42 CLASSIFICATION BASED ON MINERAL COMPOSITION. 



examining the crystals produced, under proper conditions. This is simply a 

 method of making qualitative tests upon material in bulk too small to be 

 tested in the ordinary way.* 



In 1876 Professor J. Szabo published f his method of determining feld- 

 spars by means of their fusion, reactions, and coloration produced in the 

 Bunsen flame, which gave, according to him, a means for estimating the 

 percentages of alkalies, etc., in the specimen examined. 



Still a third method is from the crystals formed in blowpipe beads under 

 proper conditions. This method was invented by Mr. George H. Emerson 

 in 1863, t and later expanded by Gustav Rose,§ W. A. Ros8,|| and H. C. 

 Sorby.^ The results are essentially similar to Boficky's method, qualita- 

 tive, but can be used with small fragments. 



The last and most important method is that of separating the feldspars 

 by means of liquids of different specific gravities. In this way considerable 

 material of a certain specific gravity can be obtained for chemical analysis, 

 and its nature ascertained.** All these methods have their defects : as, for 

 instance, the feldspars which give character to the rock are of more than 

 one species, usually; they contain more or less glass and mineral impurities; 

 and they are subject to alteration. These factors change their specific 

 gravity and chemical relations, and make the determination of a few crystals 

 of but limited value in fixinoi; the condition and character of the remainina; 

 feldspars. Of all the methods the specific gravity one promises the most, 

 but it is not believed at present to lead to any essentially valuable results 

 in determining minerals like the feldspars, whose very species are so inde- 

 terminate. While the before-mentioned methods, and many others not 

 mentioned, have added greatly to the knowledge of minerals, they seem to 

 have blinded most observers to the general characters of the rocks they were 

 studying. 



In the coarsely crystalline rocks crystals of feldspar, of sufficient size 

 for analysis, can often be obtained ; but that analysis, to be of any value, 

 must proceed on the supposition that the crystal is pure, unaltered, and 



* Arcliiv der naturwisseuschaftliclien Landesdurcliforscliung Bolinieus, 1877, iii- 5tli Abtli., pp. 1-SO. 



f Ucbcr cine ncue Metliode die Feldspatlie audi in Gesteincn zu bestiiumen, Budapest, 1S76. 



t Amer. Jour. Sci., 18G4(2), xxxvii. 414, 415 ; Proc. Am. Acad., 1SG5, vi. 47G-494. 



§ Monatsb. Berlin Akad., 18ry7, pp. 129-147. 



II Chemical News, 18G8 (Amer. Reprint), ii. 74-7G, 147, 148, 157-lGO, 196 ; Pyrology or Fire Chemis- 

 try, London, 1875. 



^ Month. Micro. Jour., 1SG9, i. 349-352. 



** Thoulet, Comptes Rendus, 1878, Ixxxvi. 454-456; Bull. Min. Soc France, 1879, p. 17 ; Cburcli, Miu. 

 Mag., 1877, i. 237, 238; Goldsclimidt, Neucs Jahr. Min., 1881 (Bcilage-Baud), pp. 179-238. 



