Vol. 59.] TORRIDONIAN, ETC. OF THE ISLE OP RUM, 207 



and where, too, disturbance posterior to the intrusion of the gneiss 

 was sufficiently intense to produce noteworthy brecciation in that 

 rock. It may be plausibly conjectured that the movements of 

 which this last effect is the witness were initiated somewhat earlier, 

 and that the fluxion in the gneiss itself is related to it. 



V. The Composition and Origin op the Gneisses. 



I have now to discuss the petrographical characters of the 

 gneissic rocks of Rum. In doing so, I shall adopt the conclusion 

 to which the field-evidence has led me, and treat the gneisses as a 

 special facies of the Tertiary plutonic rocks of the region. It will 

 be seen that this position is strengthened by many peculiarities of 

 the gneissic rocks themselves, to be noticed below. One explana- 

 tory remark is called for at the outset. Considered simply as 

 intrusions, the gneisses are to be correlated, as we have seen, 

 with the granites, but, having regard to their actual materials, the 

 granite-magma has supplied only one element, though the dominant 

 one, of the gneissic complex. The basic and ultrabasic rocks have 

 contributed in a minor degree, and so also to some extent have the 

 Torridonian sediments. These various subsidiary components are 

 for the most part much disguised, and have often quite lost 

 their individuality in the resulting complex, for the acid magma by 

 which they have been enveloped and impregnated has not only 

 metamorphosed, but often partly or wholly digested them. In a 

 word, much of the gneiss is a hybrid rock, and this composite 

 origin frequently betrays itself in unusual mineral associations. 

 Since the effect of my study of the rocks is to connect the gneisses, 

 on petrographical as well as on geological grounds, with the Tertiary 

 suite of intrusions, we will first glance at the occurrence of gneissic 

 structures in rocks admittedly of Tertiary age in the same petro- 

 graphical province. 



Nine years ago, Sir Archibald Geikie & Mr. Teall l described 

 the highly developed gneissic banding in the Tertiary gabbros 

 of Druim an Eidhne, in Skye, and pointed out the instructive 

 bearing of the jDhenomena described upon the origin of such 

 gneisses as those of the Lewisian Series. The present writer has 

 found that such banding affects in varying degree a considerable 

 portion of the gabbros of both Skye and Eum, while it is much 

 more prevalent, and attains a more striking development, in the 

 more variable group of ultrabasic plutonic rocks in the same 

 islands. The authors cited proved clearly that (at the locality 

 described by them) the banding has resulted from the intrusion of 

 a heterogeneous magma, which was drawn out into parallel streaks 

 by flowing movement without any effective intermingling of the 

 different portions. Such is undoubtedly the explanation of the 



1 Quart. Journ. Geol. Soc. vol. 1 (1894) pp. 645-59 & pis. xxvi-xxviii. 



