﻿Microscopic Constitution of certain Cornish Rocks. 105 



present in these rocks was soluble when the finely divided sub- 

 stance was digested with nitric acid, the following process was 

 adopted for its estimation. 



About 100 grains of the finely pulverized rock was attacked 

 by nitric acid in a platinum dish, and evaporated nearly to dry- 

 ness to remove the excess of acid. The insoluble matters were 

 removed by filtration, and the filtrate reduced to a very small 

 bulk by evaporation ; molybdate of ammonium was then added 

 and the beaker, carefully covered, set aside in a warm place 

 for twenty-four hours to allow of the formation of a precipi- 

 tate. 



The precipitate (when one was obtained) was thrown on a filter 

 and washed with cold water, to which had been added molybdate 

 of ammonium and a little nitric acid. 



Strong ammonia was now poured on the filter, and the filtrate 

 evaporated to dryness in order to render insoluble any silica 

 which, at the same time as the yellow precipitate, might have 

 been dissolved by ammonia. The residue was then taken up by 

 hydrochloric acid, insoluble matter separated by filtration, and 

 ammonia and magnesium mixture added. After being set aside 

 for twelve hours to allow of the formation of the magnesium-salt, 

 the clear liquor was drawn off by a pipette, the precipitate re-dis- 

 solved by a few drops of hydrochloric acid, and re-precipitated 

 by the addition of ammonia. This precipitate was washed, 

 dried, ignited, and weighed as pyrophosphate of magnesium, 

 from the weight of which the amount of phosphoric acid was cal- 

 culated. 



When the rock under examination was found to contain an 

 appreciable amount of manganese or (as in the case of the ser- 

 pentine) oxide of nickel, the general routine of analysis required 

 the following modification. 



The filtrate from silica was neutralized by ammonia, chloride 

 of ammonium added, and finally sulphide of ammonium. The 

 precipitate thus produced was dissolved in hydrochloric acid to 

 which a few drops of nitric acid had been added, evaporated to 

 dryness for the purpose of separating traces of silica, and re-solu- 

 tion effected in hydrochloric acid. This, after filtration, was 

 treated by carbonate of barium, by which oxide of iron and alu- 

 mina were precipitated* ; these were subsequently estimated in 

 the usual way. 



* In order to remove any sulphuric acid resulting from the action of ni- 

 trohydrochloric acid on the sulphides, a few drops of chloride of barium 

 were added to the filtrate before the addition of carbonate of barium. In 

 the event of a trace of alumina having escaped precipitation by carbonate 

 of barium, it is readily separated from oxide of manganese or oxide of 

 nickel by digestion in an excess of hydrate of potassium or hydrate of 

 sodium. 



