THE FELDSPAES AS A BASIS OF CLASSIFICATION, 43 



typical of the predominating feldspar in the rock ; microscopic analysis 

 shows that the larger crystals in our rocks are generally abnormal, often 

 foreign to their present surroundings, containing numerous inclusions, some- 

 times three fourths of the crystal being glass, microlites, etc. If a thin 

 section is prepared before the chemical analysis is made, it only proves 

 that the part examined is pure or impure, as the case may be, offering no 

 proof regarding the rest, only a probability ; further, the larger crystals are 

 usually the subordinate ones, being unlike the generality in the mass of the 

 rock. This method is, also, inapplicable in the cases where it is most needed ; 

 in the fine-grained and compact rocks, which contain few or no feldspars of 

 sufficient size. The larger feldspars are most subject to alteration, passing 

 from the basic towards the acidic,* some becoming greatly changed while 

 the smaller crystals are untouched ; yet the analyst names the rock from 

 the altered, and not from the unaltered feldspar, — euphotide, for instance.! 

 The secondary formation of feldspars, like orthoclase in rocks, adds 

 greatly to the difficulty of making the classification dependent upon the 

 kind of feldspar present. In other cases the feldspathic material is seen 

 to be largely replaced by quartz and other minerals — the presence of the 

 first not being suspected until the crystal was examined under tlie micro- 

 scope. The twinned character of the triclinic feldspars, seen both in com- 

 mon and polarized light, is not a constant character, as has been pointed 

 out before. It has been customary to regard all unstriated feldspars in 

 basic rocks as plagioclase, cut parallel to the brach\q:)inicoid, but in the 

 acidic rocks as ortlioclase. Through the great alteration to which the 

 feldspars have been subject in the older rocks, all signs of twinning have 

 been frequently obliterated, thereby causing such crystals in granitoid 

 rocks to be classed as orthoclase. t The chief value, therefore, of the 

 optical method for distinguishing the feldspars is apparently to deter- 

 mine the predominance of plagioclase, or of orthoclase ; while the chief 

 use of the present micro-mineralogical study of tlie feldspars in litliology 

 is the determination of the more or less acidic or basic composition of tlie 

 rocks, according to the predominance of orthoclase or plagioclase in them. 

 From the above it follows that a systematic classification cannot be properly 

 based on any such variable, indeterminate materials. 



* Geo. W. Hawes, Geology of New Hampshire, iii. part. iv. 90-92. 



t T. Sterry Hunt, Am. Jour. Sci. (2), 1859, xxvii. 336-3i9 ; J. D. Dana, Hid. (3), 187S, xvi. 34.0. 



X Bull. Mus. Comp. Zoul., ISSO, vii. 55, 56. 



