﻿Vol. 66.'] CLASSIFICATION OF IGNEOUS KOCKS. 479 



be derived from one homogeneous earth-magma. The only escape 

 from this lies in the assumption of hypothetical primary differences 

 in the earth-mass. Rosenbusch l and many others doubtless assume 

 the original homogeneity of the earth-magma. Differentiation in 

 the abstract might be used as a systematic factor, if we knew the 

 composition of the supposed primordial magma, and if we were 

 sure that the composition of that magma could never be repeated 

 by the mixture of differentiates. But such division of rocks is 

 not likely to be proposed, for divers evident reasons. 



Aschistic and diaschistic magmas. — Differentiations may 

 lead to a general distinction between parent magmas and their 

 derivatives, passing over the supposition that the great differences 

 in composition of parent magmas themselves are due to differ- 

 entiation. This would be making a general division corresponding 

 to that which Brogger has found useful between the parent magma 

 of a given centre and the partial magmas derived from it, and 

 expressed by the respective terms aschistic and diaschistic. 2 



The introduction of this distinction into general system would be 

 difficult and of doubtful utility. Thus the parent magmas of various 

 regions, or centres of eruption, differ so much that the early differ- 

 entiates of one would certainly be practically identical with the 

 parent magma of another centre ; many rocks do not visibly belong- 

 to a series, and the personal equation would necessarily enter into 

 many cases where judgment is necessary. For these and other 

 reasons no proposition of this kind is likely to be made. 



The 'dyke rocks' of Rosenbusch. — Leaving general con- 

 siderations, we now come to the definite proposition of Rosenbusch 

 to distinguish the class of 'dyke rocks,' really on a genetic 

 chemical basis, though this is obscured by the name. Here is oppor- 

 tunity for a discussion touching practically all phases of the problem 

 as to the application of chemico-genetic factors in petrographic 

 system. 



For the analysis of this important proposition it will be well to 

 have before us the most recent definition of the 'dyke rocks' as a 

 systematic division. In the 4th edition of the ' Mikroskopische 

 Physiographic : Massige Gesteine ' vol. ii, pt. i (1908) pp. 487, 488, 

 is the following: — 



' Als Ganggesteine bezeichne icli nur solche Eruptivmassen welche als 

 sell) standi ge geologische Kcirper nach dem augenblickiichen Stande unserer 

 Erfahrungen nur in typischer Gangfonn unci im Vergleioh zu den Tiefen- 

 gesteinen ge ringer Masse des einzelnen Gesteinskorpers auftreten. .... Die 

 stoffliche Natur der wichtigsten Ganggesteine ist [dagegen] bedingt, nicht durch 

 diese selbst, sondern durch gewisse Tiefengesteine, an die sie gebunden Bind und 

 ohne die sie nicht sein wiirden.' 



1 'Ueber die Chemiscben Beziehungen der Eruptivgesteine ' Tscherm. Min. 

 Petr. Mitth. vol. xi (1889) p. 144. 



2 ' Die Eruptivgesteine des Kristianiagebietes, 1 : Die Gesteine der Grorudit- 

 Tinguait Serie ' 1894 (Vidensk. Selsk. Skrifter, I. Mathemat.-Naturv. Klasse 

 No. 4) p. 125. 



