126 MR. F. C. CALVERT AND MR. R. JOHNSON ON THE 



quantities is most injurious, the more so as such minute quan- 

 tities as a few thousandths more or less will completely alter 

 the value of this metal for many uses. To determine phos- 

 phorus, the process which we followed was similar to that 

 employed for the sulphur, with this difference, that we added 

 to the liquid from which we had separated the silica, a 

 little hydrochloric acid, and then ammonia in excess, 

 instead of acetic acid, as we had done in the analysis for 

 sulphur. The liquid was allowed to stand to see if any 

 alumina separated, and if not, we added hydrochloric acid in 

 excess, then pure chloride of calcium and then ammonia 

 again, when phosphate of lime, having the following formula 

 P0 5 3CaO, was precipitated, from which the quantity of phos- 

 phorus was calculated. We always took care to operate on 

 such a bulk of fluid as to prevent the precipitation of any 

 sulphate of lime, and we also washed rapidly to prevent any 

 carbonate of lime being formed. We verified this method 

 several times during our analyses, by determining the amount 

 of lime in our precipitates and the amount of phosphoric acid 

 by M. Reynoso's process. 



Aluminum. 



If there was any alumina it was separated during the last 

 process, and its amount determined. 



We also made several fusions of iron dissolved in aqua regia, 

 and evaporated with a mixture of alkaline carbonates to which 

 we had added a little caustic alkali and we found no alumina 

 or mere traces in the iron analysed by us. 



Manganese. 



Five grammes of iron were dissolved in aqua regia, and the 

 whole evaporated to dryness and calcined with alkaline car- 

 bonates. The fused mass was treated with boiling water, and 

 to the solution were added small pieces of Swedish paper, to 

 reduce the manganese. 



