204 MR. A. HARKER ON THE OVERTHRUST [May I903, 



the Torridonian strata with which they are in most cases associated. 

 It remains to enquire their relation to the next marked epoch in the 

 geological history of Hum, namely, that of the overthrusting, which 

 we may safely correlate with the great Palaeozoic crust-movements in 

 other parts of the Scottish Highlands. In no case is there anything 

 to indicate that the gneisses themselves have "been moved. The 

 boundaries of the lenticular masses, which on that supposition would 

 be favourable places for shearing and faulting movements, seem to be 

 in fact, so far as they are exposed, normal igneous junctions. The 

 lenticular form and the general parallelism with the local strike are 

 features common in ordinary laccolitic intrusions, and due to the 

 natural tendency of an intruded magma to follow the direction of 

 least resistance. Moreover, this parallelism does not extend to the 

 gneissic banding, which is certainly of the nature of a primary 

 flow-structure. "With rare exceptions, the gneisses show no sign of 

 crushing, brecciation, or internal fracture of any kind. Some of the 

 occurrences are in the immediate neighbourhood of, or even in con- 

 tact with, crush-breccias; but the gneisses have not contributed to 

 the composition of these breccias, a fact inexplicable if the crushing 

 be supposed posterior to the intrusion of the gneiss. The Dibidil 

 district does, indeed, afford a certain exception to this generalization, 

 for there brecciated gneiss is seen in one or two places. Such 

 occasional phenomena do not, however, invalidate the argument ; 

 for, apart from all consideration of these gneissic rocks, it is probable 

 that, here as elsewhere, there was in Tertiary times a renewal of 

 mechanical disturbance along the old lines, though with much less 

 intensity. Brecciation and other effects of crushing are found on a 

 much more extensive scale than at Dibidil in undoubted Tertiary 

 granites and gabbros in some parts of Skye. The observations 

 which have been adduced warrant us then in believing that the 

 gneissic rocks of Bum were intruded at some time posterior to the 

 great crust-movement of the region. This conclusion still leaves 

 two alternatives open : the intrusion of the gneisses may conceivably 

 have followed the overthrusting after no long interval, being a later 

 incident of the same great system of disturbances, or it may be 

 referable to a distinct and long posterior epoch. The only later 

 igneous intrusions known to have affected this part of Scotland are 

 those of the Tertiary suite, and the question of the date of the 

 Rum gneisses turns, therefore, upon the relations of these rocks to 

 those of admitted Tertiary age, which are abundantly represented 

 in the same tract. We shall see that the observed relations afford 

 somewhat strong ground for believing that the gneisses are of late 

 geological date, and are merely special phases of the plutonic 

 intrusions of the mountains. 



The rocks building the large intrusive bodies fall into three 

 principal groups, which succeeded one another in order of decreas- 

 ing basicity : (i) the ultrabasic group, comprising peridotites, 

 olivine-anorthite-rocks, enstatite-anorthite-rocks, and many other 

 varieties ; (ii) gabbros ; (iii) granites, including granophyres. In 



