Intelligence and Miscellaneous Articles. 157 



in which it occurs cannot be worked any further until a tunnel 

 has been run, and that it is quite uncertain when this will be done. 



Although by no means perfect, my results approach the truth 

 and give a fair idea of the composition of the mineral, even if the 

 evident admixture of other minerals, varying in the different sam- 

 ples analyzed from about one to perhaps over twelve per cent., does 

 not permit one to calculate the atomic ratio of the constituents and 

 establish the constitution of this species. There is especially an 

 uncertainty with reference to the quantities of silicic acid, alumina, 

 and potassa which belong to the Roscoelite, or which may have been 

 introduced by admixtures of feldspathic and other minerals, as will 

 appear from the results given below, which show that the mineral, 

 when decomposed with sulphuric or dilute hydrofluoric acid, gene- 

 rally gives only about six per cent, of potassa, while fusion with 

 calcic carbonate and amnionic chloride yields from eight to nine per 

 cent. Some of these uncertainties could have been removed if a 

 larger quantity of the mineral had been at my disposal. 



Particular attention was paid to the correct determination of the 

 vanadium, and the form in which it exists in the Roscoelite. 



The separation of vanadium is attended with great difficulties ; 

 and I have not found any of the methods of separation to give 

 fully reliable results. This is in part owing to the incomplete pre- 

 cipitation of the vanadic acid, and in part to the impossibility of 

 washing the precipitates completely without loss of vanadium. It 

 was therefore always determined by the only method which I found 

 to give fully reliable results — by titration with potassic permanganate. 



After the separation from the other elements, the vanadic acid 

 was reduced by hydros ulphuric acid into V 2 4 , which, after the ex- 

 cess of hydros ulphuric acid had been expelled by continued boiling, 

 was reoxidized into Y 2 6. by permanganate. I have satisfied my- 

 self by numerous experiments that, no matter whether only a very 

 minute quantity of sulphuric acid is present, or a very large excess, 

 the V 2 4 is completely oxidized into V 2 O. by this process. 



For the determination of the state of oxidation of the vanadium 

 in the Roscoelite, a quantity of the mineral was dissolved in dilute 

 sulphuric acid in a sealed tube at a temperature of about 180° C, 

 and was titred after cooling; the liquid was then reduced by 

 hydrosulphuric acid ; and after boiling-off the excess of the latter, it 

 was again titred. From the quantity of oxgyen required for oxi- 

 dation in both cases, it was found that vanadium in the mineral is 

 present as V 6 O u =2V 2 3 , V 2 5 . 



The determinations of the other elements were made by the usual 

 methods. 



The finely powdered mineral was dried (unless otherwise stated) 

 for two days over sulphuric acid ; and the different samples gave the 

 following results : — 



a (purest scales). The analysis was made by dissolving one 

 portion in sulphuric acid and determining in this the quantity and 

 state of oxidation of the vanadium, the silicic acid, and insoluble 

 impurities. The latter were left behind in dissolving the silicic 

 acid in sodic carbonate and gave 085 per cent. ; a second portion 



