, 116, 242-245, 249, 282, 284, 288, 290, 293, 307, 323-325. 338, 349 Berthier, 319 Berthollet, 32, 35, 37-46, 48, 76-78, 86, 88, 89, 96, in, 124, 153, 314 Berthollet (younger), 78 — i | 294, i 49, 262, 315, 53, Berzelius, 34, 45, 50, 66, 69, 71, 74, 76, 81, 84-95, 97-'02, 104-106, 108-110, 112-115, 118-121, 123, 126, 131-133, 135, 136, 138, 142, 149, 158, 162, 164, 168-172, 174- 176, 178, 182, 188-190, 207, 223- 225, 227-230, 298, 325, 339 Bessemer, 343 Beudant, 96, 97 Bineau, 148, 192, 300 Biot, 332 Black, ii Blagden, 22 Boerhave, 8, 20 Boullay, 120, 122, 125, 133, 138 Boyle, 9, 27, 59, 335, 336 Braun, 325 Brauner, 313 Brewster, 317 Brodie, 198-200 Briihl, 329-331 Briining, 258, 260 Brugnatelli, 70 Buchner, 349 Buckle, i Buckton, 228, 239, 248 Buff, H. L., 247-249 Bunsen, 126, 127, 305, 313, 316, 317, 326-328, 341, 342 Butlerow, 240, 256, 264 /^ADET, 126, 127 V Cagniard de la Tour, 320 Cahours, 212, 227, 285, 301 Cailletet, 322, 343, 344 Cannizzaro, 298-300, 304, 337 358 INDEX OF AUTHORS' NAMES. Carius, 268, 269 Carlisle, 69 Caro, 279, 282, 287 Cauchy, 329 Cavendish, n, 12, 19, 22, 25, 28, 76 Chancel, 186, 210, 262 Chappuis, 345 Chevreul, 115 Chiozza, 212, 213, 215, 293 Ciamician, 319 Claisen, 294, 354 Clark, 118, 152, 158 Classen, 341 Clausius, 201, 301 Clement, 81 Cleve, 311 Cloez, 247 Coppet, 331 Couper, 253-255, 260, 269, 308 Courtois, 8 1 Crafts, 167, 210, 228, 291, 300, 343 Crookes, 318 Cruickshank, 70, 341 Curtius, 348 DALE, 330 Dalton, 45, 48, 49, 53-61, 63-66, 91, 93, 101, 102, 112, 175, 176 Danicll, 325 D'Arcet, 77 Darwin, 2 Davy, 70-86, 101. 119, 131, 138, 158, 161, 325, 339, 341 Debray, 303 De la Rive, 341 De Luc, 22 Democritus, 47 Demole, 267 Dessaignes, 268 Deville, 299, 301-307, 320, 349 Dewar, 283, 344 Dobereiner, 114, 122, 311 Draper, 326, 327 Drion, 320 Dulong, 83, 84, 95, 96, 99, 104, 106, 119, 158, 160, 162, 298 Dumas, 101-105, IIO> I22'I25> I27> 128, 132-138, 140-146, 148-151, 157, 162-168, 170, 176, 178-180, 182, 189, 203, 206, 211, 216, 217, 225, 226, 229, 232, 240, 299, 311, 327 Duppa, 257, 264, 289 Dusart, 291 T7BELMEN, 319 i\ Emmerling, 286 Empedocles, 4, 5 Engelhard, 258 Erlenmeyer, 256, 263, 264, 280, 289, 290 Erman, 89 Esson, 315 Exner, 331 FARADAY, 106, 117, 118, 141, 199, 281, 321, 322, 325, 339 Favre, 197, 323 Fes ting, 319 Fischer, E., 280, 348, 350, 354 Fischer, E. G., 52, 53 Fischer, O., 280 Fittig, 270, 271, 276, 280-282, 290, 295 Foucault, 317 Fourcroy, 32, 58, 89 Franchimont, 281 Frankland, 203, 204, 225-227, 230- 235, 242, 249, 262, 264, 290, 291, 308 Fraunhofer, 316-318, 327 Freund, 263 Friedel, 167, 210, 228, 261-263, 289, 291, 320, 355 Fritzsche, 287 Fuchs, 96 , 69 VJT Gaudin, 319 Gautier, 270 Gav-Lussac, 45, 58-62, 74, 76-79, 81, 82, 84, 91-93, 96, 101, 104, 105, 107, 113, 114, 119, I2K 122, 124, 125, 140, 196, 319, r I35' 3/ Lreber, o Gehlen, 96 Geoffrey, 36, 86 Gerhardt, no, ill, 143, 149, 161, 167, 171, 175, 180-194, 196, 197, 202-205, 2io-22i, 223, 225, 229, 230, 233, 235, 240-242, 249, 252, 253, 256, 285, 293, 298 INDEX OF AUTHORS' NAMES. 359 Gerichten, 286 Gerland, 257 Geuther, 260 Gibbs, 304, 333, 334, 341 Gladstone, 301, 330 Glan, 319 Glaser, 281 Glauber, 156 Gmelin, 106, 107, 144, 146, 150, 174, 175, 187-189, 228, 235, 311 Goldschmidt, 343 Goldschmiedt, 282 Grabe, 276, 278-281, 285, 287 Graham, 152-155, 157, 207, 208, 331 Greiff, 296 Griess, 275, 277, 280 Griffin, 158 Grimaux, 290 Groth, 332 Grove, 301 Guldberg, 314, 315 Guyton de Morveau, 8, 9, 32, 89 HAAGEN, 330 Hall, Sir James, 319 Hampson, 344 Hantzsch, 352 Harcourt, Vernon, 315 Hausmann, 319 Hautefeuille, 320, 345 Hauy, 96 Hellriegel, 349 Helmholtz, 27, 301, 304, 325, 326, 340 Hennell, 132, 133, 206 Henry, 6 1, 79, 335, 336 Heroult, 342 Herschel, Sir [., 317 Hess, 323, 338 Heumann, 355 Higgins, 54 Hillebrandt, 346 Hintze, 332 Hisinger, 69, 71 Hittorf, 318, 325, 340 Hobrecker, 295 Hoff, van 't, 268, 269, 333, 335, 337. 34°, 349-351 Hofmann, 185, 203, 204, 212, 233, 239, 247, 248, 266, 270, 279, 280, 285, 286, 289, 292, 310 Hoogewerff, 285 Hooke, 7, 21 Hooker, 282 Horbaczewsky, 290 Horsford, 246 Horstmann, 304, 316, 333 Huber, 285 Hiibner, 277, 295 Hufner, 289 Humboldt, 58 Hunt, 291 T NGENHOUSZ, 327 Isambert, 303 JACOB!, 341 Jahn, 326 Jelett, 316 Joule, 301 KANE, 209, 227, 293 Kanonnikoff, 331 Kant, 47 Kay, 243 Kekule, 221-223, 233, 240, 241, 249-253, 255, 256, 258, 260, 261, 269-276, 278-281, 289, 291, 294,. 304, 308, 337, 351, 354 Kiliani, 341 Kirchhoff, 316-318 Kirwan, 12, 13, 19, 28 Klaproth, 44, 112 Knorr, 354, 355 Koch, 319 Konigs, 284-286 Korner, 277, 283 Kohlrausch, 325, 340 Kolbe, 116, 136, 172, 182, 203, 220, 223-231, 233-239, 249, 255, 257-259, 261-264, 269, 280, 289, 291, 292 Kopp, 4, 96, 216, 252, 253, 304, 328, 329, 337 Krtiger, 355 Kriiss, 319 Kundt, 300, 345 LAAR, 354 Ladenburg, 277, 286, 290, 295> 345, 353 3<5° INDEX OF AUTHORS' NAMES. Landolt, 329, 330, 332, 339 Laplace, 22, 26, 28, 29, 322, 324, 330 Laurent, in, 128, 136, 138, 142- 150, 161-164, 167, 169, 176, 179, 180, 185, 187, 191-197, 202-205, 209, 213, 217, 232, 233, 235, 240, 266 Lautemann, 258 Lavoisier, 4, 11, 13, 16, 17, 20-29, 31-36, 49, 50, 53^ 58, 71, 77, 80- 83, 89, 101, 108, 111-113, IJ9> 124, 126, 138, 322, 324 Le Bel, 268, 350, 353 Leblanc, 229 Lecoq de Boisbaudran, 313, 319 Lemery, 7, 12, ill Lenssen, 311 Leucippus, 47 Lewkowitsch, 268 Lieben, 291, 294 Liebermann, 279, 281 Liebig, 109, no, 112, 114, 117, 125, 126, 128, 131-138, 140, 142, 149, 154-163, 170, 175, 178-180, 1 86, 188, 190, 202, 205-207, 211, 224, 226-229, 235, 240, 292 Limpricht, 221, 287, 289 Linde, 344 Lippmann, 340 Lockyer, 346 Lowig, 227 Lommel, 327 Lorentz, H. A., 330 Lorenz, L., 330 Lossen, 348 ' Louren9o, 247 Luckow, 341 Lucretius, 47 ]\ /T ACQUER, 9, 10, 13 1V1 Madrell/ 258 Magnus, 131, ^23, 341 Malaguti, 142^ 162, 171, 233 Marcet, 224 Marignac, 250, 306, 310 Mariotte, 59r335> 33^ Matthiessen/34i Maxwell, 301, 304, 332 Mayer, J. R., 301 Mayow, 7, 21 Melsens, 171, 224 Mendelejeff, 103, 311-313, 320 Mendius, 221, 291 Menschutkin, 315 Meusnier, 58 Meyer, Lothar, 103, 311, 332 Meyer, O. E., 331 Meyer, V-, 287, 300, 343, 348, 352, 353, 355 Michael, 351 Miller, W. A., 325 Miller, W. v., 341 Millon, 199 Mitscherlich, A., 318 Mitscherlich, E., 95-98, 106, 139, 158, 181, 182, 207, 224, 235, 275, 299, 301, 319 Moissan, 342, 348 Mond, 348 Morkownikoff, 264 Mousson, 316 Miiller, H., 292 AJASINI, 331 I > Natanson, 279 Natterer, 322 Naumann, 303, 304 Nernst, 340 Neumann, 96 Newlands, 103, 311, 312 Newton, 8 Nicholson, 69 Nilson, 313 ODLING, 223, 228, 233, 242, 243, 245, 249 Olszewsky, 343 Ostermeyer, 281 Ostwald, 315, 333, 338 Oudemans, 339 PASTEUR, 267, 268, 350 Pean de Saint Gilles, 315 Pebal, 305 Peligot, 135, 142 Pelouze, 291, 292 Perkin, 257, 261, 279, 289, 292, 294, 332 Persoz, J. , 280 Petermann, 277 Petit, 95, 96, 99, 104, 106, 298 INDEX OF AUTHORS' NAMES. 2 25> Pringsheim, 327 Proust, 42-46, 48, 112 Prout, 102, 311 RAMSAY, 284, 345-347 Rankine, 301 Raoult, 331, 336, 337 Raschig, 349 Rayleigh, Lord, 345 Regnault, 96. 136-138, 142, 143, 162, 165, 216, 228, 298, 301, 302 Reich, 318 Reichenbach, 271 Reimer, 290, 292 Rey, Jean, 7, 21 Richter, J. B., 45, 48, 50-52, 64, 153 Richter, Th., 318 Ritter, 73, 339, 340 Robinson, 305 Rochleder, 269, 285 Roozeboom, 335 Roscoe, 46, 310, 326-328 Rose, H., 137, 170, 298 Rosenstiehl, 279 Rossi, 291 SAUSSURE, 113 Scheele, 10, u, 13, 16, 19, 20, 24, 78, 80, ill, 115, 326 Schiel, 216 Schiff, H., 311 Schiff, R., 278 Schmitt, 280 Schonbein, 200 Schorlemmer, 265 Schweizer, 227 Seebeck, 76 Seguin, 58 Senarmont, 320 Serullas, 224 Silbermann, 197, 323 Silva, 290 Simon, 70 Simpson, 292 Skraup, 284, 285 Smith, Angus, 48, 50 Smith, E. F., 341 Soret, 200, 345 Spencer, 341 Stadeler, 263 Stahl, 5-8, II, 14, 20 Stas, 103 Stewart, Balfour, 317 Strecker, 246, 257, 289 Swan, 317 TAIT, 200 Talbot, 317 Than, 306 Thenard, 74, 76-79, 81, 113, 114, 119, 121, 127 Thenard, Paul, 212 Thiele, 349 Thilorier, 322 Thomsen, 278, 315, 323-325, 331, 338 Thomson, James, 301 Thomson, Thomas, 48, 53, 64, 92, 102 Thorpe, 308 Tiemann, 290, 355 Tillet, 13 Tollens, 271, 290 Troost, 299, 307, 345 Tschermak, 310 Tyndall, 316 ULRICH, 258 Uslar, 221 VALENTINE, Basil, 123 Van Dorp, 285 Van Helmont, n Vauquelin, 44, 58, 112 Verguin, 279 Vierordt, 318 Volhard, 289 Volta, 69, 72 Vortmann, 341 362 INDEX OF AUTHORS' NAMES. WAAGE, 314, 315 Waals, van der, 333-336 Wachendorff, 296 Walden, 351 Wallach, 355 Walter, 166 Wanklyn, 228, 235, 305 Warburg, 300 Watt, 28 Weidel, 284, 286 Weltzien, 311 Wenzel, 50 Werner, 352 Wertheim, 285, 289 Wiedemann, G., 316, 331 Willgerodt, 355 Williamson, ill, 196, 201, 202, 204-215, 221, 223, 226, 229, 233, 235. 237, 240-243, 249 "Windier, S. C. H.," 179 Winkler, 292, 313 Wischnegradzky, 264, 286 Wislicenus, 265, 292, 351 Wislicenus, W., 354 Wohler, 109, 116, 125, 126, 131, 140, 167, 170, 227, 288, 342 Wolf, 320 Wollaston, 48, 49, 64-66, 102, 106, 174, 187, 316 Woskresensky, 278 Wroblewsky, 322, 343 Wurtz, 199, 204, 210, 211, 214, 217, 223, 233, 242, 245-249, 257- 259, 263, 264, 269, 271, 290, 291, 294, 307, 3io yEISE, 133, 134 l^ Zimmennann, 313 Zincke, 290 Zinin, 289 INDEX OF SUBJECTS. A BSOLUTE boiling-point, 320 ' /\ Accumulators, 340 Acetamide, 228 Acetic acid, 120, 141, 164, 171, 208 | etc., 226 etc., 233, 235 etc., j 254 acid, formulae, 190, 191 ether, 208 Aceto- acetic ether, syntheses by means of, 264 etc., 291 Acetone, 209, 236, 238 Acetonic acid, 263, 264 Acetonitrile, 250 Acetyl, 137, 171 theory, 138 Acetylene, 269 Acichlorides, discovery, 212 Acid amides, 212 Acidifying principle, 25, 26, 82 Acids, dibasic, 183, 194 etc., 210, 212 etc., 229, 238 ideas of Davy and of Dulong, 83, 84, 161 Lavoisier's views concerning, 25> 31' 33. 78, 108 monobasic, 183, 195 poly basic, 185, 190, 194 etc., 202, 242, 292 polybasic, existence admitted by Kolbe, 238 tribasic, 239 views of Liebig, 157 etc. , 161 Acridine, 287 Acrylic aldehyde, 294 Active mass, 314 Additions, kinds distinguished by Gerhardt, 182 Affinity, 85 etc., 314, 315 Berthollet's views, 37 etc. degree, 254 Affinity, elective, 254 Guldberg and Waage's investi- gations, 314 etc. tables, 36, 41 Alanine, 257 synthesis, 256, 289 Alcohol, formulae, 005, 254. radicals, 203 synthesis, 116, 288 views of Berzelius concerning, 132 views of Dumas concerning, 141 views of Gay Lussac concerning, 122 views of Liebig concerning, 133 etc. "Alcoholic hydrogen," 261 Alcohols, 25$ conversion into acids (Kolbe's views), 236 new class predicted by Kolbe, 237, 262, 264 polyatomic, 245 etc. polyatomic, theory of, 242 tertiary, 264 Aldehyde, 228, 236 etc. Aldehydene, 136 Aldehydes, 295 Aldehydines, 295 Aldol, 294 Alizarine, 279, 281 Alkalies, decomposition, 67, 74 etc. relations of caustic and mild, 10, II Alkaloids, relations to pyridine, 285, 286 Alkyl bases, synthesis, 292, 293 oxybenzoic acids, 275 Allanto'in, synthesis, 290 364 INDEX OF SUBJECTS. Allotropy, 119 Alloxantine, synthesis, 290 Allyl alcohol, 253 Aluminium ethyl, 227 Amid-acids, 185, 195, 215, 221 Amide, 137 theory, 138 Amides, 185, 195 of dibasic acids, 213 Amido-acids, 257 Amines, 21 1 etc., 214, 217 Ammonia, discovery, n theory, 138 Ammonium, 130 amalgam, 76 carbamate, vapour density, 300 carbamate, dissociation, 307 carbonate, dissociation, 303 chloride, dissociation, 305, 306 j hydrosulphide, dissociation, 303» 307 theory, 131, 138 Amyl alcohol, 268 ether, 208 Amylene hydrate, 263, 264 Anhydrides, 195, 295 dibasic, 215 mixed, 210, 213 of dibasic acids, 213, 215 of monobasic acids, 212, 213, 295 Anilid-acids, 185, 195 Anilides, 185, 195 Aniline dyes, 279 Anisic acid, 229 Anthracene, 281, 295 synthesis, 289, 293 Anthraquinone, 281, 282 Antimony pentachloride, vapour density, 301 Antipyrine, 355 Aqiia regia, 1 1 Argon, 345, 346 Aromatic compounds, 270 compounds, hydrogenised, 351, ! 352 . compounds, isomensm in, 273 j etc. compounds, ortho-, meta-, and para-, 276 etc., 295 compounds, position of substi- tuted atoms or groups, 276 etc. Aromatic compounds, theory of, 271, 273 etc., 277, 279 hydrocarbons, oxidation into acids, 275 Arsenic, vapour density, 299, 343 Ascending series, methods of, 291 Aspartic acid, 268 Assimilation in plants, relation of light to, 326 etc. Asymmetric carbon atom theory, 268 etc., 351 etc. nitrogen, 352, 353 Atom, 49, 65, 101, 102, 104, 142, 150, 190, 192 etc., 230, 240, 297 and equivalent, 190 Atomic magnetism, 339 refraction, 331 theory, 47 etc., 53, 106, no, 175 etc., 201 weight, 174 weights, 54, 56 etc., 99, 100 etc., 175, 187 etc., 298 etc. weights, " new," 241 weights, views of Cannizzaro, 298 etc. Atomicities, metals possessing seve- ral, 242 Atomicity, 240, 254, 282, 307 and basicity, distinction, 162, 256, 259 Atoms, determination of number in compounds, 56, 64, 91 etc. migration of, 266 wandering of, 266 Avidity, 315 Avogadro's hypothesis, 61 etc., 103 etc., 1 10, 133, 192, 195, 201, 211, 299 Azo-compounds, 275 dyes, 280 dyes, substantive, 355 BARIUM sulphate, formula, 218 Barred formulae (Berzelius), 90 Baryta, discovery, 10 Bases, polyatomic, 242 Basicity and atomicity, distinction, 162 law of, 183, 184, 220 of acids, 157 Benzamide, 125 INDEX OF SUBJECTS. 365 Benzene, 271, 272 etc., 283, 285, 287 Kekule's formula, 272, 278 ortho-dicarbonic acid (Phthalic acid), 285 prism formula, 278 sulphonic acid, 222, 223 synthesis, 284, 288, 293 Benzoic acid, 125, 142, 147, 170, 257, 271, 273 ether, 125, 134 BenzoyI, 125, 126, 170 acetic ether, 291 chloride, 125, 170 compounds, 125 sulphuric acid, 182 Benzyl alcohol, 274 chloride, 274 Beryllium, atomic weight, 312 Bismuth, nitrate, 242 oxide, 242 Bitter almond oil, 140, 147 almond oil, Liebig and Wohler on, 109, 125 Boiling point, elevation, 336 points of liquids, 328 Butyric acid, 263 ether, 264 Butyro-lactic ether, 258 CACODYL, 126, 127, 221, 228, 231, 232 oxide, 127 Cacodylic acid, 127 Cadmium, vapour density, 299 Caesium, discovery, 318 Calcium carbide, 342 carbonate, crystallised, artificial, 3'9 carbonate, dissociation, 303 Calx, 6 reduction by hydrogen, 12 Capillary electrometer, 340 Carbazol, 287 Carbon, in all organic compounds, | in quadrivalence, 250 etc., 272 Carbonic acid, 243, 261 anhydride, II, 25, 28 anhydride, dissociation, 302 anhydride, isotherms, 321 anhydride, liquefaction, 322 Carbonic ethers, 265 oxide, 269, 270 oxide, discovery, II oxide, dissociation, 302 Carborundum, 342 Carbylamines, 270 Catalytic action, 207 Chemical difference (Brodie), 198, 200 mass, 37 etc., 314 system of Berzelius, 85 etc. system of Kolbe, 230' etc., 235 etc. volume, 315 Chloral, 140, 141, 147 hydrate, dissociation, 307 Chlorides of dibasic acids, 215 Chlorine, chemical nature, 79 etc., 119 discovery, 10 Chlorlactic ether, 257 Chlormethyl-sulphonic acid, 224 Chloroform, 140, 147, 250 Chlorophyll, 327 Chlorosulphonic acid, 214 Chlorotoluenes, 2/4 Chlorpicrin, 250 Chlorpropionic ether, 258 Chlorpropionyl chloride, 258 Choline, synthesis, 290 Chromyl chloride, 170 Chrysene (Naphthalene - phenan- threne), 282 Cinnamic aldehyde, synthesis, 293 Citric acid, n, 154, 242 acid, synthesis, 290 Classification of the elements, 310 Collidine, synthesis, 289, 294 Combination, a consequence of de- composition, 198 fixed or varying proportions, 35, 40 etc. Combining weight, 103, 106, 174 etc. Combustion, views respecting, 5 etc., 31 Lavoisier's theory of, 25 Complete equilibrium, 334 Components, 334 Concentration cells, 340 Condensation, syntheses by, 293 etc. Condensations, internal, 294 etc. Coniine, splitting of synthetic, 350 synthesis, 290 366 INDEX OF SUBJECTS. Conjugated compounds, 226 radicals, 220 etc., 228, 230 Constitution of compounds, 225, 329 Continuity of liquid and gaseous states, 321 Copula, 171 etc., 184, 225 Copulre, theory of, 227, 233 Corresponding conditions, van der Waals's theories, 335 Coupled compounds, 171, 184, 185 compounds, basicity of, 183, 184 compounds, Gerhardt's first views, 180, 182 etc. compounds, Gerhardt's later views, 183 Coupling, Frankland's views, 130 etc. Creatine, synthesis, 289 Cresols, 274 Critical pressure, 321 temperature, 320, 321, 343 Crotonic acid, synthesis, 289 aldehyde, synthesis, 294 Crypton, 346, 347 Cumarine, synthesis, 292, 295 Cuprous chloride, formula, 343 Cyanic acid, 116, 117, 147, 154, 255 Cyanogen, 124 etc., 147 compounds, 124, 125 Cyanuric acid, 155, 259 TAAPHNETINE, 295 1 J Densities of gases and va- pours, 104 etc. Dephlogisticated air, 18 Desmotropy, 354 Diamines, 295, 296 Diamonds, artificial, 342 Diazo-compounds, 275 Dibasic lactates, 258, 260 lactic ether, 258 Dibromo-benzenes, 277 pyridine, 286 Dichlormethyl-sulphonic acid, 224 Diethyl (Butane), synthesis, 290 Dimethyl benzene, 271, 275 carbinol, 238, 262 (Ethane), synthesis, 290 identical with ethyl hydride, 265 Dimethylamine, 262 Dimorphism, 106, 117 Dioxy - anthraquinone (Alizarine), ' 279 Diphenyl, synthesis, 293 Diphenylene methane, 282 Dissociation, 301 analogous to evaporation, 349 Deville's experiments, 301 electrolytic, 337 etc. theory of, 301 etc., 304 Disulphetholic acid, 248 Disulphobenzolic acid, 239 Disulphometholic acid, 239 Dithionic acid, 243 Divisibility of elementary molecules, 198, 200 etc. Double atoms, 90, 175 salts, 350, 353 Duaiism, 85, 108, 109, 119, 169 etc., 173 Dualistic theory, 88 etc. Dyads, Laurent's, 192, 195 Dynamo-electrical machine, 342 EKA-ALUMINIUM (Scan- dium), 313 Eka-boron (Gallium), 313 Eka-silicon (Germanium), 313 Electric furnace, 342 Electro-chemical theory of Berzelius, 74, 85 etc., 108 etc., 121, 168 etc., 172, 173, 230 chemical theory of Davy, 72 etc. chemistry, 339 etc. Electrolysis, 325, 341 Electrolytic conductivity, 339 dissociation, 337 etc. production of metals, 341, 342 Electrotyping, 341 Element, definition, 27 Elements, classification according to valencies, 310 of Becher, 5 of Empedocles, 4, 5 Enantiomorph forms, 267 Endothermic and exothermic re- actions, 324, 325 Enol formula, 354 Equivalence, 50 etc., 139 Equivalent, 49, 64, 101 etc., 139, 142, 150, 151, 174 etc., 187, 1 88, 190 etc., 196, 231, 240, 297» 3°7 INDEX OF SUBJECTS. 36? Equivalent weight, 307 Equivalents, 53, 65, 66, 107, 151, 193 Gmelin's, 175, 188 Gerhardt's, 241 Ester formation, limit, 315 formation rate, 315 Ethenyl-toluylene-diamine, 295 xylene-diamine, 295 Ether, 204 etc., 210, 233 formula?, 205, 208, 254 views of Berzelius concerning, | 132 views of Dumas concerning, 134 views of Gay-Lussac concerning, 122 views of Liebig concerning, 133 etc., 179 Williamson's experiments, 204 Etherirication, theories of, 206 etc. Etherin, 124 etc., 135, 137 theory, 122 etc., 137, 206 theory attacked by Liebig, 132 Ethers, mixed, 206, 208, 210 Ethionic acid, 131 Ethyl, 133, 137, 209, 226, 227, 229 alcohol, 237 amyl ether, 208 benzene, 271 carbinol, 262 methyl carbonate, 210 methyl oxalate, 210 pyridines, 284 sulphonic acid, 258 sulphuric acid, 184, 206 etc., 223, 224 theory, 138, 226 Ethylamine, 217, 247, 262 a substituted ammonia, 204 Ethylene, 247, 248, 269, 295 oxide, 246, 247, 295 sulphocyanide, 248 sulphurous acid, 248 synthesis, 288 TTARADAY'S law, 106, 325 Ji Felspars, artificial, 320 Tschermak's research on, 310 Fermentation without living organ- isms, 349 Ferric oxide, 242 sulphate, 194 iy 194 Ferrosuni) 194 Ferrous oxide, 242 sulphate, 193, 194 Fire, an element, 4, 8 Fixed air, 1 1, 24 Fluoranthene (Idryl), 282 Fluorene (Diphenylene methane), 282 Fluorine, isolation, 348 Formic acid, 120, 235, 236, 292 acid, synthesis, 116, 288 ether, tribasic (Kay), 243 Formulae, barred, 90 constitutional, 253, 255 empirical, 186 equivalent, 193 etc. four-volume, 189 graphic, 260, 265 molecular, 194 structural, 265 synoptic, 186 two-volume, 192 Fuchsine, 279 Fraunhofer lines, 316 etc. Freezing points, depression, 331, 336, 337 Fulminic acid, 117, 154 Fumaric acid, 351 Furfuran, 287 , n OT Gallium, 313, 31 8 Galvanic current, decompositions by aid of, 68 etc. Gaseous volumes, law of, 58 etc., 91 etc., 108, 196 Gases, liquefaction of, 321, 322, 343 etc. transpiration of, 331 Germanium, 313 Glucoses, 268 Glyceric acid, 259 Glycerine, Berthelot's researches, 242 etc. discovery, 1 1 5 synthesis, 289 Glycocoll, 257 synthesis, 289 Glycol, 246 etc., 255, 257, 259 Glycollic acid, 246, 257 etc. Gycols, 242, 248 Guanidine, synthesis, 289 368 INDEX OF SUBJECTS. HEAT, Lavoisier's views con- cerning, 26 etc. Heats of formation, 324 of neutralisation, 338, 339 Heliotropine, 355 Helium, 344, 346 Heterologous series, 217 Homologous compounds, 216 compounds, regular differences in boiling-points, 216 series, 216 Homologues of marsh gas, 203 Hydracids, 82 Hydrazine, 348, 349 Hydrazoic acid, 348 Hydrobenzamide, 147, 148 Hydrochloric acid, discovery, 11 acid, dissociation, 302 Hydrocyanic acid, n, 78, 124, 147, 255, 292 Hydrofluoric acid, discovery, 1 1 Hydrogen, n, 12 acids (or hydracids), 82, 119, 158, 161 identity with phlogiston, 12 presence in metals suspected, 76 etc. Hydropthalic acids, 352 Hydroquinone, 278 Hydroxylamine, 348, 349 Hypochlorous anhyride, 195 Hypothesis of Avogadro, 61 etc., 103, etc., no, 133, 192, 195, 2OI, 211, 299 of Prout, 102, 311 IMIDES, 195 Increase of weight during com- bustion, 7, 21, 24 Indigo, 290, 296, 355 Indium, atomic weight, 313 discovery, 318 Indol, 286, 287, 295 Inflammable air, 12 Internal condensation, 294 etc. oxidation, 296 Iodine, discovery, 81 vapour density, 300, 343 lodo-, iodoso-, and iodonium com- pounds, 355 lonisation, 337 etc. lonone, 355 Isatine, 295, 296 Isethionic acid, 131, 222, 224, 258 Isobutyric acid, 263 ether, 264 Isologous series, 216 Isomers in benzene series, 273 etc. in benzene series, Korner's rule, 277 in pyridine series, 284 Isomeric amyl alcohols, 263 propyl alcohols, 262, 263 triphenyl-phosphine oxides, 309 valerianic acids, 263 Isomerism, 117 etc., 262, 265, 288 early observations, 117 in the benzene series, 273 etc. physical, 268 Isomorphism, 95 etc., 106, 140 Isonitriles, 270 Isophthalic acid, 276 J AHRESBERICHTE, 102 ^ ETONE formula, 354 Ketones, 236, 295 mixed, 209 KirchhorFs law, 317 Kolbe's reaction, 292 ACT AM ETHAN, 258 Lactic acid, 11, 246, 257 etc., 351 acid, discussion regarding its constitution, 256 etc. acid, formulse, 256, 257 acid, views as to its basicity, 256,257 Lactone acids, 295 Lactones, 295 Lactyl chloride, 258 Law of chemical mass action, 315, 334 of Dulong and Petit, 95 etc., 1 06, 298 of Oudemans and Landolt, 339 of selective absorption, 317 of the even number of atoms, 192, 202 of thermo-neutrality, 338 Leucic acid, 262 INDEX OF SUBJECTS. 369 Leucine, synthesis, 289 Light, chemical effects of, 326 etc. Linking of carbon atoms (Kekule), 250 etc., 272 MAGNETIC rotatory power, 339 Malaguti's ethers, 171, 233 Maleic acid, 351 Malic acid, 11, 154 acid, synthesis, 289 Malonic ether, 291 Mandelic acid, 268, 292 Manganese, 170 discovery, 10 peroxide, 170 Manganic acid, 170 Marsh gas, 250 gas, synthesis, 288 gas, type, 250 Matter, indestructibility of, 15, 22 Mauvei'ne, 279 Mercaptans, discovery, 134 Mercury ethyl, 227 fulminate, 250 vapour density, 299 Mesitylene, 276 synthesis, 293 Metal-ammonia and metal-ammo- nium compounds, 310 Metalepsy, 142 Metals, existence as such in salts, 159 polybasicity, 242 Metameric substances, 118, 119 Metaphosphoric acid, 154, 158, 243 Metargon, 347 Methyl (Ethane), 203, 226, 227 alcohol, 237 aldehyde, 236 benzene, 271, 273 benzenes, 283 carbinol, 262 chloride, 250 cyanide, 226 ethyl ether, 205, 206, 210 propyl-phenanthrene, 282 pyridines, 283, 284 sulphonic acid, 224, 239 toluene, 271 Methylamine, 262 Methylene, 135, 147 Minerals, syntheses, 319, 320 Mixed anhydrides, 210, 213 ethers, 206, 208 ketones, 209 types, 221 etc. Mixtures, 35, 45, 107 Molecular compounds, 308, 309 magnetism, 331 physics, 328 etc. refraction, 330, 331 volumes, 328, 329, 331 weights, 187 etc., 196, 209, 299, 337 Molecule, 192 etc., 230, 240, 297, 298 chemical, 201, 203, 210 etc. physical, 21 1 Molecules, 53, 61 etc., 195, 196 Monochloracetic acid, 229 Morphotropy, 332 Multiple point, 335 proportions, law of, 48, 53 etc., 91, 108, 115, 175 Muriatic urn, 80, 81, Mustard oil, synthesis, 289 NAPHTHALENE, 276, 280, 281, 283, 285, 287, 295 isomeric derivatives, 281 mono-sulphonic acids, 281 phenanthrene, 282 a-Naphthol, 280 Nascent state, 196 Neon, 347 Neurine, 247, 290 Neutrality, law of, 50 Nickel carbonyl, discovery, 348 Nitric acid, 195, 236, 242, 243 acid, composition, 25 oxide, discovery, 1 1 Nitriles, 195, 270 conversion into acids, 226, 292 Nitro-aerial spirit, 21 Nitrobenzene, 181 Nitrogen, assimilation, 349 equivalent of, 148 sulphonic acids, 349 Nomenclature of Berzelius, 89 new system of chemical, 32, etc. 109 2 A 37° INDEX OF SUBJECTS. Notation of Berzelius, 90 Nucleus theory of Laurent, 143 etc. , 167, 174, 1 80, 202 OIL of the Dutch chemists, 136, 141, 149 Olefiant gas, 122 etc. Organic acids, Liebig's researches, 154 analysis, 28 etc., 113 etc. analysis of nitrogenous sub- stances, 114 chemistry, new nomenclature, 355 chemistry, separate treatment, 108, in etc. compounds, classification by Gerhardt, 215 etc. compounds, constitution of, 117, 121, 288 compounds, derivatives of in- organic compounds, 235 Organo-metallic compounds, 227, 232 Ortho-condensation, 296 (a /?) pyridine dicarbonic acid, 285 Orthotoluidine, 279 Orthrin, 126 Osmotic pressure, 336 Othyl, 209 Oxalic acid, u, 122, 226, 243, 255, 259 Oxamide, 185, 214 Oxanilide, 185 Oxidation of substituted pyridines, 284 Oximes, 348 Oxindol, synthesis, 296 Oxybenzoic acid, 257 Oxycumarines, 295 Oxygen, assumed presence in hydro- chloric acid, 76, 78 etc. discovery, n, 16, 24 Oxy-isobutyric acid, 264 Oxypropionic acid, 257 Ozone, density, 345 discovery, 200 PARA - OXYBENZOIC acid, 277 Perfumes, artificial, 355 Periodic law, 103, 311 etc. Perkin's reaction, 292 Phase rule, 334 etc., 349 Phases, theory of, 334 etc. Phenanthrene, 281 Phenol aldehydes, synthesis, 292 dyes, 280 Phenols, synthesis, 291 Phenyl, 229 hydrazine, 348 sulphonic acid, 239 Phenylene diamines, 277 Phlogistians, chemical knowledge of, 10 Phlogisticated air, 18 Phlogiston theory, 5 etc., 12 etc. Phosgene, 170 Phosphoric acid, 236, 244 acids, Graham's investigations, 152 etc. acids, supposed isomerism, 118, 152 anhydride, 25 Phosphorus, acids of, 243 pentabromide, vapour density, 301 pentachloride, vapour density, 301 pentachloride, dissociation, 305, 307 vapour density, 106, 299 Photo-chemical induction, 328 Phthale'ines, 280, 295 Phthalic acid, 238, 276 anhydride, 282 »Picene, 283 Picoline, synthesis, 289, 294 Picric acid, 221 Piperidine, 285 synthesis, 290 Piperonal, 355 Polybasic acids, 185, 190, 194 etc., 202, 242, 292 acids, theory of, 151, 154 etc. Polyethylene alcohols, 247 Polarity (Brodie), 198 etc. Polymerism, 118 Potassamide, 76, 77 Potassium, discovery, 67, 74 etc. ethyl, 228 iodide, formula, 343 INDEX OF SUBJECTS. Principle of Hess, 323 of maximum work, 324 Probability theory (Maxwell's), 304 Proin, 126 Propionic acid, 257, 258 acid, synthesis, 235 ether, 258 Propyl aldehyde, 253 pyridines, 284, 286 Propylene, 269 glycol, 257 Prout's hypothesis, 102, 311 Purpurine, 281 Pyrazol group, 355 Pyrene, 283 Pyridine, 283 etc. carbonic acid from oxidation of nicotine, 285 dicarbonic acid, 286 series, isomerism in, 284 series, position of substituted atoms or groups, 284 synthesis, 284 tricar bonic acid, 286 Pyrophosphoric acid, 153, 158, 243 Pyrrol, 286, 287 QUINIZARINE, 281 Quinoline, 283 etc., 295 from decomposition of al- kaloids, 285 synthesis, 284, 285, 296 Quinone, 278, 279 formulae for, 278 Quinones, Grabe's examination of, 278 etc. RACEMIC acid, 118, 267 etc. acid, Pasteur's modes of de- composing, 267 etc. acid, synthesis, 289 and partially racemic substances, 353 compounds, splitting of, 350 Racemism, 353 Radical, 31, 33, 124, 126 etc., 167, 169, 217, 225 compound, 109 Liebig's definition, 128 theory, no, 121, 130 etc., 174, 176 etc., 230 Radicals, 186, 217 binary, 170 conjugated, 220 etc., 228, 230 containing metals, 221, 231, 235, 242 different, recognised by Kolbe, 229 isolation, 226 polyatomic, 240, 247 with basicity greater than one, 214, 221 Refraction equivalents, 330 Refractive index, 329 etc. Relations between electrical and chemical forces, 325 between optical and chemical properties, 326 by weight in chemical changes, II, 21 Replacement, 139 Residues, theory of, 180 etc., 293 Respiration, 29 etc. Retene (Methyl - propyl - phenan - threne), 282 Rosaniline, 279, 295 Rosolic acid, 280, 295 Rotation of plane of polarisation, 332 of plane of polarisation, electro- magnetic, 332 Rubidium, discovery, 318 Rubies, artificial, 319 SALICYLIC acid, 229 Salts, 242 amphid, 84, 120 haloid, 84, 120 neutral, 161 views regarding, 120 Salylic acid, 271, 273 Saturating capacity (Frankland), 231, 242 Scandium, 313, 318 Secondary batteries, 340 Semi-permeable membranes, 336 Sesquioxides, 242 Silicic acid, derivatives, 310 Silicon ethyl, 228 fluoride, discovery, n replacement of carbon by, 167 Sodium, discovery, 67, 74 etc. ethyl, 228 372 INDEX OF SUBJECTS. Solution, van 't Hoff's theory, 335, 336 tension, 340 Specific volumes of liquids, 252, 328 Spectra, absorption, 319 band, 318 emission, 319 line, 318 Spectrum analysis, 316 etc. analysis, quantitative, 318 bright bands, 317 Statiqne Chimique, Berthollet's, 35 Stellar chemistry, 318 Stereo-chemistry, 268, 350, 353 Stibethin, 232 Stochiometry, 52, 113 Suboxides, 89 Substituted ammonias, 204, 214, 217 Substitution, 139 etc., 150, 162 etc., 165 etc., 202, 225, 266 early observations, 140 of carbon by silicon, 167 Succinic acid, 238 Sugar group, 348, 350, 351, 354 Sulphacetic acid, 229, 239 Sulphanilic acid, 185 Sulphite of perchloride of carbon (Kolbe), 224 Sulpho-acids, 239 Sulphobenzide, 181, 182 Sulphobenzoic acid, 182, 221, 222, 239 Sulphocamphoric acid, 166 Sulphovinic acid, 182, 206 (see also Ethyl sulphuric acid) Sulphur, vapour density, 105, 201, 299 Sulphuric acid, 120, 195, 236, 239, 243' 247 acid, a dibasic acid, 183, 184, 185, 187, 214 acid, dissociation, 305 acid, Nordhausen, 222 anhydride, 25, 247 Sulphurous acid, discovery, n anhydride, dissociation, 302 Sulphuryl chloride, 214 Superoxides, 90 Synthesis by condensation, 294 of aromatic hydrocarbons, 290, 291 of hydrocarbons, 290 of minerals, 319 Synthesis of organic compounds. 1 1 6, 288 etc. emetic, 159 ±_ Tartaric acid, 154, 156, 159, 190 acid, formula, 202 acid, inactive, 267 etc. acid, isornerism, 118, 267 acid, left rotating, 267 etc. acid, right rotating, 267 etc. Taurine, synthesis, 289 Tautomerism, 354 Tellurium, atomic weight, 313, 346 ethyl, 227 Temperatures, high, attainment of, 342 etc. low, attainment of, 343 etc. Terephthalic acid, 275, 277 Terme de comparaison, Laurent's, 191, 209 Terpenes, 268, 354, 355 Tetra-phenol (F\irfuran), 287 Thallium, discovery, 318 Thermal effect, 325 Thermo-chemistry, 322 etc. researches of Thomsen, 315, 323 Thioacetic acid, 241 Thiofurfuran, 287 Thiophen, 287 Toluene, 271, 273 Toluylic acid, 275 Transition temperature, 335, 349, 35