Avogadro,the Chemists,and Historians of Chemistry: Part 2

The basis of Prout's ideas is discussed in Knight, Transcendental chemistry (ref. 14), 108–10.

Mauskopf , op. cit., Isis(ref. 14).

Fisher , op. cit. (ref. 85, part 1).

Levere T. H. , Affinity and matter: Elements of chemical philosophy, 1800–1865 (Oxford, 1971), 113–22.

Mauskopf , op. cit., Isis(ref. 14); Cole , op. cit. (ref. 14).

Baudrimont , op. cit. (ref. 49), i, 283–5.

Laurent A. , Comptes rendus mensuels , i (1845), 141; Baudrimont A. , Comptes rendus , xx (1845), 960; Laurent , op. cit. (ref. 49). For a more detailed analysis of Laurent's views before he accepted the hypothesis, see Brooke , op. cit. (ref. 20), ref. 22.

Laurent , op. cit. (ref. 49), 297–8; Baudrimont , op. cit. (ref. 49), i, 198.

In the same year as Laurent, 1846, Favre and Silbermann also concluded from a study of heats of reactions that oxygen gas should be regarded as O₂; “Recherches sur les chaleurs produites pendant les combinaisons chimiques”, Comptes rendus, xxiii (1846), 199–206. This provided independent support for Avogadro's views, but was not much noticed at the time.

The recent discovery of organic substitution had begun to undermine this confidence, notably in Baudrimont, Laurent, and Gerhardt.

Deville H. St-C. , “Sur la dissociation ou la décomposition spontannée des corps sous l'influence de la chaleur”, Comptes rendus, xlv (1845), 875–61. The anomalies were due to the fact that some components such as ammonium chloride dissociate on heating: NH₄Cl⇄NH₃ + HCl, giving half the vapour density one would expect. This problem (for references, see Partington , History (ref. 5), 494–5) does not seem to have worried many chemists before Deville's work; after all, it is only when a rule becomes established that there can be anomalies, and there were no firm rules governing vapour densities until Avogadro's hypothesis was accepted. Deville was still a prominent opponent of atoms in the Paris atomic debates of the late 1870s.

Brodie B. C. , “On the conditions of certain elements at the moment of chemical change”, Philosophical transactions, cxl (1850), 759–804. Brodie as an opponent of atomism is the central figure in Brock (ed.), op. cit. (ref. 81).

Clausius R. , “Ueber die Art der Bewegung, welche wir Warme nennen”, Annalen der Physik, c (1857), 353–80; Maxwell J. C. , “On the dynamical theory of gases”, BAAS Report (transactions) , xxix(1859), 9; Philosophical magazine , xix(1860), 19–32; xx (1860), 21–37. Clausius deduced an approximation to Avogadro's hypothesis from his initial assumptions, and Maxwell the hypothesis proper. I am grateful to Jon Dorling for pointing out to me in response to an earlier version of this paper that none of the possible theories explaining gas pressure held by physicists before Clausius and Maxwell had pressure depending only on temperature and the number of molecules in a given volume, as Avogadro's hypothesis requires. Clearly Avogadro's hypothesis was incompatible with contemporary physics, as well as with chemistry.

See ref. 50 above. Cannizzaro perhaps became acquainted with Gerhardt's reforms through his first teacher of chemistry, Piria, a pupil of Dumas but later a supporter of Gerhardt. For an account of Cannizzaro's unsettled Wanderjahre for a long time he could not settle on any one subject to study, and his career both as a student and as a teacher kept being interrupted by political sabbaticals see Leicester H. M. , “Cannizzaro, Stanislao”, in Gillispie C. C. (ed.), op. cit. (ref. 16), iii (New York, 1971), 45–49.

For a description see Partington , History (ref. 5), 419–24.

Cf. ref. 49 above.

Gerhardt C. F. , Introduction à l'étude de la chimie par le système unitaire (Paris, 1848), 55–57, and passim. Gerhardt's well known structural agnosticism' (he maintained that the chemist could not argue from the phenomena revealed at the moment of sometimes violent change to the constitutions of bodies at rest) is discussed by Russell , op. cit. (ref. 2); by Brooke J. H. , “Laurent, Gerhardt, and the philosophy of chemistry”. Historical studies in the physical sciences , vi (1975), 405–30; and perhaps most emphatically, and in the context of its repercussions in subsequent chemical theory, by Rocke A. J. , “Kekulé, Butlerov, and the historiography of the theory of chemical structure”, The British journal for the history of science, xiv (1981), 27–57.

Gerhardt C. F. , “Recherches sur la classification chimique des substances organiques”. Revuè scientifique, xii (1843), 592–600, p. 593.

Idem, [Introduction to vol. v], Comptes rendus mensuels des travaux chimiques , v(1849). i–viii.

Kekulé A. , “Aequivalent und Aequvalenz”, in Neues Handwörterbuch der Chemie , ed. von Fehling H. (Braunschweig. 1874). 77–89. p. 87. For resistance to Gerhardt's reforms, cf. Brooke, op. cit. (ref. 20).

Williamson A. W. “Theory of etherification”. Philosophical magazine , xxxvii (1850), 350–6; reprinted in Alembic Club reprints, xvi (reissued Edinburgh, 1949), 7 17. For ether, cf. the text above at ref. 86. It is interesting, seeing how an interest in reaction mechanism was a prerequisite for an appreciation of Avogadro (cf. text at ref. 102), that Williamson went on to consider the dynamics of etherification; ibid. (ACR), 18–24. In the 1850s, however, Williamson was no Avogadrian. i.

Odling W. quoted by H. Hartley (his pupil), in Studies in the history of chemistry (Oxford. 1971). 83–84. 146. 205.

See Fisher , op. cit. (ref. 85, part 2).

Pebal , quoted by his friend Meēr L. , op. cit. (ref. 82). 56n.

Gerhardt C. F. Traité de chimie organique (Paris. 1854–56). pp. i–ii.

Kekulé A. , Lehrbuch der organischen Chemie. oder der Chemie der kohlenstoffverbindungen. i (Frlangen. 1861): The first signature (pp. 1 230) appeared in 1859.

Kopp H. , Jahresbericht über die Fortschritte der Chemie…für 1858 , xi (1859), 11–13, 14n. It should be noted that Kopp's interest was primarily in Cannizzaro's use of Dulong and Petit's law; the name of neither Avogadro nor Ampère was mentioned, although there was some discussion of the application of vapour densities to weight determinations.

Of the 126 chemists named in Anschütz, op. cit. (ref. 10) as attending the congress, more than half were of German birth.

Cannizzaro S. , “Application à la chimie de la théorie d'Ampère et d'Avogadro sur la constitution des corps gazeux”, Répertoire de chimie pure, i (1858–59), 201–5. For Wurtz's long-standing interest in the atomic theory, see Yoshida Akira , “Charles Adolphe Wurtz et la théorie atomique”, Japanese studies in the history of science, xvi (1977), 129–35. Yoshida ascribes the Cannizzaro abstract to one of Wurtz's editorial assistants. Félix Leblanc (ibid., 133) but he cites no evidence for this. Wurtz's other major preoccupation at this time was with the glycols and with lactic acid (see Fisher , op. cit. (ref. 91), 38–39). These topics, too, got a thorough airing in the Répertoire. It was very much Wurtz's journal.

Daubeny C. G. B. , BAAS report , xxvi (1856), xlviii–lxiii, pp. xlix–liv; Foster G. C. , BAAS report (transactions), xxvii (1857), 45–47.

Odling W. , ibid ., xxviii (1858), 58; Mercer J. , ibid., 57–58.

The journey to Aberdeen, long and uncomfortable as it must then have been, must have paled into insignificance for the select few invited out to Balmoral at the end of the meeting to catch a glimpse of the Queen. The horrors of this expedition are described by Carnie W. , Reporting reminiscences (3 vols, Aberdeen, 1902–6), i, 336–9, and in the St James's chronicle, 27 September 1859, 6. The party left Aberdeen by train at 6 a.m. for Banchory. Proceeding by stage coach and omnibus for the remaining thirty-three miles, they were drenched by “a perfect torrent” at Ballater, and they had further to endure a Highland Games at Balmoral before they were eventually fed in mid-afternoon. On the way back to Aberdeen, they found plenty to eat and drink at Banchory station, which, it was implied, was to be paid for by the Association. Later, however, “a man came round and demanded and received one shilling for every cup of tea or coffee consumed, and sixpence for one or two biscuits. On being remonstrated with for this excessive charge, and for the manner in which the ladies and gentlemen had been thrown off their guard, he coolly said, ‘We have not the British Association here every day’”. I am grateful to Mr C. A. McLaren, Aberdeen University Archivist and Keeper of Manuscripts, for this anecdote.

Playfair L. , BAAS report (transactions) , xxix (1859), 65–66. At Oxford a year later, Miller W. A. made the ‘inconvenience’ attendant on the revision a positive reason for rejection of Gerhardt's reforms: “Chemistry, it must be remembered, is not merely a science: It is also an art, which has introduced its nomenclature and its notation into our manufactories, and, in some measure, even into daily life; it is therefore specially necessary to beware of needless innovation”; ibid ., xxx (1860), 70–71, p. 70.

Foster G. C. and Odling W. , BAAS report, xxix (1895), 1–22.

ibid., 20–21.

ibid., 2. The chemical question was the formulation of the aldehydes, and the physical data Hermann Kopp's determination of the specific gravities of liquid organic compounds, which (correctly) suggested to Kopp that Gerhardt's formulae should be revised. For Gerhardt's problems with the aldehydes, see Fisher , op. cit. (ref. 85, part 2), Appendix, 231–3. As a final point about the Aberdeen meeting, it should be noted that it was there that Maxwell (professor of natural philosophy at Marischal College, Aberdeen) first announced his kinetic theory of gases, including his deduction of Avogadro's hypothesis; see ref. 111, above.

Odling W. , “Atomic weights”, in A dictionary of chemistry, ed. Watts H. (8 vols in 9, London, 1859–81), i, 452–73.

ibid., 456.

ibid., 455.

ibid., 457.

ibid., 465–6; for Liebig, see ref. 98, above.

It is not clear how Odling got to know of Cannizzaro's “Sketch”. It would appear from the British union catalogue of periodicals that only Cambridge University Library was taking. Il nuovo cimento in 1858. Presumably Cannizzaro, whose liberality with offprints is well known, had sent one to Odling who was known to be interested in atomic weights.

Odling , op. cit. (ref. 135), 466.

The exceptions arose from dissociation, as in the case of ammonium chloride (cf. ref. 109), and also from dimerization. Historians have noted the importance of the former as a deterrent to the acceptance of Avogadro's hypothesis, and have, like Odling, discussed its ‘explaining away’ by Deville and others. Few if any historians have noted the inverse problem: The vapour density of aluminium chloride gives a molecular formula of Al₂Cl₆, as compared to the chemically determined formula AlCl₃. Odling was more worried by this than by dissociation; it could not so easily be explained away. One has to agree with him that there were few obvious benefits, and distinct inconveniences, in doubling the accepted formulae; ibid., 468, 470.

ibid., 461.

Ibid ., 471. By the time he came to write “Metals, atomic weights and classification of”, in ibid., iii, 957–76, published in 1865, Odling had accepted Cannizzaro's reforms (again, primarily for chemical reasons), and he inserted a footnote to this effect in the 2nd edn of vol. i (London, 1879), against the above quotation: “This opinion has since been modified by the writer”; p. 471n. See also ref. 182, below.

Farrar W. V. , “Odling, William”, in Williams (ed.), op. cit. (ref. 3), 396.

Cf. ref. 84, above. There was a rationale, usually ignored by historians, for the well known antipathy of nineteenth century journal editors towards theoretical papers: Theory belonged in the classroom, not in the research report.

For Williamson as a teacher, see Tilden's comments quoted in Harris J. and Brock W. H. , “From Giessen to Gower Street: Towards a biography of Alexander William Williamson (1824–1904)”, Annals of science , xxxi (1974), 95–130, p. 110. For Wurtz, see Brooke J. H. , “Wurtz, Charles-Adolphe”, in Gillispie C. C. (ed.), op. cit. (ref. 16), xiv (New York, 1976), 529–32: Wurtz had a “real vocation as a teacher”, and “became one of the most enthusiastic and outstanding teachers of his generation” – the closest thing France had to a Liebig (pp. 529–30). For Kekulé, see von Meyer E. , A history of chemistry from the earliest times to the present day, tr. by McGowan G. , 2nd English edn from 2nd German edn (London, 1898), 308.

In ‘Cannizzaro's reaction’ certain aldehydes disproportionate in the presence of concentrated alkali to give the acid and the alcohol; This reaction was the subject of Cannizzaro's first published papers, “Ueber den der Benzoësäure entsprechenden Alkohol”, Annalen, lxxxviii (1853), 129–30; xc (1854), 252–4; xcii (1854), 113–17. In the second and third of these papers, Cannizzaro had repeatedly to correct earlier mistakes, and he never really found out what happened in the reaction, detecting only the reduction to the alcohol. In the second half of his research career, he busied himself with the natural product santonin, which gives plenty of derivatives with all sorts of reagents; Cannizzaro was content to obtain and analyze as many as he could, with no attempt at interpretation. This, was not a very productive research programme.

Leicester , op. cit. (ref. 112), 46–47.

For the lack of laboratory, see ibid., 46. Cannizzaro's skill as a teacher was recognized by Walker James , professor of chemistry at Edinburgh, in his preface to the English translation of the “Sketch”, ACR (ref. 10), iii: “The eminence of Cannizzaro as a teacher is plain in every page of the summary of his lecture course on chemical philosophy which is here translated. The facts are marshalled and their bearing explained with absolute mastery of pedagogic method, and one is impelled to the conclusion that Cannizzaro's students of 1858 must have had clearer conceptions of chemical theory than most of his scientific colleagues of a much later date”.

Cannizzaro S. , “Considerations on some points of the theoretic teaching of chemistry”, Journal of the Chemical Society, xxv (1872), 941–67; reprinted in Chemical Society, Faraday lectures (London, 1928), 17–43. The latter will be cited in subsequent references.

This point has not always been sufficiently appreciated by historians, though Ihde noted it briefly in his op. cit. (ref. 56), and in his Development (ref. 5), 227. Cf. also Nye , op. cit. (ref. 81), 260–2.

See Russell , op. cit. (ref. 2), 36–43. For Frankland and other adherents of the radical theory, these compounds had no implications for vapour-density considerations. Their theoretical importance lay in leading Frankland to the origins of the concept of valency.

Cannizzaro , ACR(ref. 10), 21–24, 31.

Cannizzaro's primary reliance on Dulong and Petit was noted by Odling in his op. cit. (ref. 135), 471. Tilden W. A. , in his Short history of the progress of scientific chemistry in our own times (London, 1899), mentioned Cannizzaro's name only in connexion with the systematic applications of Dulong and Petit's law to all the elements (p. 75); though in his “Cannizzaro memorial lecture” thirteen years later (op. cit. (ref. 12)), he gave a much fuller exposition of Cannizzaro's reforms, including his use of Avogadro's hypothesis. It was his reform of inorganic weights (through specific heats) that was his greatest achievement; it “finally disposed of the superstition which had hovered so long in the minds of chemists that organic chemistry was subject to laws different from those prevailing among mineral substances. There is, in fact, but one science of chemistry and one set of atomic weights” (p. 1691). Meldrum took much the same line: Gerhardt and Laurent had reformed and systematized organic chemistry where “strictly chemical” methods were applicable; Cannizzaro's reform of inorganic weights complemented this, and served to recombine the two branches; op. cit. (ref. 1), 90, 100.

Cannizzaro , ACR (ref. 10), 1.

Cannizzaro , op. cit. (ref. 151), 18.

Ibid. Interestingly, analogous arguments for avoiding the chaos of empiricism are to be found in a recent American introduction to chemistry for high school students, put together by a team called “Chemical Education Material Study”, who published Chemistry: An experimental science , ed. Pimentel G. C. (San Francisco, 1963); the accompanying Teachers' guide, ed. McLellan A. L. , synopsizes the discussion of “thirty-five scientists and educators” on how the atomic theory should be presented in ch. 2: The consensus was that a mature scientist believes in the usefulness of the atomic hypothesis because he himself has felt innumerable successes in a wide variety of applications based on this hypothesis…. For these reasons our task in Chapter 2 is not to present the evidence for atoms; the evidence is too diverse, too manifold, too esoteric [and too modern!] Rather, we must begin with a small part of the evidence, and, in an atmosphere of tentativity. let the particulate model grow as new facts are learned (p. 75). Just as had Cannizzaro, the team chose Avogadro's hypothesis as their small starting point. And they were quite clear that the historical approach of the Conant-style ‘case-study’ was counter productive: If it is intended only to clarify the logic by which chemical evidence supports the atomic theory, there is no obligation to display the tortuous process by which this logic was recognized…. Indeed, if one's intent is to ensure that the student understands the logic, it is undesirable to relate the logic to the chronology because it accents difficulty…. Freeing the development of the atomic hypothesis from chronology makes it possible to use the simplest and most easily grasped presentation of the logic by which the chemical evidence supports the atomic theory (p. 77). Faced with the same problem, Cannizzaro put forward the same solution: His contemporaries were trapped in the confusion of history — they were part of it, indeed. For teaching purposes, Cannizzaro turned from the confusion of induction to the clarity and distinctness of deduction from Avogadro's hypothesis.

Cannizzaro , op. cit. (ref. 151), 22.

ibid., 24.

Cf. ref. 1, above.

See for example de Milt C. , (i) “Carl Weltzien and the congress at Karlsruhe”, Chymia, i (1948), 153–69; idem, (ii) “The congress at Karlsruhe”, Journal of chemical education, xxviii (1951), 421–4. Ihde's “Karlsruhe” (ref. 5) is on the whole a very fair summary of nineteenth century chemistry, but he does exaggerate the importance of the congress not inappropriately, perhaps for a “centennial retrospect”.

“The congress of chemists at Karlsruhe”, Chemical news, ii (1960), 226–7. De Milt , op. cit. (ref. 162. ii), 422, ascribes this report to l'abbé Moigno, one of the foreign correspondents of the Chemical news.

Weltzien's speech is reproduced in translation by de Milt in op. cit. (ref. 162, i), 163–4.

von Meyer E. , “Die Karlsruhe Chemiker-Versammlung im Jahre 1860”, Journal für praktische Chemie, cxci (1911), 182–9; Hartley , op. cit. (ref. 120), ch. 8: “Stanislao Cannizzaro, FRS (1836–1910), and the first international chemical conference at Karlsruhe in 1860”. Kekulé's caution in coming down publicly on one side of a theoretical question or the other is discussed by Rocke , op. cit. (ref. 115).

These included Mendeleev; see his letter to his teacher Vokressensky reproduced in translation by de Milt in op. cit. (ref. 162, ii), 422. See also his recollection of twenty-nine years later, in the 1889 Faraday lecture, “The periodic law of the chemical elements”, Journal of the Chemical Society, xlv (1889), 634–56, p. 636: “I vividly remember the impression produced by [Cannizzaro's] speeches, which admitted of no compromise, and seemed to advocate truth itself, based on the conceptions of Avogadro, Gerhardt, and Regnault, which at the time were far from being generally recognized”.

Meyer L. , op. cit. (ref. 82), 59.

How many European chemists shared Meyer's ability to read Italian? Any further examination of the influence of Cannizzaro will have to consider the question of language. Presumably Cannizzaro addressed the Karlsruhe congress in French, since France was the only foreign country in which he had spent any amount of time. (He must have spoken Italian to his English wife, since he confessed himself “not much accustomed” to English in 1872; op. cit. (ref. 151), 17.) In Wurtz's report, all the speeches are recorded in French (Anschütz, op. cit. (ref. 10)), but his account of Cannizzaro's major speech is so detailed (and almost as long as all the other minutes of the three days put together) as to suggest that Cannizzaro wrote this speech in a foreign tongue out in full, and later gave it to Wurtz. Before 1872, this was the only exposition of his ideas other than in Italian, but it remained unprinted in the nineteenth century. His “Sketch” was reprinted in Italy several times (for details, see Meyer , op. cit. (ref. 82). 60–61). but was not translated into German until 1891, or into English until 1911. by which time it was already a ‘classic’. Significantly, I know of no translation into French. The question of Cannizzaro's immediate personal influence on others than Meyer is of limited importance, of course, since the infection of Meyer was enough to spread the germ. But the lack of any evidence of an epidemic of acceptance is indicative of a continued resistance of chemists towards deductive chemistry.

Benfey O. T. “Meyer. Julius Lothar”. in Gillispie C. C. (ed.). op. cit. (ref. 16). ix (New York. 1974). 347–53; which draws heavily on Bedson P. P. “Lothar Meyer memorial lecture”. Journal of the Chemical Society . lxix (1896). 1402–39: Which draws heavily on Seubert K. , “Lothar Meyer”. Berichte der deutschen chemischen Gesellschaft. xxviii (1895). 1109–46.

Meyer O. E. , Die kinetische Theorie der Gase (Breslau, 1877: 2nd edn, Breslau, 1895), trans, by Baynes R. E. as Kinetic theory of gases (London, 1899).

Meyer L. Modern theories of chemistry . trans, by Bedson P. P. and Williams W. C. from the 5th (1884) German edn (London. 1888). p. vii: “Author's preface to the fourth edition”. The publishing history of this work, which Bedson says “must be regarded as his magnum opus“, (op. cit. (ref. 169). 1405). is given by Bedson , ibid., 1429. with an analysis of changes between the 1864. 1872. 1876. 1883. and 1881 edns: And by Benfey. op. cit. (ref. 169). 353. who notes the translations into Russian from the 1st edn and into English and French from the 5th.

ibid., 1–62.

“Introduction to the first edition”, reprinted in the 5th edn. and translated in ibid., xix–xxxii.

ibid., xxii.

Partington , History (ref. 5), 889. Partington, born in 1886, was of course educated while Meyer was still exerting that influence.

Bedson , loc. cit., (ref. 171). Cf. Meyer , op. cit. (ref. 171), viii–xv.

van Spronsen J. W. , The periodic system of chemical elements: A history of the first hundred years (Amsterdam, 1969), 124–31.

Meyer L. , Outlines of theoretical chemistry, trans, by Bedson P. P. and Williams W. C. (London, 1892), v: “Author's preface to English translation”.

Remsen I. , “Lothar Meyer”, American chemical journal, xvii (1895), 471–2, p. 471.

Meyer , op. cit. (ref. 171), 31.

Wurtz A. , “Sur quelques points de philosophie chimique”, in Wurtz , Leçons de chimie professées en 1863 (Paris, 1864); idem, “Cours de philosophie chimique fait au Collège de France”, Moniteur scientifique, i (1864), 481–6, 543–53…, ii (1865), 182–7 (11 lessons); idem, “Sur l'atomicité des éléments”, Bulletin du Société chimique de Paris, ii (1864), 247–53.

Odling W. , “On the proportional numbers of the elements”, Quarterly journal of science, i (1864), 642–8. Cf. ref. 144, above.

Kekulé A. , “Sur la théorie atomique et la théorie de l'atomicité”, Comptes rendus, lx(1865), 174–7.

Williamson A. W. , “On the atomic theory”, Journal of the Chemical Society, xxii (1869), 328–65; “Discussion on Dr Williamson's lecture…Sir Benjamin C. Brodie…in the chair”, ibid., 433–41. The debate is discussed in Brock , (ed.), op. cit. (ref. 81).

Colmant P. , “Querelle à l'Institut entre équivalentistes et atomistes”, Revue des questions scientifiques, cxliii (1972), 439–519. As Colmant points out, the seventeen contributions to the debate fill seventy-two quarto pages of the Comptes rendus, lxxxiv (1877), passim. The debate was basically between Wurtz for atoms, and Berthollet and Deville; see Duhem , op. cit. (ref. 81). I am grateful to M.J. Nye for showing me an unpublished manuscript, “Berthollet's anti-atomism: A 'matter of taste?” read to the Midwest Junto, Madison, Wisconsin, April 1980.

Van Spronsen , op. cit. (ref. 177), 102–46; Partington , History (ref. 5), 886–97; Ihde , Development (ref. 5), 240–51.

Ihde , ibid., 248–9.

Bedson , op. cit. (ref. 169), 1421–2.

Muir M. M. P. , A treatise on the principles of chemistry (Cambridge, 1884), ch. 3: “The periodic law”.

ibid., 30. Note the book's dedication to Sir William Thomson.

Muir M. M. P. , Heroes of science: Chemists (London, 1883), 138.

Meldrum , op. cit. (ref. 1), 43.

Divers E. , op. cit. (ref. 55).

Clearly a great deal more work needs to be done on the gradual acceptance of Avogadro's hypothesis into chemistry, and especially into chemical textbooks, in the last third of the nineteenth century. To judge by one examination question, at least, the hypothesis had penetrated to the most elementary level of chemical instruction by 1899: “7. What is the law which is named after Avogadro? How does this law assist us in assigning a formula to hydrochloric acid gas?” Junior Cambridge local examination, “Theoretical chemistry”, December 1899. quoted in School world, i (1899), 396.

Cf. the quotation from Dumas in the text at ref. 94, above.

Ostwald W. “Elements and compounds”, Faraday lecture. 1904, Journal of the Chemical Society. lxxxv (1904). 506–22, p. 508.