MINERALOGICAL CLASSIFICATION OF IGNEOUS ROCKS 97 



of Fig. 20 would then fall into the dotted compartments of 

 Fig. 21. Computed modes, however, would be more difficult to 

 place. As a matter of fact it is usually not difficult to separate 

 the albite in microperthite from the orthoclase. Should this change 

 be made, Family 6, for example, would become the family of the 

 alkali-granites, and would contain potash-granite, alkali-granite, 

 alkali-adamellite, alkali-granodiorite, and soda-tonalite. The latter 

 would then again become soda-granite, the first potash-granite, 

 and the intermediate rocks soda-potash granites. Covite, mariupo- 

 lite, most essexites, etc., would fall in Family 21 without differentia- 

 tion. Such a combination would reduce the number of families by 

 72, and if the anorthite were united in Class 3, as suggested above, 

 the total reduction would be 144 families. Personally the writer 

 is inclined to favor separating the feldspars into the Or, Ab, and An 

 molecules. 



5. Would it be desirable to indicate, in the name of the rock 

 itself, that the mineral proportions have been determined, and that 

 the rock falls into a certain compartment, for example by a slight 

 change in the spelling, such as granyt, dioryt, etc. ? Of course 

 terms like monzonite Brogger, theralite Rosenbusch, etc., 

 might be used, but they seem cumbersome. (Granyte, dioryte, 

 etc., cannot be used, since this spelling was suggested and used by 

 Dana to contrast with the -ite endings of minerals.) 



Appendix. — An alternative classification could be based upon 

 four double tetrahedrons, representing four classes, according to 

 the amounts of light and dark constituents, and each subdivided 

 as in Fig. 27. The corners of the tetrahedrons would be quartz, 

 Kf, Naf, Caf, and Foids, and the division points 0-25-75-100. 

 There would be fewer varieties than in the preceding classification, 

 and it would be much simpler, but the families would not corre- 

 spond so closely to those in the old classifications as does the one 

 given above. 



