METHODS OF CLASSIFICATION 47 



part from the alteration of the non-crystalline materials, and in part from 

 the fixct of the more or less denudation which they have suffered. When 

 buried enough to allow of a more or less slow solidification, the tendency is 

 to form a completely crystalline structure, approaching more and more to 

 the granitic. While the chief portion of the granitic structure,' like that 

 seen in gabbros, some diabases and diorites, true granites and syenites, is 

 indigenous, all does not seem to be so, and great depth does not appear to 

 be indispensable ; the only requirement seems to be slow solidification. 



Section VII. — Ifdhods of Clamfication. 



The framework of any descriptive or systematic science is its classifica- 

 tion, and upon it depends much of the value and suggestiveness of the work. 

 It hence becomes a most important and vital point that the classification 

 used shall be as correct as possible. The common classifications of rocks are 

 well ]\nown to be artificial, and the writer has found them unsatisfxctory in 

 his work. Instead, therefore, of endeavoring to invent a new one, he has 

 striven to discover the laws and principles of the natural system, so far as 

 the rocks studied might enable him to do so. 



In studying rocks by any system, two methods are open to the observer. 

 One is to simply describe tlie characters of the minerals in the rock, thus 

 making the minerals the principal object, and the rock the subordinate one. 

 In this case lithology becomes simply a mineralogical study, and the litholo- 

 gist a mineralogist, Avho looks upon his rocks as small mineral cabinets_, not 

 realizing that the minerals are for the most part changing and changeable, 

 and that the true method is to trace the history and variations of the I'ock 

 as manifested in its mass and its constituents. 



The other method is to study the rocks for the purpose of determining 

 their natural relations, the various changes they have luidergone, and the 

 characters by which they may be known in all these various alterations. 

 In this the rock is the unit, the paramount object, and the mineral the 

 subordinate quantity. From this point of view the minerals in a rock 

 answer to the teeth and bones in an animal — very important, but not 

 superior in value to the animal as a whole. In a rock the mineral is the 

 accident, it may or may not exist ; and when the rock is entirely composed 

 of crystallized minerals, they should be used as the teeth and bones are used 

 in determination, when the zoologist has them alone in his specimen — in 



