4:00 Foote and Langley — Method for Determining^ etc. 



The density method was also applied to the partial analysis 

 of a Branchville cohimbite and for comparison tantalum and 

 colnmbium were separated and determined by the Marignac 

 method. The specimen contained no titanium. The oxides 

 obtained in the course of analysis were sohible in hydrofluoric 

 acid, so tliat preliminary fusion with acid fluoride was unneces- 

 sary. The solution was precipitated with ammonia, Altered, 

 redissolved in hydrofluoric acid and evaporated with sulphuric 

 acid, etc., as previously described. The results of the duplicate 

 analysis are given in the following table: 



Per cent 

 Ta306 + 

 CbjOs in 

 mineral 



1. 'ZS-T? 



2. 79-04 



Density 



of mixed 



oxides 



4-957 

 4-918 



Per cent 



TaaOr, in 



oxides by 



density 



method 



21-0 

 20-0 



Per cent 

 Ta,Or, in 

 oxides by 

 Marignac 



method 



18-74 

 17-21 



Per cent 

 oxides in 

 mineral by 

 density 

 method 



Per cent 



oxides in 



mineral by 



Marignac 



method i 



TaaOs CbuOs Ta^O^ CbaOe 



16-54 62-23 

 15-81 63-23 



14-43 63-68 

 13-60 65-75 



It will be seen that the results are approximately alike by the 

 two methods, but we believe the results by the density method 

 are more accurate and they are certainly much easier to obtain. 

 The specific gravity of the columbic oxide from one Marig- 

 nac separation was found to be 4-612 corresponding to 96 per 

 cent of columbic oxide and 4 per cent tantalic oxide, showing 

 that some tantalum went with the columbium. It is probable 

 also that some columbium crystallized with the tantalum double 

 fluoride but the amount of tantalic oxide was too small for a 

 density determination. 



The authors wish to express their thanks to Dr. W. E. Ford 

 and Dr. T. B. Osborne for the tantalum and columbium oxides 

 used in this investigation. 



Sheffield Chemical Laboratory, 



New Haven, Connecticut, 



July, 1910. 



