On definite Proportions. 40? 



was the small quantity of potassium contained in the amal- 

 gam ; for I have often been obliged to take 60 grammes, 

 or more, in order to work with one-third of a gramme of 

 potassium. If now the larger weights are uncertain even 

 to Y q ,*, p p , we may easily obtain, when they are changed 

 after the extraction of the potassium, a resuit too great or 

 too small by a milligramme, which is of some importance 

 in the calculation. 1 therefore always employed, in weighing 

 the amalgam, some small weights, which amounted to little 

 more than the potassium I expected to find, in order to be 

 able to make as few changes as possible. The different 

 degrees of dryness of the amalgam, before and after the ex- 

 periment, may be a second cause of error. I therefore left 

 the amalgam, which I obtained, in a small well stopped 

 vessel, of which it filled four-fifths, for some time, on a very 

 hot sand-bath, so that all the water adhering to it was de- 

 composed by the potassium ; I then poured the pure mass 

 into a little phial, which it filled to the neck, and weighed it. 

 After the extraction of the potassium, I dried the mercury' 

 again in a strong heat, so that it was completely freed from 

 water. The amalgam must be weighed in a stopped [Sw. 

 not "dried" as G.] vessel, otherwise it acquires weight 

 ■during the operation, by the formation of drops of a solu- 

 tion of potass on its surface. An error may arise in the 

 third place from the different effect of the solvent. If the 

 amalgam was oxidated in pure water, the hydrogen escaped 

 without the slightest smell, even when the oxidation pro- 

 ceeded pretty rapidly : but when I added muriatic acid, the 

 hydrogen acquired a strong smell, resembling that which 

 is perceived during the solution of zinc in this acid. Con- 

 sequently the gas must have held something in solution ; 

 and this could be nothing else than potassium : hence the 

 experiments, in which the muriatic acid was used for the 

 tioUition of the potassium, always gave a smaller result 

 than the rest. The same circumstance occurs when an 

 amalgam of the base of one of the earths is dissolved in 

 dilulcd muriatic acid; even if instead of the acid we only 

 add sal ammoniac j while in this case such an addition is 

 the more necessary, in order to obtain the earth in a state 

 of solution. 



The numerous experiments which I have made, respect- 

 ing the component parts of potass, gave at first the propor- 

 tion of oxygen varying from 16 to 20 per cent. I shall 

 here only adduce those which were conducted with the 

 gri-aicst care, and of which the results agree tolerably well 

 with each other. They indicate a greater proportion of 

 C c 4 oxygen 



