11 & • Gn some new analytical Researches. 



the same as that procured from it by electcicily ; and the two 

 series of facts seem I'uUy to establish the decompositioa 

 and reconiposition of the acid. 



From the large quamity of potassium required to decom- 

 pose a small quantity of the acid, it is evident that the 

 borAcic acid must contain a considerable prpporlion of oxy- 

 gen. I have endeavoured to determine the relative weights 

 of the peculiar inflammable mailer and oxvgen, which com- 

 pose a given weight of boracic acid ; and to this end 1 made 

 several analytical and synthetical experiments; I shall give 

 the results of the two which I consider most accurate. 



Twenty grains of boracic acid and thirty grains of potas- 

 sium were made to act upon each other by heat in a tube 

 of brass } the result did not effervesce when washed with di- 

 luted muriatic acid ; and there were obtained after the pro- 

 cess, by slight lixiviation in warn) water, two grains and 

 about -^g-t lis of the olive-coloured matter. Now thirty grains 

 of potassium would require about five grains of oxygen, to 

 form thirty-five of potash ; and according to this estima- 

 tion, boracic acid must consist of about one of the peculiar 

 inflammable substance, to nearly two of oxyeen. 



A grain of the inflammable substance in very fine powder, 

 and diffused over a large surface, was set fire to in a retort, 

 containing twelve cubical inches of oxygen ; three cubical 

 inches of gas were absorbed, and the black residuum coU 

 lected after the boracic acid had been dissolved, was found 

 to equal five-eighths of a grain. This, by a second combus- 

 tion, was almost entirely converted into boracic acid, with 

 the absorption of two cubical inches and one-eighth more of 

 oxygen. The thermometer in this experiment was at 58*^ 

 Fahrenheit, and the barometer at 30" 2, 



According to this result, boracic acid would consist of 

 * one of the inflamriiable matter, to about rs of oxygen j 

 and the dark residual substance, supposing it to be simply 

 the inflammable matter combined with less oxygen than is 

 sufiicient to constitute boracic acid, would be an oxide, 

 consisting of about 47 of inflanunable matter, to V5b of 

 oxygen. 



These estimations I do not however venture to give as 

 entirely correct, in the analytical experiments, there are 

 probably sources of error, from the solution of a part of the 

 inflammable matter, and it possibly may retain alkali, which, 

 cannot be separated by the acid. In the synthetical process, 

 in which washing is employed, and so small a quantity of 

 matter used, the results are still less to be depended upon; 

 |hey must be considered only as imperfect approximations, 



Froav 



