548 



NATURE 



[April 7, 189S 



all as constituted of the same matter, affected by attributes 

 more or less fortuitous and accidental. Thus, all the varieties 

 of inflammable air were at bottom identical, with properties 

 modified by their origin or their varying content of the hypo- 

 thetical principle phlogiston — that is the principle that was 

 assumed to make them burn. 



From Watt's published correspondence we are able to judge 

 how he regarded Priestley's further work on this so-called con- 

 version of water into air. He admits that the facts are " in 

 some degree contradictory to each other." The apparent con- 

 version would seem to depend upon the material of the vessel in 

 which it was made. In a glass vessel no air was produced, nor 

 was any found in a gun-barrel when the distillation was done 

 slowly ; but when confined by a cock, " and let out by puffs, it 

 produces much air; which," sa^s Watt, "agrees with my 

 theory, and also coincides with what I have observed in steam- 

 engines. In some cases I have seen the tenth of the bulk of 

 the water, of air extricated or made from it." Davy once said 

 " the human mind is governed not by what it knows, but by 

 what it believes ; not by what it is capable of attaining, but by 

 what it desires." However willing to catch at anything in 

 support of his belief, it is possible that Watt might have been 

 led to doubt the soundness of Priestley's experiment, if an 

 apparent and wholly -unlooked for confirmation of it had 

 not arisen. 



To make the account exact, and in view of what is to follow, 

 it is necessary to go back a little, in point of time. In the 

 spring of 1781, Priestley performed what he styled "a mere 

 random experiment made to entertain a few philosophical 

 friends." It was practically a repetition of Volta's experiment 

 of firing a mixture of the inflammable air from metals, that is, 

 hydrogen, with common air in a closed glass vessel by means 

 of the electric spark. After the deflagration the vessel was 

 found to be hot, and on cooling its sides were observed to be 

 bedewed. Neither Priestley nor any of his philosophical friends 

 seem to have paid particular attention to the deposit of 

 moisture, or, at all events, if ihey did they failed to perceive its 

 significance. One of them, however, Mr. John Warltire, a 

 lecturer in natural philosophy in Birmingham, imagined that 

 the experiment might afford the means of showing whether heat 

 was ponderable or not ; and accordingly he repeated it, using 

 for greater safety a copper globe, weighed before and after the 

 passage of the spark. A minute loss of weight was always 

 noticed, "but not constantly the same ; upon the average it was 

 about 2 grains."^ 



Priestley, who, with Withering, was present when the ex- 

 periments were made, confirmed the apparent loss of weight ; 

 but he added, with a caution that was not characteristic, that he 

 did not think " that so very bold an opinion as that of the 

 latent heat of bodies contributing to their weight should be re- 

 ceived without more experiments, and made upon a still larger 

 scale." 



Priestley's volume — the sixth in the series— was published in 

 1781, and was certainly known to Watt; indeed, in the Ap- 

 pendix are printed a number of observations made by him 

 apparently as the work was passing through the press. Al- 

 though, therefore, he must have had his attention drawn about 

 this time to the formation of the dew in Priestley and Warl- 

 tire's experiment, there is nothing to show that he attached any 

 importance to the circumstance, or that, if he did, he dissented 

 from Warltire's conclusion that common air deposits its moisture 

 when it is phlogisticated. 



For some time previous to the publication of Priestley's book, 

 Mr. Cavendish was engaged upon an inquiry "to find out the 

 cause of the diminution which common air is well known to 

 suffer by all the various ways in which it is phlogisticated, and 

 to discover what becomes of the air thus lost or condensed." 

 In other words, it was an investigation to determine the 

 changes experienced by air when bodies were made to burn in 

 confined portions of it. On the appearance of Priestley's book 

 he repeated Warltire's experiment, thinking "it worthwhile to 

 examine more closely, as it seemed likely to throw great light 

 on the subject I had in view." He confirmed the observation 

 on the formation of dew ; but although he made the experiment 

 on a larger scale, and with varying proportions of the two airs, 

 he was unable to satisfy himself as to the loss of weight after the 



1 The account of these experiments is given in a letter to Priestley, and 

 constitutes No. v. of the " Appendix to Priestley's Experiments and Ob- 

 servations relating to various branches of Natural Philosophy, &c.," vol. ii. 

 (Birmingham, 1781). 



NO. 1484, VOL. 57] 



explosion. As the result of a number of trials, made both with 

 the inflammable air from zinc and from iron — that is, hydrogen 

 — and mixed with common air in the proportion of 423 

 measures of the inflammable air to 1000 of common air, he 

 says, " we may safely conclude that when they are mixed in this 

 proportion, and exploded, almost all the inflammable air and 

 about one- fifth part of the common air lose their elasticity, and 

 are condensed into the dew which lines the glass." In order to 

 examine the nature of this dew, large quantities of the hydrogen 

 were burnt with two and a half times its volume of common air, 

 and the product of the combustion was caused to pass through a 

 long glass tube whereby it was condensed. " By this means 135 

 grains of water were condensed in the cylinder [»'. e. the tube], which 

 had no taste nor smell, and which left no sensible sediment when 

 evaporated to dryness ; neither did it leave any pungent smell 

 during the evaporation ; in short, it seemed pure water. . . . 

 By the experiments with the globe, it appeared that when the in- 

 flammable and common air are exploded in a proper proportion, 

 almost all the inflammable air and nearly one-fifth of the com- 

 mon air, lose their elasticity, and are condensed into dew. And 

 by this experiment it appears that this dew is plain water, and 

 consequently that almost all the inflammable air and about one- 

 fifth of the common air are turned into pure water." 



The idea that common air was for the most part a mixture of 

 two gases — oxygen or the dephlogisticated air of Scheele and 

 Priestley, and nitrogen or the mephitic air of Rutherford, the 

 azote of Lavoisier — was familiar to chemists at this period as the 

 result of the teaching of Scheele and Lavoisier, and there is 

 reason to suppose that this opinion was shared by Cavendish. 

 He had been engaged for some time past in an elaborate inquiry 

 into the constitution of atmospheric air, the results of which 

 admitted of no other interpretation than that common air was 

 composed of two different gases, mixed or combined in constant 

 relative proportions. It is true that in the memoir containing 

 the results of his inquiry he nowhere directly gives his estimate of 

 these relative quantities, but, from the data he affords, it is easy 

 to deduce the amount and the constancy of the proportion. 

 Cavendish's papers are characterised by remarkable conciseness 

 and brevity; an experiment which must have involved the putting 

 together of elaborate and complicated apparatus, and which 

 must have occupied considerable time in its performance, is de- 

 scribed in a few lines, and hence it is not always possible to 

 gather with certainty the precise disposition of the arrangements. 

 He never sets out his reasons or his conclusions with any great 

 amount of detail, and his published words occasionally give 

 little indication of his line of thought. But that he clearly 

 recognised that only one portion of common air was concerned in 

 the formation of water, and that this portion was the dephlogist- 

 icated air, or oxygen, is obvious from the next series of experi- 

 ments in which he fired a mixture of about two measures of 

 hydrogen and one measure of oxygen in a previously exhausted 

 glass globe furnished with an apparatus for firing air by electricity. 

 When the included air was fired, almost all of it lost its elasticity, 

 so that fresh quantities of the explosive mixture could be intro- 

 duced and the process repeated until a sufficient quantity of the 

 moisture was obtained for examination. In these experiments 

 Cavendish clearly and definitely demonstrated that the weight 

 of the water was practically equal to the weight of the mixed 

 gases which had combined to form it. In some cases the water 

 was perfectly neutral in its reaction ; in others it was slightly 

 acid, and the cause of this acidity caused Cavendish much ex- 

 perimenting, but he is never in any doubt as to the main result ; 

 he says distinctly, "if those airs could be obtained perfectly 

 pure, the whole would be condensed." Now if Cavendish had 

 published this main result at the time he obtained it, namely in 

 the summer of 1781, or even if he had formally communicated 

 it to one of the meetings of the Royal Society during the 

 ensuing session, there would have been no Water Controversy. 

 But even if he were ready, it was characteristic of him to delay, 

 not from inertia or indolence, but from a morbid shyness, an 

 unconquerable reticence, which constantly led him to postpone 

 any public announcement of his work. He had the additional, 

 and to him all-sufficient, reason that he had not yet worked out 

 the cause of the occasional acidity of the water. What he did, 

 however, was to communicate the facts of his experiments to 

 Priestley, as Priestley himself states in a subsequent paper 

 published in the Philosophical Transactions for 1 783. When or 

 how he communicated them to Priestley does not appear, nor 

 have we any means of knowing precisely what was said. 

 Something, however, on this point may be inferred from what 



