118 MESSRS. J. R. DAKTNS AND J. J. H. TEALL ON THE 



gives rise to the formation of magmas more and more acid. Now, if 

 plutonic rocks represent the subterranean aspects of volcanic phe- 

 nomena, there should be a resemblance between the sequence of 

 plutonic and the sequence of volcanic rocks. Given one reservoir and 

 continuous cooling, there should be one sequence. The basic rocks 

 should precede the acid rocks. This has been the case in many- 

 regions, but the rule is by no means without exceptions. The 

 exceptions may be due to the existence of two or more reservoirs, or 

 to the accession of heat or fresh material during the process of con- 

 solidation. Taking all the facts into consideration, it does not seem 

 unreasonable to suggest that with one reservoir undergoing con- 

 tinuous consolidation there will be one definite sequence. 



Another speculation of considerable interest is suggested by the 

 facts before us. The consolidation of a plutonic magma may give 

 rise to rocks of which there are no volcanic equivalents. Such 

 rocks may result from the separation of minerals from a magma 

 having a very different composition from that of the rocks in 

 question. The absence of volcanic representatives of the typical 

 peridotites and their rarity as dykes may be due to this. Again, 

 since magnetite is the first mineral to form, it is quite possible that 

 it may arise on a large scale as a product of the consolidation of a 

 plutonic mass without ever existing as an actual magma. The view 

 that certain magnetic iron ores are of plutonic origin is one which 

 has been maintained by Wadsworth. If we look at the diagram 

 representing the composition of the rocks which have been analysed 

 we see that the maxima for alumina, lime, and soda lie within the 

 limits of the diagram. The maximum for magnesia is evidently 

 near the left-hand limit, but somewhat beyond. The maximum for 

 iron is not even approached within the limits of the diagram. Prom 

 what we know as to the order of formation of the minerals and the 

 influence of this in determining the character of the rocks, we may 

 speculate on the constitution of the diagram to the left. Lime and 

 alumina would fall. Magnesia and iron oxide (magnetite) would 

 rise. Then magnesia would fall, iron oxide still continuing to rise. 

 When the point of origin was reached we should have a rock com- 

 posed of magnetite. 



As possible rocks not represented in our series we may therefore 

 mention magnetite- rock and magnetite-olivine rock or cumber- 

 landite of Wadsworth. 



The view that subterranean reservoirs (magma-basins) of 

 homogeneous molten material may become differentiated into local 

 magmas of varying composition by the operation of various 

 imperfectly understood processes is rapidly gaining ground. Prof. 

 Kosenbusch deals with this subject in a paper ' Ueber die chemischen 

 Beziehungeu der Eruptivgesteine,^ ^ and Prof. Brogger has applied 

 the idea to the explanation of the sequence of rocks in the 

 Christiania district.^ It is also worthy of note that Prof. Brogger's 



1 Tschermak's ' Min. u. petr. Mitth.' Band xi. (1889) p. 144. 

 ^ 'Die Mineralien der Syenitypegmatitgange, etc.,' Zeitschr. f. Kryst.'u. Min. 

 Band xvi. (1890) p. 80. 



