W. F. Tlillebrand — Danger in Mineral Separations. 439 



Art. XXXVI. — A Danger to he Guarded Against in Making 

 Mineral Separations oy Means of Heavy Solutions; by 



W. F. HlLLEBKAND. 



That one or another of the heavy liquids and solutions used 

 for making mineral separations acts chemically on a few min- 

 erals has long been known. Thus, the solution of potassium 

 and mercuric iodides, known as Sonstadt's or Thoulet's solu- 

 tion, attacks metals and probably certain oxides and sulphides. 

 Eohrbach's solution of barium and mercuric iodides probably 

 has similar action. Klein's solution of cadmium-boro-tungstate 

 decomposes carbonates as well as attacks metals. Doubtless 

 others have their specific effects. 



So far as known, the actions hitherto observed have resulted 

 beyond doubt in complete solution or profound disruption of 

 the mineral subject to attack, but such action, being easily 

 noticed, involves no serious risk. I have now to record an 

 instance in which the attack involves a chemical interchange 

 between heavy solution and mineral without producing any 

 alteration in the latter, that is visible to the eye. 



There was sent me for examination, some two years ago, by 

 the Foote Mineral Company, of Philadelphia, a canary-yellow 

 crystallized mineral from Paradox Yalley, Montrose County, 

 Colorado, which was supposed to be carnotite, this being essen- 

 tially hydrous potassium- uranium vanadate. This mineral 

 proved to be very similar to carnotite in composition, but with 

 calcium replacing potassium. Later, the attempt was made to 

 purify some of it by the potassium -mercuric iodide solution, 

 but analysis then gave for it the composition of carnotite. 

 Suspecting the reason, another specimen was divided into two 

 portions, one of which was analyzed in its original state, the 

 other after action upon it of the heavy solution for 18 hours 

 and complete removal of the mercury salt. This was found to 

 have lost most of its calcium and to have taken up the exact 

 equivalent of potassium. The composition of the two portions 

 with respect to the major constituents is shown below, a num- 

 ber of minor elements being disregarded, also the quartzy 

 gangue. 



It should be remarked that the material analyzed in this case 

 was a mixture, the copper belonging in great part if not alto- 

 gether to a greenish vanadate, to which probably a portion of 

 the calcium also belongs. 



The only essential change in the composition of the mineral 

 appears to be with respect to the calcium and water. The 

 molecular ratio of CaO+K 2 in the two portions is 954 to 952, 

 indicating exactly equivalent replacement of calcium by potas- 



