656 SIR A. GEIKIE AND MR. J. J. H. TEALL ON THE [Nov. 1 894, 



and, so far as this is the case, they do not imply heterogeneity in 

 the magma at the time of intrusion. 



Passing on to still larger areas, composed of typical plutonic rocks, 

 such as that of Garabal Hill and Meall Breac, 1 we often find what 

 is evidently a connected series of intrusions exhibiting a wide range 

 in chemical and mineralogical composition, and certain definite, 

 though probably not very great, differences in age. Such a plutonic 

 area, taken as a whole, though forming a petrological complex, is a 

 geological unit. The complexity in such cases cannot wholly be 

 accounted for by differentiation in situ. The more abrupt changes 

 require the hypothesis of successive intrusions, which probably repre- 

 sent differentiation elsewhere of the same general character as that 

 which has taken place to a smaller extent in situ. 



"We recognize, therefore, two causes of petrological complexity in 

 plutonic areas of the same kind as those referred to by Prof. Judd 

 in his paper ' On Composite Dykes in Arran ' : (1) differentiation 

 in situ, and (2) successive intrusions. But the facts described in 

 this communication appear to point distinctly to yet another cause, 

 namely (3) the intrusion of a heterogeneous magma. That the 

 banding in the gabbros is not due to differentiation in situ is proved 

 by the fact that the succession of more basic and more acid material 

 bears no definite relation to the margins of the banded series, as is 

 the case with dykes of the Bainy Lake Pegion described by Prof. 

 Lawson, 2 and with the Huk Dyke described by Prof. Vogt. 3 And 

 even if we could suppose that some physical cause, acting independ- 

 ently of the margins, might determine a segregation into parallel 

 bands, we should still be unable to account for the folding described 

 in the first part of this communication, which must have taken 

 place during intrusion and before consolidation. The hypothesis of 

 successive intrusions seems equally inapplicable, for, taking the 

 banded series as a whole, every intermediate type of rock between 

 the most basic and the most acid may be observed ; and where the 

 junctions between the dark and the light bands are fairly sharp, the 

 minerals, as we have already explained, interlock with each other 

 across the junction-line, just as they do in the interior of the bands. 

 These considerations have, therefore, led us to the conclusion that 

 the banding is the result of the intrusion of a heterogeneous magma. 

 We think that after reaching its present position, and before con- 

 solidation, the molten mass possessed those variations in composition 

 which are represented in Mr. Player's analyses (p. 653). According 

 to this view the heterogeneity must have been produced by causes 

 operating elsewhere, and probably at lower levels in the earth's 

 crust. We know nothing as to the original forms of the more or 

 less differentiated masses. Then came the intrusion of the hetero- 

 geneous magma as sills, and it was by the deformation of the molten 



1 ' On the Plutonic Rocks of G-arabal Hill and Meall Breac,' by J. R. 

 Dakyns and J. J. H. Teall, Quart. Journ. Geol. Soc. toI. xlviii. (1892) p. 104. 



2 'American Geologist,' vol. vii. (1891) p. 153. 



3 • Om dannelsen af de vigtigste .... Grupper 'Jernmalmforekomster,' Geol. 

 Foren. Forh. vol. xiii. (1891) p. 484. 



