On Tribasic Phosphoric Add. L95 



I. Method with MO,. 



The analyses, according to this method, were made in ex- 

 actly the same way as previously described; the acid fluid, 

 however, that was filtered off from the precipitate containing 

 the phosphoric acid, was super-saturated with ammonia, and a 

 current of sulphydric acid passed through it. The solution, of 

 a dark-red color, was washed out from the residue; the latter 

 dissolved in hydrochloric acid, and treated once more in the 

 same way, before finally dissolving in IIC1, precipitating by 

 NH 3 , washing, drying, and weighing. 



The results in this case, were as follows : 

 (a) (A) 0.698 (B) 0.7354 (C) 0.6375 



PCX 32.31 32.39 32.36 



A1,0 3 19.19 19.17 L9.18 



HO 48.50 48.44 t8.46 



mean P0 5 32.35 



Al a O s 19.18 



110 48.47 



II. Method with Si0. 2 (Berzelius' by fusing). 



I have to remark on this method and on the following one, 

 that a small quantity of alumina will get into solution, depend- 

 ing probably on the amount of silica used, for more or less of it 

 will change the character of the otherwise insoluble residue. 

 About 1 gr. of the substance was fused over a single Bunsen 

 burner, with 10 gr. of the above mentioned mixture containing 

 the silica. The "mass looked crystalline and separated easily 

 from the platinum crucible ; after digestion with hoi water,the 

 soluble part was washed out and separated from silica by twice 

 evaporating to dryness. The filtrate was then supersaturated 

 with carbonate of soda, evaporated to dryness, and heated for 

 a few minutes to fusion. The fused mass dissolved clearly in 

 water. After acidulating with 11C1 the solution was boiled, 



