1825.] of Claude-Louis Berthollet. 173 



ing. Since, therefore, in this mass of oxidized tin, much of the 

 metal had remained uncalcined, the solution, conformably to 

 Berthollet's theory, ought to contain an oxide of tin at a mini- 

 mum, or at least at one of the infinitely numerous inferior degrees 

 of oxidation. But so far is this from being the case, that the 

 solution when examined with reagents is found to hold a muriate 

 of tin at exactly a maximum of oxidation. The slow calcination 

 of tin, therefore, does not afford the slightest evidence of an 

 ascending oxidation. 



From the results of this experiment, and from the investiga- 

 tion of other imperfect calcinations, Proust's theory was placed 

 on a tolerably broad foundation, and he soon gained another 

 advantage by an experiment of Berthollet's, which, had it suc- 

 ceeded as the latter had expected, must have established his 

 theory, but which proved so untractable as to involve him in no 

 small difficulty, and from which it required all his ingenuity to 

 extricate himself w ith any eclat. Berthollet took a solution of 

 nitrate of mercury, in which he presumed the acid might be 

 obtained combined with the metal inevery stage of its oxidation 

 between the maxima and minima proportions. He very reason- 

 ably inferred, therefore, that by adding muriate of soda to the 

 solution, a variety of analogous compounds might be formed of 

 muriatic acid with these various oxides of mercury. He made 

 the experiment with every caution, but his exterunlor contingent 

 principles constantly interfered with the pure operation of 

 affinity, for the results were only two compounds, calomel and 

 corrosive sublimate. Both of these are well-defined and inva- 

 riable combinations of oxygen and mercury with muriatic acid 

 (to use the chemical language of that period), the first being an 

 union of the acid witli the metal at a minimum ; the second of 

 the acid with the metal at a maximum of oxidation. Proust's 

 explanation of this apparent anomaly was a plain and obvious 

 one. All solutions of the nitrate of mercury must consist, 

 besides the acid, of the metal at a maximum of oxidation, or at 

 a minimum of oxidation, or finally of a mixture of these oxides. 

 In decomposing these solutions with muriate of soda, the product 

 of the first will be muriate of oxide of mercury at a maximum of 

 oxidation, or corrosive sublimate ; of the second, the product 

 will be muriate of the oxide of mercury at a minimum of oxida- 

 tion, or calomel ; and the products of the third will be a mixture 

 of the two substances just mentioned, or both corrosive subli- 

 mate and calomel, in every possible proportion. 



The only means by which Berthollet could account for these 

 results was to suppose that the mercury assumes these tv/o con- 

 stant states of composition only at the instant when it is on the 

 point of separating itself into two combinations, and that these 

 two definite oxides are formed only at the very point of time 

 when the muriatic acid decides their separation into a soluble 



