330 M. Vauquelin on d [Mav, 



212° Fahr. and weigh it again, or I bum it, and also the same 

 quantity of filtering paper, and keep the a&hes for a counterpoise 

 of those of the filter. 



As to the manganese which remains in the solution, I precipi- 

 tate it by subcarbonate of potash : it appears in the state of 

 white flocks, which soon diminish in volume and become gra- 

 nular. The vessel must be shaken for some time in order to 

 prevent the substance from adhering to it. I entreat chemists 

 to excuse me for entering into these minute details, but they 

 may not, perhaps, be useless to some persons. 



The quantity of the bases here obtained being too great for the 

 saturation of the tungstic acid, the following experiment was 

 made. 



. Five grammes of wolfram treated by potash gave 1 gramme 

 59 centigrammes of iron and manganese ; the analysis of these 

 r59 yielded 69 cent, of iron, 65 of manganese, and 19 of silica; 

 these quantities divided by 5, in order to convert them into hun- 

 dredths, give 13-8 of iron, 13 manganese, and 3-8 silica ; but on 

 examining the manganese, there were found 8 cent, of carbonate 

 of lime, which also divided by 5, give 16 milligrammes to be 

 deducted 5 the oxide of manganese is thus reduced to 11-4 

 grammes. 



As four analyses were performed which gave the same result 

 within about one-hundredth, and as the above experiment was 

 performed with the greatest precaution, there is ground for 

 hoping that it is as correct as can be expected by chemical 

 means. 



In adding together the different quantities of matter which 

 have been mentioned, we get 31*48 ; but of this quantity there 

 are 4'68 of silica and lime, which unquestionably form no essen- 

 tial part of the mineral : this reduces the iron and manganese 

 to 26"8 ; thus subtracting these 4*68 of silica and lime, it appears 

 that only 95*32 cent, of wolfram were operated upon, which 

 consequently raises the quantity of iron to 14*462, manganese to 

 11*949, and tungstic acid to 73*599 cent. On this occasion the 

 analysis of common wolfram was repeated, and the results were 

 the following ; 



Peroxide of iron 19*5 



Peroxide of manganese 5*4 



Silica 4*5 



Alumina 1*6 



31*0 



But subtracting from this amount 6*1 of silica and alumina, 

 which ought to be considered as foreign admixture, the iron will 

 become 20*745, the manganese 5*744, and the tungstic acid 

 consequently 73*511. It is then evident that this kind of 



