84 WHITMAN CROSS 



rights. Rather, it must be said that he does not recognize his 

 rights. He does not perceive the true relationship between 

 geology — petrology — and petrography. Nor does the petrog- 

 rapher who accepts that proposition recognize his rights. 



Defining^ petrology and petrography, as has been done, 

 bearing in mind the complex and variable character of the rock 

 and its manifold relationships, it seems to me that the petrogra- 

 pher should esteem it his duty to produce a systematic classifi- 

 cation of rocks with a consistent nomenclature, which shall first 

 of all possess stability. The nearer it approaches to a natural 

 system the better, but the character of the rock precludes the 

 hope of securing a fully natural system. The right of the 

 petrographer under this principle is that he may apply the test 

 of adaptability to each criterion offered. It may be said by 

 some that the ultimate object of petrography must be to secure a 

 thoroughly natural classification, and that when knowledge of 

 the rock is extensive enough such a system will be possible. 

 I believe that that position is incorrect, if, by a natural classifi- 

 cation, is meant one expressing all the relationships of rocks. 

 It is not because of ignorance that we cannot set up such a nat- 

 ural system for rocks. The nature of the rock is the cause of 

 this inability, not ignorance concerning it. 



The petrologist must classify rocks from every standpoint. 

 He must apply many material facts, all of which cannot possibly 

 be used in the systematic classification of petrography, so many 

 sided is the rock. To illustrate this point, a sandstone is a rock 

 which may be described as inorganic, derived, compound, clastic, 

 stratified, sedimentary, aqueous, surficial, noncombustible, etc., 

 and each of these terms expresses a criterion that has been used 

 in some proposed systematic classification. The petrologist 

 must also classify some rocks on bases of theory or hypothesis, 

 with an expressive nomenclature. For the good of his science 

 he should be able to change such classification and dependent 

 nomenclature as required by advancing knowledge. This 

 amounts to a revolution if the general classification must also be 

 revised in each case. Is it not then a logical principle, for the 



