188 HALL AND SMITH— COLUMBIUM. [May 19, 



Calculated Found. 

 KjCbOFsHjO. 



Oxide 44.52 44.38 



K 2 SQ 4 57.81 58.13 



The sp. gr. of the oxide was found to be 4.481. 0.3160 gram of 

 oxide gave a color with hydrogen peroxide equivalent to .0022 

 gram Ti0 2 = . 7 per cent. 



Crystals "B" (page 181) analyzed as follows: 



0.8098 gram of sample gave 0.3614 gram of oxide and 

 0.4708 gram of K 2 S0 4 

 0.3614 : 0.4708 :: x/2 : 174 x = 267.0. 



Calculated Found. 

 KXbOF 6 H,0. 



Oxide 44-52 44-63 



K 2 SO, 57-Si 58-14 



The sp. gr. of the oxide was found to be 4.864. 0.3600 gram of 

 oxide gave with hydrogen peroxide a color equivalent to .0013. 

 gram of Ti0 2 = .36 per cent. 



Having failed to get any evidence of the existence of neptunium 

 in the last fractions of the double fluoride from the South Dakota 

 mineral, it was decided to test some of the mineral from Haddam, 

 Conn., the source of the material used by Hermann in his investi- 

 gation. The last fractions of the potassium double fluoride from 

 5.87 kilos of columbite from Haddam, Conn., amounting to 100 

 grams, were dissolved in boiling water and an excess of sodium hy- 

 droxide added. The precipitate obtained was partly crystalline 

 and partly flocculent, as described by Hermann. It was filtered 

 out, dried on a porous plate, and boiled with 25 parts of water. 

 The crystals dissolved leaving a yellowish residue evidently con- 

 taining much iron. This residue gave a yellow colored bead in 

 the reducing flame containing so much iron that it was impossible 

 with a small blowpipe to keep it all reduced. Is it not probable 

 that this is what Hermann supposed was neptunium ? 



This salt was fused with acid potassium sulphate to remove the 

 iron, the oxide remaining after extracting the fusion with boiling 

 water was changed to double fluoride, dissolved in boiling water, 

 and an excess of sodium hydroxide added. The precipitate ob- 

 tained was crystalline and perfectly soluble in water, leaving an 

 inappreciable residue. From all of which it may be inferred that 



