ERUPTIVE ROCKS. 593 



have appeared in the amiuouium tartrate sohitiou above meutioued. As a trace of. 

 lead sulpliate might have been formed by the ijinitiou of the- precipitate by hydrogen 

 sulphide and have escaped solution in the nitric acid added, the residue on the filter 

 was exhausted with ammonium tartrate and tested with ammonium sulphide. The 

 contents of the watch-glass were then evaporated with two or three drops of sulphuric 

 acid, and finally gently heated to expel nitric acid. If lead was ])resent it could now 

 invariably be seen at the center as white powder. This was collected on the smallest 

 possible filter, washed with alcohol, dried, ignited, and weighed as sulphate. The 

 latter was then scraped as far as possible on charcoal, and carefully reduced with a 

 very little soda. The yellow coating of lead oxide was invariably formed, and in the 

 soda appeared minute metallic buttons, malleable and soluble in nitric acid. The 

 solution, concentrated to a drop or two, showed a bluish-black precipitate with hydro 

 gen sulphide. 



The portion of the rock insoluble in nitric or nitrohydrochloric acids, composed 

 of quartz and silicates, was decomposed with hydrofluoric and sulphuric acids purified 

 by distillation from a platinum retort, and dissolved in slightly acidified water, after 

 expulsion of the hydrofluoric and excess of sulphuric acids. Solution and possible 

 residue were then treated as in the foregoing for the separation and estimation of lead. 



For the estimation of zinc and cobalt, large quantities (30 grains) were taken 

 and decomposition was effected by hydrofluoric acid. After evaporating with sulphuric 

 acid and igniting to expel the excess of the latter, solution was efl'ected in hot water 

 slightly acidified; the solution saturated with hydrogen-surphide; from the filtrate, 

 after oxidation, alumina and iron thrown down by ammonia; the precipitate redis- 

 solved in hydrochloiic acid after filtration, and reprecipitated. This being repeated 

 once more, the combined filtrates were evaporated to a moderate volume ; the alumina 

 still in solution was thrown down while boiling by ammonia, and this precipitate redis- 

 solved and reprecipitated. To the again combined filtrates ammonium sulphide was 

 added to throw down zinc, manganese, cobalt, and nickel, if present; the precipitate 

 was treated on the filter with a mixture of one part hydrochloric acid of 1.12 sp. gr. 

 and six parts solution of hydrogen sulphide. The zinc and manganese in solution 

 were thrown down again by ammonium sulphide, the manganese (being present in very 

 small quantity) extracted by dilute acetic acid, while the zinc sulphide on the filter 

 was then brought into a weighed platinum crucible by means of hydrochloric acid, 

 evaporated to dryness, and ignited with mercuric oxide iu the manner recommeuded 

 by Volhard. The oxide, after weighing, gave the characteristic green coloration on 

 igniting with cobalt nitrate. The cobalt sulphide left on the filter after extraction 

 of zinc and manganese was ignited with the filter, digested with nitrohydrochloric 

 acid; the solution rendered alkaline with ammonia; ammonium carbouate added; the 

 slight precipitate separated by filtration and the cobalt thrown down by potassium 

 hydrate. The ignited oxide tested by the method of Jorissen showed no trace of nickel. 



MON XII 38 



