﻿Vol. 66.] CLASSIFICATION OF IGXEOUS ROCKS. 498 



occurrence of hundreds, if not thousands, of rocks of all kinds is 

 disregarded by their classification in Kosenbusch's work into deep- 

 seated or effusive rocks, to say nothing of the dyke rocks. 



The connexion of texture with geological occurrence in a 

 systematic way dates from days of ignorance concerning the origin 

 of texture, and of very limited knowledge of the rocks of the 

 hypabyssal zone of the earth, with their multiplicity of textural 

 forms. Once this systematic proposition seemed a close approxi- 

 mation to a statement of natural relations. Now we know that 

 it is not in accord with the nature of things ; its use to-day is 

 purely arbitrary (often a matter of personal convenience), and yet 

 it is commonly referred to as ' natural ' ! This course simply serves 

 to befog the discussion of truly natural classification. 1 



To classify igneous rocks as plutonic, intrusive and extrusive, or 

 by other equivalent terms, with the idea that because such a 

 scheme expresses certain geological relationships it must perforce 

 be accepted as ' a natural system,' may be compared with dividing 

 vertebrate animals into aquatic, terrestrial, and aerial, or those 

 that swim, walk, and fly, respectively. This division is natural in 

 its way, and agrees with zoological system in its general results. 

 The zoologist, however, does not call the whale a fish because it 

 lives in the water and swims, nor the bat a bird because it flies, 

 although it is more confined to this form of locomotion than is 

 any bird. 



VII. Quantitative Classification. 



Introductory. — If the considerations thus far presented are 

 well founded and sufficiently comprehensive, it appears that a 

 natural classification of igneous rocks, expressing a relation between 

 their most notable chemical or physical properties and the origin of 

 those properties in geological occurrence, is impossible. The natural 



1 As an instance of recent advocacy ot this systematic proposition, I would 

 cite a paper by Dr. F. H. Hatch. 'The Classification of the Plutonic Bocks ' 

 (Science Progress, Oct. 11)08). where the appropriateness of recognizing the 

 two great divisions of plutonic and volcanic rocks is stated without further 

 discussion in these terms : — s Apart from mode of occurrence, the difference 

 in internal structure or texture between a eugranitic plutonic rock, that has 

 cooled slowly as a deep-seated mass under pressure and in the presence of 

 occluded water, and aporphyritic or semi-vitreous volcanic rock, that has been 

 erupted at the surface and bas cooled quickly, is so great that a system can 

 scarcely be termed natural that places two such different rocks in the same 

 category. In a natural classification, therefore, the plutonic and the volcanic 

 rocks must be recognized as distinct classes.' (Sep. cop. p. 1.) 



The necessity for a distinction on this ground is less obvious when we con- 

 sider that coarsely crystalline or even granular basic rocks are known, which 

 have cooled slowly in a surface-flow, while there are equivalent porphyritic or 

 semivitreous rocks, which have evidently formed as a result of rapid cooling 

 in the hypabyssal zone or below. Classification is not for granitic rocks alone, 

 nor for plutonics, though many suggestions in classification have seemingly 

 been made without consideration of the needs of other rocks. 



