52 PROCEEDINGS OF THE AMERICAN ACADEMY 



we obtain a simple mixture of phosphate and metavanadate. Ammonic 

 chloride is then to be added, and the analysis conducted in the usual 

 manner^ This method permits the separate determination of each of 

 the pentoxides and so far offers an advantage. Oa the other hand, it 

 oul}"^ applies to salts with alkaline bases. It is far more convenient, 

 and at least equally accurate, to determine the vanadium by titrition 

 with hypermanganate under appropriate conditions. The end reaction 

 is usually well defined in a sufficiently dilute solution. The sum of 

 the vanadic and phosphoric oxides may be accurately determined by 

 precipitating the solution of the two pentoxides at a boiling heat by 

 means of mercurous nitrate, adding mercuric oxide until a small but 

 distinct excess is present so as to perfectly neutralize the free acid, and 

 then boiling for a few minutes. Since the ignited mass of oxides 

 retains mercuric oxide with some tenacity, it is best, after burning off 

 the filter with a very free access of air, to add a weighed portion of 

 sodic tungstate and ignite until a constant weight is obtained. When 

 the process is carefully conducted, the fused mass contains all the 

 phosphoric oxide and all the vanadium as pentoxide. Care should be 

 taken to oxidize the whole of the vanadium to pentoxide as more 

 certain. Any small quantity of dioxide present is however precipi- 

 tated when an excess of mercuric oxide is present. This method 

 applies to all salts or compounds containing oxides of phosphorus 

 and vanadium. As regards the details of the determination of vana- 

 dium by titrition with potassic hypermanganate the following points 

 require attention. 



For the reduction of vanadic pentoxide to dioxide, it is best, after 

 adding dilute sulphuric acid, to boil with a solution of sulphurous acid 

 and afterward to evaporate the solution nearly, at least, to dryness 

 upon a water bath, so as to insure a perfect reduction and at the same 

 time to expel the last traces of sulphurous oxide. Sulphydric acid gas 

 also readily reduces the vanadic pentoxide in presence of sulphuric 

 acid, but there is almost always a separation of free sulphur in a state 

 of extreme sitbdivision. This passes readily through the filter, and the 

 results of the titrition are not so satisfactory as when sulphurous acid 

 is employed as the reducing agent. The solution of vanadic dioxide 

 should be diluted till the color is pale blue, heated to about 80°-100°, 

 and titrated hot. In the hot solution the disappearance of the hyper- 

 manganate takes place instantly, and the end reaction is much more 

 quickly and satisfactorily attained. The actual results are precisely 

 the same as when a cold solution is employed, as the following titritions 

 sufficiently prove. 



