by a relabelling exercise: in 1787 the French chemist proposed a new system of nomenclature in his magisterial Traité elementaire de chimie, the adoption of which would make it virtually impossible to practice chemistry without implicity endorsing oxygen. Imagine what would happen, Cavendish complained, if everyone who came up with a new theory concocted a new terminology to go along with it. In the end, chemistry would become a veritable Tower of Babel in which no one could understand anyone else. He derided the ‘rage of name-making’ and dismissed Lavoisier’s Traité as a mere ‘fashion’. Until there were experimental results that could settle such disputes, he said, it was better to stick with the tried-and-tested terminology, since new names inevitably prejudice the very terms within which theoretical questions can be framed.
That Cavendish’s opposition was not motivated by mere traditionalism is clear from the fact that he gradually abandoned phlogiston and accepted Lavoisier’s oxygen as the evidence stacked up in the French chemist’s favour. Even in 1785 he was prepared to concede that phlogiston was a ‘doubtful point’, and by early 1787 the phlogistonist Richard Kirwan in England wrote to Louis Bernard Guyton de Morveau, a colleague of Lavoisier’s, saying that ‘Mr Cavendish has renounced phlogiston.’ By the turn of the century, Cavendish was prepared even to use Lavoisier’s terms: dephlogisticated air became oxygen, and inflammable air was hydrogen – the gas which, thanks to the researches of Warltire, Priestley and Cavendish as well as his own, Lavoisier saw fit to call the ‘water former’.
But that is rather leaping ahead of the matter. In the late 1770s Lavoisier decided that, since his oxygen was the principle of acidity, its combination with hydrogen should produce an acid. In 1781–2 he looked for it in experiments along the same lines as Priestley and Warltire, but saw none. Working with Laplace, he combined oxygen and hydrogen in a glass vessel and found that their combined weight was more or less equal to that of the resulting water.
They were not the only French scientists to try it. When Joseph Priestley conducted further experiments of this kind in March 1783, the French scientist Edmond Charles Genet in London wrote a letter describing the work to the French Académie des Sciences, the equivalent of the Royal Society. Genet’s letter was read to the academicians in early May. Lavoisier was there to hear it, and so was the mathematician Gaspard Monge from the military school of Mézières, who promptly repeated the experiment in June.
Lavoisier and Laplace did likewise – but by then they knew of Cavendish’s results too, for Charles Blagden told them about his colleague’s investigations in early June while on a trip to Paris. Lavoisier, as ambitious as Cavendish was diffident, quickly repeated the measurements on 24 June (forgoing, in haste, his usual quantitative precision) and presented them soon after to the Académie. He referred to the earlier work by both Monge and Cavendish, magisterially indicating that he ‘proposed to confirm’ Cavendish’s observations ‘in order to give it greater authority’.
Curiously, Lavoisier continued at this point to call oxygen ‘dephlogisticated air’ – but for him this was more or less just a conventional label, and did not oblige him to fit phlogiston into his explanations. That enabled him to see through to the proper conclusion with far more directness and insight than Cavendish. ‘It is difficult to refuse to recognize’, he said, ‘that in this experiment, water is made artificially and from scratch.’
And, in a master stroke, he verified that this was so by showing how water might be split into its two constituents. Lavoisier felt that the right way to investigate the composition of matter was to come at it from both directions: synthesis, or making a substance from elemental components, and analysis, which meant separating the substance into those fundamental ingredients. He described how, in collaboration with the engineer Jean Baptiste Meusnier, he ‘analysed’ water by placing it together with iron filings in an environment free of air, held in an inverted bowl under a pool of mercury. The iron, he reported, was converted into rust, just as it is when it absorbs dephlogisticated air (that is, oxygen) from common air; and ‘at the same time it released a quantity of inflammable air in proportion to the quantity of dephlogisticated air which had been absorbed by the iron.’ ‘Thus’, he concluded, “water, in this experiment, is decomposed into two distinct substances, dephlogisticated air . . . and inflammable air. Water is not a simple substance at all, not properly called an element, as had always been thought.”
Cavendish’s experiment was beautiful because of his attention to detail, a characteristic that redirected attention towards the formation of water and pointed clearly to the conclusion that Lavoisier subsequently drew. But Lavoisier’s follow-up studies surely deserve a share of that beauty, because of the way he found the right interpretation and then went on to make it irrefutable.
Needless to say, not everyone saw it quite like that. The shroud of phlogiston that made Cavendish’s explanation of his experiment somewhat ambiguous also helped to protect him from the kind of reactionary responses that Lavoisier’s starker message attracted. An English chemist named William Ford Stevenson showed how reluctant some scientists were to abandon the ancient elemental status of water when he called Lavoisier’s claims a kind of ‘deception’. How on Earth could water, which puts out fires and was for that reason ‘the most powerful antiphlogistic we possess’, how could this substance truly be compounded from an air ‘which surpasses all other substances in its inflammability’? Cavendish betrayed that he had not quite grasped the true implications of his results when he too expressed doubts about Lavoisier’s conclusions. Priestley, a staunch believer in phlogiston, had no time for them. Blagden, meanwhile, was more angered (and with some justification) by Lavoisier’s failure to give sufficient credit to what Cavendish had already achieved – although at that point Cavendish had still not submitted his report to the Royal Society.
Water wars
Even that was not the full extent of the controversy. No sooner had Cavendish’s paper finally been read to the Royal Society in January 1784 than it awakened a new dispute. The Swiss scientist Jean André De Luc heard about the report and asked Cavendish for a copy, whereupon he wrote to his friend James Watt, suggesting that Cavendish was a plagiarist who had copied Watt’s ideas ‘word for word’. For Watt had repeated Warltire’s experiment several years earlier while he was still a university technician at Edinburgh, working under Joseph Black. It was not so much the experiment itself that incited De Luc’s charges, but Cavendish’s interpretation in terms of ‘dephlogisticated water’, which seemed very much along the lines of what Watt had deduced: he had claimed that water was a compound of pure air and phlogiston.
At least, that is what some historical accounts indicate; but again, there is ambiguity about whether Watt truly identified water as a substance produced by the chemical reaction of two ‘elements’. Drawing on Joseph Priestley’s experiments in early 1783 on the spark ignition of dephlogisticated and inflammable air (which were themselves stimulated by Cavendish’s still unpublished work), Watt suggested in April of that year that ‘water is composed of dephlogisticated and inflammable air, or phlogiston, deprived of part of their latent heat.’ Is this a statement that water is a compound substance? Latent heat is the heat a gas releases when it condenses into a liquid – and so Watt’s conclusion seems to blend notions about both the combination of two gases and the condensation of water. It’s hard to know quite what to make of it.
At that time, Watt expressed his ideas about water in letters to Priestley, De Luc and Joseph Black. He had intended that they be read out formally to the Royal Society, but then withdrew his formal communication after learning that Priestley’s further investigations seemed to point to some inconsistencies with other ideas that Watt’s letter contained. Yet when De Luc saw Cavendish’s report, he decided that it had appropriated Watt’s ‘theory’ without attribution.
Watt was annoyed, although unable to conclude for sure that intellectual theft was involved. ‘I by no means wish’, he wrote to De Luc,
to make any illiberal attack on Mr C. It is barely possible he may have heard nothing of
