¹ Roger Joseph Boscovich, ed. Whyte, L. L. (London, 1961)Google Scholar. Maxwell's famous article “Atom” in the Encyclopaedia Britannica (Ninth Edition, 1875)Google Scholar refers several times to Boscovich, but not once to Dalton.

² On the “Vortex Atom”, see Silliman, R. H., Isis, liv (1963), 461.Google Scholar

³ Annals of Philosophy, vi (1815), 321.Google Scholar Thomson had previously (ibid., ii (1813), 114) suggested that some atomic weights might be multiples of that of oxygen.

⁴ Ibid., vii (1816), III.

⁵ Elements of Chemical Philosophy (1812)Google Scholar. Davy distinguished between “undecompounded substances” and “the true elements of bodies” and supposed that the metals and inflammable bodies might be “different combinations of hydrogen with another principle as yet unknown”.

⁶ Traité Élémentaire de Chimie (1789)Google Scholar; the list of elements is headed by “Substances simples qui appartiennent aux trois règnes & qu'on peut regarder comme les élémens des corps: Lumière: Calorique: Oxygène: Azote: Hydrogène”.

⁷ A discussion of Meinecke's independence of Prout is in Prout's Hypothesis (Alembic Club Reprints No. 20).

⁸ Ann. der Physik, xxiv (1816), 162.Google Scholar

⁹ J. für Chemie und Physik, xxvii (1819), 46.Google Scholar

¹⁰ “It has been imagined by some philosophers that all matter, however unlike, is probably the same thing, and that the great variety of its appearances arises from certain powers communicated to it, and from the variety of combinations and arrangements of which it is susceptible. From the notes I borrowed from Newton in the last lecture, this does not appear to have been his idea. Neither is it mine. I should apprehend there are a considerable number of what may be called elementary principles, which are never metamorphosed, one into another, by any power we can control” (quoted by Roscoe, and Harden, , A New View of the Origin of Dalton's Atomic Theory, p. 112).Google Scholar

¹¹ History of Chemistry, 1830, 31, ii, 295.Google Scholar

¹² Dalton, like other chemists, admitted that some supposed elements might in fact be compounds, but this would be merely an error in classification. During the early part of the century there was some confusion about which substances were or were not elements; thus, chlorine and nitrogen were thought (notably by Berzelius) to be oxides of “murium” and “azotium”. These confusions, which were largely at an end by 1830, are not chronicled in this paper.

¹³ In the New System of Chemical Philosophy several pairs of elements are given the same atomic weight. Too much importance should not be attached to this, since Dalton probably regarded the numbers as provisional. Or did he envisage atoms of the same weight having: different shapes?

¹⁴ An attempt to establish the first principles of chemistry by experiment (London, 1825, 2 vols.).Google Scholar

¹⁵ Jahresbericht, vi (1827), 77.Google Scholar It seems that Berzelius later regretted the severity of this attack; see Partington, J. R., History of Chemistry, iv, 226.Google Scholar Thomson's compatriot Andrew Ure (1778–1857) also criticized the book in intemperate terms (Quart. J. Sci., xx (1825), 113).Google Scholar

¹⁶ Thomson “assumed that chemists would consider him capable of making the analysis in the way it should be made” (Annals of Philosophy, x (1825), 363)Google Scholar and not in the way described in his book.

¹⁷ See reference 37. Pettenkofer regarded Berzelius as having, by his great authority, saved chemistry from the empty speculations of Naturphilosophie which had become such a nuisance (Plage) in other sciences. He may have been right; Meinecke certainly published in Oken's journal. Prout's tendencies in this direction were checked by his adherence to Paley.

¹⁸ Even Berzelius may have caught “Multiplenfieber” in the end. In a letter to Laurent (Comptes rend., xix (1844), 352)Google Scholar, he wrote: “M. Erdmann, a Swedish chemist, has studied the atomic weight of zinc … (and) found 406·591 (0 = 100). MM. Swanberg and Norlin have determined the atomic weight of iron; the mean of 14 experiments is 349·523. The atomic weight of zinc would not be a multiple of the equivalent of hydrogen, but of the atom = 6·25. The atomic weight of iron is practically a multiple of 12·50.” However, this is so un-Berzelian that it is more likely that he was transmitting, without comment, a communication from Erdmann, who was a follower of Prout.

¹⁹ Phil. Trans. Roy. Soc., cxxiii (1833), 523Google Scholar; for an earlier paper see ibid., cxix (1829), 291.

²⁰ Phil. Trans. Roy. Soc., cxxix (1839), 13.Google Scholar

²¹ C. G. B. Daubeny (1795–1867), Professor of Chemistry and Botany at Oxford. The letter was published as an appendix to Daubeny, 's Introduction to the Atomic Theory (Oxford, 1831)Google Scholar. In this book Daubeny suggests that the prime matter need not be hydrogen “or any other of the bodies which have yet come under our cognisance”.

²² Prout independently formulated “Avogadro's Hypothesis” and was aware of the distinction between the molecule (which he usually called the atom) and the atom (which he usually called the sub-atom). He wrongly supposed the sub-atom to be capable of still further division, of the same nature.

²³ Chemistry, Meteorology, and the Function of Digestion, considered with reference to Natural Theology (London, 1833; later editions 1834, 1845, 1855).Google Scholar

²⁴ Ibid., p. 137. “Although we have rendered it probable, that the molecules of bodies considered at present as elementary, are immediately compounded of many others, more or less resembling them; yet it is obvious, that there must be a point at which these, and other elements, exist in a primary or ultimate form; and beyond which, if the elements can be supposed to be subdivided, they must become something altogether different … With respect to the nature of the sub-molecules … they may naturally be supposed to possess the most intense properties, or polarities. Indeed, such sub-molecules may be imagined to resemble in some degree, the imponderable matters, heat &c, not only by their extreme tenuity, but in other characters also; and this very intensity of property and character may be reasonably considered as one, if not the principal reason, why they are incapable of existing in a detached form. Lastly; are not these ultimate and refined forms of matter extensively employed in many of the operations of nature; and particularly in many of the processes of organization?” This is a very fair guess at the properties of free atoms and free radicals generally.

²⁵ Ibid., p. 177 (footnote). “May we not then infer, that during those periodic convulsions alluded to in the text, new elements have been developed, or old ones decomposed into others of a higher, and more elementary kind; and that in virtue of the general laws in operation, these new elements have subsequently combined to form series of new arrangements?”

²⁶ Comptes rend., xi (1840), 991.Google Scholar

²⁷ Ibid., p. 287.

²⁸ Ibid., xiv (1842), 537. Dumas was the first major chemist to follow Turner in reducing his weights to vacuum. Berzelius, with characteristic obstinacy, refused to do this, ridiculing it as “straining at a gnat and swallowing a camel”.

It will be noted that Dumas (in common with many other chemists of the time) used 6 and 8 for the atomic weights of carbon and oxygen, instead of 12 and 16. This was a consequence of the non-acceptance of Avogadro's Hypothesis; it does not greatly affect arguments about the complexity of the elements.

²⁹ de Marignac, J. C. G. (1817–1894), Professor in Geneva. Oeuvres complètes, i, 75.Google Scholar

³⁰ Redtenbacher, J. and Liebig, J., Ann. der Chemie, xxxviii (1841), 113CrossRefGoogle Scholar; J. Liebig, ibid., 195.

³¹ Ibid., p. 92.

³² Ann. chim. phys., xviii (1846), 41.Google Scholar

³³ The classical example was the benzoyl radical, discovered by Liebig and Wöhler (Ann. der Chemie, iii (1832), 249)Google Scholar. The much earlier cyanogen (CN) and ammonium (NH₄), with their marked resemblances to the halogens and the alkali metals respectively, probably had even more influence on contemporary chemical thinking.

³⁴ For example “Cyanogen” and “methyl”, which are in fact dicyanogen and ethane respectively, though this was not certainly known until the 1860's. The curious “ammonium amalgam” was also taken more seriously than it is to-day.

³⁵ Homologous series were discovered by Schiel (Ann. der Chemie, xliii, 1842, 107)Google Scholar and Dumas, (Comptes rend., xv (1843), 935)Google Scholar, and given wider currency by Laurent and Gerhardt.

³⁶ Döbereiner, J. W. (1780–1849) of Jena; Pogg. Ann., xv (1829), 301.Google Scholar

³⁷ Max von Pettenkofer (1818–1901); lecture to the Münchener Academic der Wissenschaften, 12 January 1850, printed in Münchener Gelehrten Anzeigen, xxx, 261.Google Scholar The publications of Dumas stimulated Pettenkofer to reprint his lecture in Ann. der Chemie, cv (1858), 187Google Scholar, with a short commentary and claim for priority. The idea of elements as radicals, though not worked out in any detail, is to be found in the Elements of Chemistry (2nd ed., Dublin, 1849) of SirKane, Robert (1809–1890)Google Scholar (information from Dr. D. M. Knight).

³⁸ Remarks to this effect were often made by Faraday; he made a similar observation to Crookes on the discovery of thallium. Faraday's hope was for “some more high and general power of nature even than electricity” which would split up the metals as the fixed alkalies and earths had been split.

³⁹ The Atheneum, 1851, p. 750.Google Scholar The reporter wrote plaintively “Prof. Dumas had not previously prepared diagrams or tables, but covered a large blackboard with lines, figures and formulae, to follow his train of reasoning … and symbols, volumes and names were rapidly produced and as rapidly effaced to illustrate the Professor's views”. It is not clear whether Dumas lectured in French or English.

⁴⁰ Phil. Mag., [4], v (1853), 313.Google Scholar

⁴¹ Report of the British Association, 1858, 57.Google Scholar According to Parnell, E. A., The life and labours of John Mercer (London, 1886)Google Scholar there were unpublished memoranda which carried his speculations further.

⁴² Comptes rend., xlv (1857), 74Google Scholar; xlvi (1858), 951, 1026; Ann. chim. phys., lv (1859), 129.Google Scholar

⁴³ Am. J. Sci., xvii (1854) (2), 387.Google Scholar There were numerous attempts in the next few years to expand Döbereiner's triads to tetrads, to make triads of triads, etc., but little thought was given to the possible causes underlying these numerical games. See for example, Odling, , Phil. Mag., [4], xiii (1857), 423CrossRefGoogle Scholar; Lennsen, , Ann. der Chemie, ciii (1857), 121.CrossRefGoogle Scholar

⁴⁴ Comptes rend., xlvii (1858), 746Google Scholar; xlviii (1859), 362. Replies by Dumas: ibid., xlviii (1859), 139, 372. Partington, J. R. (History of Chemistry, iv, 885)Google Scholar erroneously ascribes to Despretz the exactly contrary opinion.

⁴⁵ Quoted by Mendeleev, D. I., The principles of Chemistry, Third English Ed. (London, 1905), appendix 2.Google Scholar

⁴⁶ Reports of … the British Association, 1885, p. 969.Google Scholar Carnelley (1854–90) was Professor of Chemistry successively at Sheffield, Dundee and Aberdeen.

⁴⁷ Bull. Acad. roy. Belg., x (1860) (2), 208.Google Scholar

⁴⁸ J. W. Mallet (1832–1912), Professor in the University of Virginia; quoted in the Stas Memorial Lecture (J. Chem. Soc., 1893, 1)Google Scholar. The original is “Il faut croire qu'il y a quelque chose là-dessous”. It is interesting to note that long experience of atomic weights had the opposite effect on F. W. Clarke (1847–1931), who was almost the peer of Stas as an exact analyst. “I began this recalculation of the atomic weights with a strong prejudice against Prout's hypothesis, but the facts as they came before me have forced me to give it a very respectful consideration” (A recalculation of the atomic weights, Smithsonian Institution, 1882, p. 280).Google Scholar

⁴⁹ “Commentary on the foregoing paper” (signed “C.M.”); Oeuvres complètes, i, 693.Google Scholar

⁵⁰ Aston, F. W., Isotopes (London, 1922), p. 101Google Scholar; “In the nuclei of normal atoms the packing of the electrons and protons is so close that the additive law of mass will not hold, and the mass of the nucleus will be less than the sum of the masses of its constituent charges.” The idea of the interconversion of mass and energy can be traced back at least as far as a little-known Scottish writer, Alexander Stephen Wilson, whose book, The unity of matter (1855) expounds what is almost a theory of continuous creation. It is not likely that either Marignac or Mendeleev had read this book.

⁵¹ A further exchange of papers between Stas (Mem. Acad. roy. Belg., xxxv (1865), 3)Google Scholar and Marignac (Oeuvres complètes, ii, 281)Google Scholar altered their respective positions very little, though the latter had to admit that variability of composition was not significant in the compounds used by Stas.

⁵² Sunti di un corso di filosofia chimica (1858)Google Scholar, a pamphlet which was distributed at the Karlsruhe conference of 1860.

⁵³ Comptes rend., liv (1862), 757Google Scholar, and later papers.

⁵⁴ On the discovery of the periodic law (London, 1884).Google Scholar

⁵⁵ Ann. der Chemie (Suppl. Band), vii (1870), 354.Google Scholar The paper begins: “That the hitherto undecomposed chemical elements are absolutely undecomposable is nowadays very improbable, to say the least. It would seem rather that the atoms of the elements are not the ultimate, but only the proximate components of the molecules both of compounds and elements … which in their turn consist of particles of matter of a third, higher, order.”

⁵⁶ Ibid.(Suppl. Band), viii (1872), 206.Google Scholar By “Prout's hypothesis” Mendeleev evidently meant the proposition “atomic weights are exact integers” and not “the atoms of the elements are made up of atoms of a prime matter”. Unless this is assumed the passage is unintelligible.

⁵⁷ An ingenious extension of this idea was attempted by Risteen, A. D., The molecular theory of matter (Boston, Mass., 1896), p. 160.Google Scholar “… it is just possible that… an atom of weight A, when combining with another of weight B, does not produce a molecule of weight A + B. I am well aware that this would make perpetual motion possible, for if the weight of the given substances happened to be greater in the combined state than in the uncombined one, we should only have to let them fall some convenient distance while they are combined, and raise them again while they are uncombined, and we should gain a little energy every time the cycle was repeated; while if combination should cause loss of weight instead of gain we could attain the same end by performing the cycle in the opposite direction … If it be true, therefore, that matter is composed of some fundamental substance combined with itself in varying degrees of complexity, then whenever the law of conservation of weight would be violated on splitting a body up into its constituents, or in forming it from them, the means at our disposal can never enable us to effect either the separation or the combination; and so far as we are concerned, such a body would forever remain an element … This hypothesis explains both the existence of ‘elements’, and slight deviations from integral values that we find in the atomic weights.” The fallacy lies in the assumption that the same energy would be required for separation as would be liberated in the combination of the ultimate particles.

⁵⁸ J. Chem. Soc., 1889, 634.Google Scholar

⁵⁹ Quoted by Gladstone, reference 40.

⁶⁰ Phil. Trans. Roy. Soc., clxxi (1880), 1033.Google Scholar

⁶¹ Phil. Mag. [6], i (1901), 311.Google Scholar

⁶² Bull. de l'Acad. roy. de Danemark, 1894, p. 325Google Scholar; abstracted in Z. phys. Chem., xvii (1895), 354.Google Scholar Thomsen earlier published a pamphlet Om Materiens Enhed (1887), which the writer has not seen.

⁶³ Programme der Atommechanik (Iowa, 1867)Google Scholar; The true atomic weights of the chemical elements and the unity of matter (St. Louis, 1894)Google Scholar; The absolute atomic weights of the chemical elements (St. Louis, 1901)Google Scholar; The proximate constituents of the chemical elements (New York, 1904)Google Scholar; La matière est une (Paris, 1907).Google Scholar

⁶⁴ Elements of Chemistry (1842), p. 122.Google Scholar

⁶⁵ Proc. Roy. Soc., xii (1862/1863), 620.Google Scholar

⁶⁶ Phil. Trans. Roy. Soc., clv (1865), 71Google Scholar; J. Chem. Soc., 1866, 230.Google Scholar An earlier version appears in Ann. der Chemie, cxxvi (1863), 362.Google Scholar

⁶⁷ Dumas had a similar idea; see Leçons sur la philosophie chimique (1837), p. 278.Google Scholar

⁶⁸ The exceptions to the law of Dulong and Petit were explained much later in terms of the quantum theory.

⁶⁹ Farrar, W. V., Chymia, ix (1964), 169CrossRefGoogle Scholar, and references there cited.

⁷⁰ Ber. Dtsch. Chem. Ges., xii (1879), 1426Google Scholar; J Chem. Soc., 1879, 673.Google Scholar

⁷¹ Chem. News, xv (1867), 315Google Scholar; a lecture entitled “The chemistry of the primeval earth”.

⁷² Director of the Astrophysical Observatory, South Kensington. Lockyer (1868) discovered helium spectroscopically in the atmosphere of the Sun. This was the only authentic discovery among several reported by spectroscopists of “new” elements existing outside the earth. These supposed elements (coronium, nebulium, etc.) provided a fine set of pseudoproblems for speculators.

⁷³ Phil. Trans. Roy. Soc., clxiv (1874), 491.Google Scholar Clarke (reference 48) expressed similar views in a popular article a few months earlier (see J. Amer. Chem. Soc., lvii (1935), suppl. p. 21).Google Scholar

⁷⁴ The chemistry of the Sun (London, 1887)Google Scholar; Inorganic evolution as studied by spectrum analysis (London, 1900)Google Scholar. There is little new material in the latter book.

⁷⁵ Lockyer would have been delighted by the recent discovery of water in “cool” stars.

⁷⁶ Comptes rend., lxxvii (1873), 1352.Google Scholar Berthelot must not on this account be thought of as opposed to the general idea of the complexity of the elements. In another context (Ibid., xc (1880), 1512) he wrote of atoms as “highly complex structures, endowed with a specific architecture, and animated with a variety of internal motions”.

⁷⁷ For instance, Kayser, H.'s influential Lehrbuch der Spektralanalyse (Berlin, 1883), pp. 201–202.CrossRefGoogle Scholar

⁷⁸ Phil. Mag., [5], xxi (1886), 151.Google Scholar

⁷⁹ Reports of … the British Association, 1886, 558Google Scholar; see also J. Chem. Soc., 1888, 487.Google Scholar

⁸⁰ It is strange that Crookes, with his great experience of fractionation, should never have thought of an alternative explanation, which would now be considered correct; that in any process of geochemical fractionation, materials of similar properties will tend to segregate together.

⁸¹ Der genetische System der chemischen Elemente (Berlin, 1893)Google Scholar. Thierry William Preyer was born in Manchester and educated in England, but became professor of Physiology in Jena.

⁸² For example, Maxwell (reference I).

⁸³ Bull. Acad. roy. Belg., xix (1865) (2), 411.Google Scholar

⁸⁴ Chem. News, xlv (1882), 50.Google Scholar Here again we have able chemists overestimating the accuracy of their analytical methods. As late as 1893 the experienced J. A. Wanklyn was claiming, on the basis of some hydrocarbon analyses, that the atomic weight of carbon was 6 (Phil. Mag., [5], xxxvi (1893), 552).Google Scholar

⁸⁵ Bull. soc. chim., xxxix (1883), 263.Google Scholar

⁸⁶ Gumilevskiĭ, L., Aleksandr Mikhailovich Butlerov (Moscow, 1951).Google Scholar

⁸⁷ Variable atomic weights are also discussed by E. Vogel, of San Francisco (Nature, xxxi (1884), 42)Google Scholar, but his arguments are impenetrable.

⁸⁸ Reports of … the British Association, 1882, p. 483.Google Scholar

⁸⁹ C. F. Schönbein (1799–1868) always maintained that chlorine was an oxide. G. Ciamician (1857–1922) was misled by the early results of spectrum analysis into the belief that aluminium, for example, was an oxide of boron. Both were chemists of ability and reputation.

⁹⁰ An enquiry into the simple bodies of chemistry (1844).Google Scholar

⁹¹ Trans. Roy. Soc. Edinburgh, 1841, 231.Google Scholar

⁹² Phil. Mag., [3], xxiv (1844), 296.Google Scholar

⁹³ R. H. Brett and J. D. Smith, ibid., [3], xix (1841), 295.

⁹⁴ Mem. Proc. Manchester Lit. Phil. Soc., xvii (1878), 194.Google Scholar On Wilde see W. W. Haldane Gee, ibid., lxiii (1920), No. V; SirSchuster, A., Biographical Fragments (London, 1932).Google Scholar

⁹⁵ Original Research; the governing principles of the elements (London, 1878)Google Scholar. Nothing is known of Masson.

⁹⁶ Chem. News, lxxi (1895), 67.Google Scholar It is of interest that Rayleigh began his work on the density of nitrogen (which led to the discovery of the inert gases) as yet another attempt to confirm or refute Prout's hypothesis (reference 88, p. 440).

⁹⁷ Bolton, H. Carrington, Chem. News, lxxvii (1898), 3, 16, 69, 73.Google Scholar

⁹⁸ Samlade Skrifter av A. Strindberg, vol. xxvii (Stockholm, 1919).Google Scholar

⁹⁹ Proc. Roy. Soc., xix (1871), 236.Google Scholar

¹⁰⁰ Phil. Mag., [5], lxiv (1897), 293.Google Scholar

¹⁰¹ Faraday Lecture, J. Chem. Soc., 1881, 277.Google Scholar

¹⁰² Rév. gén. Sci., x (1899), 41.Google Scholar

¹⁰³ J. Chem. Soc., 1902, 231.Google Scholar According to Soddy, when he put this interpretation to Rutherford, the latter shouted “‘For Mike's sake, Soddy, don't call it transmutation. They'll have our heads off for alchemists. You know what they are!’ After which he went waltzing round the laboratory, his huge voice booming ‘Onward Christian so-ho-hojers …” (Howorth, Muriel, Pioneer Research on the Atom (London, 1958)).Google Scholar

¹⁰⁴ History of Chemistry (English edition, 1888), p. 201.Google Scholar