METHODS OF PETROGRAPHICAL RESEARCH. 211 



constituents of a crushed rock may be partially separated, 

 magnetite, ilmenite, augite, hornblende, olivine, etc., being 

 picked out successively ; but this refinement does not seem 

 to have come into general use. 



Cordier's method of levigation in water is too tedious 

 and too imperfect to be of much use in separating such 

 minerals as felspars and pyroxenes by their not widely 

 different specific gravities. For isolating the microscopic 

 crystals of exceptionally dense minerals (zircon, rutile, 

 anatase, monazite, etc.) from crushed or decomposing 

 crystalline rocks, however, it has been used with success 

 by some workers, e.g., by Thiirach, who has given an 

 account of the process adopted by him (3). The method 

 was originally borrowed from the miners, and Derby (4) 

 and other workers have made use of the "batea" or 

 washing-pan of the Brazilian diamond-seekers. In this 

 country Dick (5) and others have used a similar device 

 for separating the denser minerals from sands. 



Notwithstanding these special applications, it is manifest 

 that the separation of minerals by their different densities 

 can be in general but very imperfectly effected by means 

 of a liquid like water, much less dense than any of the 

 minerals. To surmount this difficulty recourse has been 

 had to special "heavy solutions," by which the majority of 

 minerals met with can be separated hydrostatically. The 

 two most familiar to petrologists are Thoulet's (1878), 

 which is potassium-mercuric iodide, and Klein's (1881), 

 cadmium boro-tungstate. Both can be prepared of specific 

 gravities exceeding those of most rock-forming minerals, 

 and can be diluted down as required, so that the floating 

 mineral - grains sink successively in the order of their 

 densities, and can be removed in turn. It is unnecessary 

 to give in this place the methods of using these dense 

 solutions, to describe the special apparatus for facilitating 

 their use, or to mention other liquids which have been 

 employed, some of superior density to the above, but with 

 practical drawbacks. Good accounts have been given by 

 Cohen (6) and others, and are summarised in the text-books 

 of Rosenbusch and Zirkel. 



