Pyroxenes. 17 



^obtained and an abundance of a solvent agent, it appears impossible 

 to explain why the smallest crystals always have the most nearly perfect 

 form, and that, even, when a large crystal and a small one lie close 

 to each other in the gangue. If a large crystal of calcite be dropped 

 in an acid solution together with a smaller one, it will invariably follow 

 that the smaller will disappear first, and that it will wholly lose its 

 external form before it so disappears. The same holds true of all 

 easily dissolved minerals, and it appears safe to assume it true of min- 

 erals like quartz, pyroxene, apatite, etc., which are more refractory. 



Again, the abundance of free silica present would render the ac- 

 counting for the silica removed by solution a task by no means easy. 

 Not only does the " country rock," which in this case is gneiss, con- 

 tain much quartz, but the walls of the veins are in many cases com- 

 pletely covered with quartz crystals, "fused," and with their apices 

 pointing towards the center of the vein. It would seem as if, had 

 the solvent action been present, the silica would have been carried 

 into the vein, not out of it, especially when such an abundance of 

 bases existed in the form of lime. 



It seems as if minerals could readily be divided into three classes. 

 First, those formed by volatilization ; second, those formed from 

 solution ; third, those formed by segregation in beds or veins while 

 undergoing metamorphism. As examples of the first class crystal* 

 of sulphur formed in volcanoes, the different chlorides, etc., found 

 under the same conditions, may be given, to which may, in all 

 probability, be added the diamond. Minerals of the second class may 

 be recognized by their fluid inclusions, such as quartz, and many may 

 be formed artificially. Of the third class, the mineral constituents of 

 rocks, such as granites, gneisses, diorites, may be given. Intrusive 

 veins, dykes, and veins of segregation, whether metalliferous or not, 

 would also come under this third division. 



Among rocks that are wholly crystalline, it is impossible that their 

 mineral constituents should be deposited from solution in the sense 

 in which the word is usually employed. Each mineral, if indeed any 

 individuals existed in the beginning, would be in a semi-fluid or pasty 

 condition. As time went on each would separate more or less per- 

 fectly from the mass, and as nearly as possible each would assume its 

 peculiar form. With large rock masses, however, which Eosen- 

 busch designates as " hypidiomorphic-granular," individualization is 

 rendered impossible from lack of space, and from the fact that the 

 factors of solution are nearly the same in the case of each. In fact 



