﻿472 DR. WHITMAN CROSS ON THE NATURAL [A-Ug. I9IO, 



or other systematic unit, means a precise statement of the qualities 

 characterizing it. Petrographic nomenclature must be based ou a 

 universal scientific system of classification in order that general 

 discussions concerning rocks may be fruitful, as well as for the 

 sake of convenience and clearness in description. There are 

 reasonable grounds for the suspicion that many petrographers of 

 to-day have a very hazy idea of what they mean in using most 

 rock-names. The reader of petrographic literature is often quite 

 uncertain as to the real character of rocks named, but not fully 

 described. 



All petrographers should realize that when they advocate ' a 

 natural classification ' of igneous rocks, or refer to existing usage 

 as ' natural/ in any particular, they must mean according to 

 the nature of those objects. Yet the relations assumed to 

 be natural are actually, in most cases, merely empirical formula? 

 long since shown to have limited application. Only generalizations 

 without known exceptions in experience can be applied to the 

 construction of a system that may be called natural. Let us have 

 regard for things as they are, not assume them to be as we wish 

 they were, or as it would be convenient to have them. 



A classification of natural objects worthy of acceptance as the 

 -systematic scheme for a branch of natural science must be logical 

 and consistent in construction. As numerous logicians and philo- 

 sophers have pointed out, a systematic classification must be con- 

 structed by the use of factors vitally related to the differing 

 properties of the objects to be classified, and their application 

 should follow the order of their importance. A scientific specialist 

 should not be satisfied with a system which is such only in the 

 mechanical plan of its construction, or with one which does not apply 

 'to all objects of the class iu question. 



With these principles in mind it appears to me that most classifi- 

 cations of igneous rocks must be called unnatural and illogical 

 or unscientific in construction. These defects arise from two 

 principal causes. In the first place, what has been found 

 or believed to be true of the rocks of a petrographic 

 province, or of a certain group, has often been too 

 quickly assumed to be true of all igneous rocks, and 

 an erroneous factor has been thereb)' introduced into 

 the system. The age distinction, for example, seemed natural 

 to Continental petrographers, while British geologists knew it to 

 be unnatural, in the light of their observations. Such assumptions 

 are perhaps unavoidable in the early history of a science, but the 

 time must come when they should be rigorously tested in their 

 application to the rocks of the whole earth. In the second place, 

 it has been the practice of many of the most eminent 

 petrographers to assume as true certain relations of 

 rocks which were known to be but approximately in 

 accordance with observed facts. The assumptions that 

 texture is a function of depth, or of pressure, or of a eutectic mixture, 

 illustrate this point. It is usually, in such cases, the convenience 



