he took occasional pupils, such as Mitscherlich and Wöhler.
Berzelius first learned of Dalton when planning his own influential textbook, Larbok i kemien, the first volume of which was published in 1808. Somehow Berzelius had acquired a copy of Richter’s writings on stoichiometry (he remarked on how unusual this was) and so learned of the law of reciprocal proportions and of the idea of equivalents. He saw immediately how useful these generalizations were for analytical chemistry. An avid follower of British chemical investigations, Berzelius learned of Dalton’s theory when he read a reference to it in Wollaston’s report on multiple proportions in Nicholson’s Journal. Because of the European wars, which made scientific communication difficult, he was unable to obtain a copy of Dalton’s New System (from Dalton himself) until 1812. Nevertheless, just from Wollaston’s brief account he saw immediately that a corpuscular interpretation of these analytical regularities was ‘the greatest step which chemistry had made towards its completion as a science’.
His own analytical results more than confirmed that, whenever substances combined together in different proportions, they were always, as Dalton had already concluded, in the proportions A + B, A + 2B, 2A + 3B, A + 4B, etc. Berzelius reconciled this regularity with Berthollet’s views on the influence of mass in chemical reactions. He agreed that Berthollet was right in supposing that substances could combine together in varying proportions; but these proportions were never continuously variable, as Berthollet had argued against Proust, but fixed according to Dalton’s corpuscular ratios.
Berzelius’ teaching duties included the training of pharmacists. He was, therefore, conscious of the fact that the Swedish Pharmacopoeia had not been revised since the days of phlogiston chemistry and that by 1810 its language had become embarrassingly out of date. In 1811, in an attempt to persuade the government to make a sensible decision on its pharmaceutical nomenclature, Berzelius devised a new Latin classification of substances, which exploited the electrochemical phenomena that he and Davy had studied, and firmly founded the organization of ponderable matter on the dualistic system that lay at the basis of Lavoisier’s antiphlogistic nomenclature.
Ponderable bodies were divided into two classes, ‘electropositive’ and ‘electronegative’ according to whether during electrolysis they were deposited or evolved around the positive or negative pole. Since these definitions reversed
FIGURE 4.1 Berzelius’ classification of substances. (Based on C. A. Russell, Annals of Science, 19 (1963): 124.)
the convention that Davy had already introduced, Berzelius was soon obliged to conform to the definition that electropositive substances were attracted to the negative pole. It was because of the theoretical implications of galvanic language that Faraday, in 1832, introduced the valueneutral nomenclature of electrodes, cathodes, anodes and so on. Berzelius’ electropositive and electronegative substances then became anions and cations respectively.
Oxygen, according to Berzelius, was unique in its extreme electronegativity. Other, less electronegative substances, like sulphur, could be positive towards oxygen and negative towards metals. On combination, a small residual contact charge was left, which allowed further combination to occur to form salts and complex salts. Thus, electropositive metals might form electropositive (basic) oxides (as electrolysis demonstrated), which would combine with electronegative acidic oxides to form neutral salts. The latter, however, might still have a residual charge that allowed them to hydrate and to form complex salts:
The scheme allowed the elements to be arranged in an electrochemical series from oxygen to potassium, based upon the electrolytic behaviour of elements and their oxides. Because salts were defined as combinations of oxides, Berzelius had to insist for a long time that chlorine and iodine were oxides of unknown elements, and that ammonia was similarly an oxide of ‘ammonia’. It was not until the 1820s that Berzelius finally capitulated and agreed that chlorine, iodine and bromine (which he placed in the special category of forming electronegative ‘haloid’ salts) were elements and that ammonia was a compound of nitrogen and hydrogen only.
It was this electrochemical system which was to have far-reaching analogical implications for the classification of organic substances. It also allowed Berzelius in 1813 to introduce a rational symbolism based upon the Latin names of the elements. Compounds were denoted by a plus sign between the constituents, as in copper oxide, Cu + O, the electropositive element being written first. Later, Berzelius dispensed with the plus sign and set the two elements side by side as in algebra. Different numbers of elements were then indicated by superscripts, e.g. S2O3, a molecule of ‘hyposulphuric acid’. These joined symbols, which were criticized initially for being potentially confusing with algebraic symbolism, only began to be used in the 1830s. It was Liebig who, in 1834, introduced the subscript convention we still use today, though French chemists went on using superscripts well into the twentieth century. Because of the importance of oxygen in Berzelius’ system, he abbreviated it to a dot over its electropositive congener, i.e. Cu = Cu + O. In 1827 he extended this to sulphur, which was indicated by a comma, i.e. copper sulphide, Cú.
In a further ‘simplification’, which in practice wrought havoc in the classification of organic compounds and in communication between chemists, Berzelius in 1827 introduced ‘barred’ or underlined symbols to indicate two atoms of an element. (Since the bar was one-third up the stem of the symbol it involved printers making a special type, thereby losing one advantage over Dalton’s symbols; hence the use of underlined symbols in some texts.) The symbols for water and potash alum thus became, respectively:
Although Berzelius introduced symbols as a memory aid to chemical proportions, they were initially adopted by few chemists. Berzelius himself virtually ignored his own suggestions until 1827, when he published the organic chemistry section of his textbook, which appeared in German and French translations soon afterwards. Indeed, the development of organic chemistry was undoubtedly the key factor into pushing chemists into symbolic representations. Following the determination of a group of younger British chemists to introduce Continental organic research into Britain, Edward Turner employed Berzelius’ symbols in the fourth edition of his Elements of Chemistry in 1834. From then on, together with chemical equations, whose use in Britain was pioneered by Thomas Graham, symbols became an indispensable part of chemical communication.
TABLE 4.2 The development of the chemical equation.
As we have seen, Dalton angrily rejected Berzelius’ symbols mainly on the grounds that they did not indicate structure but were merely synoptic. Nor was he at all pleased with the way Berzelius had taken over his creation and transformed it electrochemically. On his part, Berzelius, after struggling for years to obtain a copy of Dalton’s New System, expressed deep disappointment with the book when he eventually read it in 18127:
I have been able to skim through the book in haste, but I will not conceal that I was surprised to see how the author has disappointed my hopes. Incorrect even in the mathematical part (e.g. in determining the maximum density of water), in the chemical part he allows himself lapses from the truth at which we have the right to be astonished.
Berzelius’ extensive account of his interpretation of Dalton’s theory was published in English in Thomas Thomson’s monthly Annals of Philosophy
