CONCLUSIONS 801 



ence (by feeding) and preservation of the race (through procreation) to 

 the conditions of the environment. In rocks, on the contrary, we deal, 

 not with sharply discrete individuals and groups of individuals, but with 

 indefinite mixtures, which grade or shade into each other in every (chem- 

 ical, mineral, and textural) direction. At the best, we may conceive of 

 and concentrate our attention on certain somewhat vague and ill-defined 

 centers around which the igneous rocks cluster, much as a nebula, spiral 

 or otherwise, may cluster about a nucleus, the outer limits of both the 

 nucleus and the nebula being hazy and strictly indeterminate. There is 

 no individual rock species, as there is an animal or vegetal species, inde- 

 terminate, transitional, and transitory as such an organic species may be. 



It would follow from the considerations just expressed that there can 

 probably be no exact definitely determinable or definable classification of 

 igneous rocks, such as we find in zoology and botany. We can classify 

 rocks, for petrological purposes, exactly, definitely, and strictly only by 

 creating arbitrary divisions, cutting them up by sharp planes and putting 

 them into man-devised pigeon-holes, as was done in the quantitative 

 classification or as seems to be necessary in any modification of it. Such 

 a classification is a pis-aller, a makeshift, a classification of convenience ; 

 it may or may not correspond to the evolution of igneous rocks as it 

 really is. That is for the future to determine. Personally, I doubt that 

 an exact petrological classification of igneous rocks can ever be attained. 

 We may arrive at some sort of approximation to an orderly arrangement 

 for purposes of petrographic description and petrological discussion, 

 which might by courtesy be called a classification; but in the attainment 

 of a classification sensu stricto we shall be apparently always restricted 

 by the inherently indeterminate characters of the things which we attempt 

 to classify, the igneous rocks. 



The prevalent belief in the possibility of the reference of igneous rocks 

 or comagmatic regions to a few definitely delimited groups is seen in the 

 conception of the so-called Atlantic and Pacific families of Harker, 

 Becke, and others. Apart from the now fairly obvious misapplication of 

 the names, there is evident throughout the literature dealing with them 

 the assumption that rocks or rock magmas may be referable to only two 

 great chemico-mineral groups. The attempts to reconcile this dualistic 

 idea with the actual occurrence of igneous rocks and comagmatic regions 

 are often futile, if not sometimes ludicrous. 



Von Wolff 44 was evidently struck by some of the difficulties inherent 

 in such a limited interpretation when he proposed that the plateau basalts 



44 F. von Wolff : Der Vulkanismus. vol. i, 1914, p. 153. 



