and naphtfhaline, und on a new acid produced. 157 



baryta was then added, the whole evaporated to dryness, 

 heated red-hot, washed with dilute muriatic acid to remove 

 the baryta uncombined with sulphuric acid, and the sulphate 

 collected, dried, and weighed. The results were inconstant; 

 but the sulphate of baryta obtained, always much surpassed 

 that furnished by the former method. Judging from this 

 circumstance that the jsulphuric acid in the salt was more 

 than an equivalent for the baryta present, many processes 

 were devised for the determination of its quantity, but were 

 rejected in consequence of difficulties and imperfections, [aris- 

 ing, principally, from the presence and action of so much 

 carbonaceous matter. The following was ultimately adopted. 

 A quantity of per-oxide of copper was prepared by heating 

 copper plates in air and scaHng them. A sufficient quantity 

 of pure muriatic and nitric acids were provided, and also a 

 specimen of pure native carbonate of baryta. Seven grains 

 of the salt to be examined were then mixed with seven grains 

 of the pulverized carbonate of baryta, and afterwards witii 

 312 grains of the oxide of copper. The mixture being, put 

 into a glass tube was successively heated throughout its 

 Onass, the gas liberated being passed through a mixture of 

 baryta water and solution of muriate of baryta. It was found 

 that no sulphurous or sulphuric acids came off', or indeed 

 sulphur in any state. The contents of the tube were then 

 dissolved in an excess of nitric and muriatic acids, above that 

 required to take up all that was soluble ; and a little solution 

 of muriate of baryta was added for the sake of greater 

 certainty. A portion of sulphate of baryta remained undis- 

 solved, equivalent to the sulphuric acid of the salt experi- 

 mented upon, with that containedaccidentally in the oxide of 



