136 DR GEIKIE ON THE HISTORY OF VOLCANIC ACTION 



can be followed for some distance along the base of the hill as a marked projecting 

 escarpment (shown in the foreground of fig. 40). Higher up, other varieties are ranged 

 in successive parallel beds, the harder kinds standing out boldly as prominent ribs, while 

 the softer crumble into a kind of sand, which forms talus-slopes between the others. 

 Alternations of this nature are continued up to the very top of the mountain. The 

 beds are nearly flat, but dip slightly into the interior or towards the south-west. 



But not only are the gabbro and associated rocks disposed in beds differing from each 

 other in pctrographical characters. The same parallel arrangement may be traced even in 

 the internal structure of some of the individual beds. The most remarkable example of 

 this nature which I have found is presented by the band of light-coloured troctolite 

 just referred to. This rock at once arrests attention by its laminar structure. Indeed, 

 hand-specimens of it, as I have said, might readily pass for pieces of schistose limestone. 

 It consists of successive layers, which on the weathered surface divide it into beds almost 

 as regular as those of a sandstone, each bed being further separated into laminae marked 

 off by the darker and lighter tints of their mineral constituents. The darker layers consist 

 of olivine, and the lighter of plagioclase. This segregation here and there takes the form 

 of rounded masses, where the minerals are more indefinitely gathered together, and the 

 affinity of the rock with intrusive sheets is further displayed by the occurrence of 

 abundant nut-like aggregates of pale green olivine. Examined under the microscope, this 

 flow-structure is admirably seen, the lath-shaped felspars being drawn out parallel to the 

 planes of movement, and giving thereby the peculiarly schistose structure which is so 

 deceptive. 



The bedded arrangement of the gabbros is conspicuous from bottom to top of the great 

 eastern cones, as shown in figs. 39 and 40, and the dip is gently inw r ards to the W. or 

 SW. On the west side also, beyond Loch Sgathaig, a distinct bedding may be traced, 

 the inclination being here once more inwards or to the E. But from Glen Harris and 

 the base of Askival this structure becomes less marked, and gradually disappears. There 

 is thus a central or southern amorphous region, while round the margin towards the north 

 and east the rock appears in frequent alternating beds. 



It is clear that in the broad features of their architecture, the hills of Rum follow 

 closely the plan shown in the Cuillin Hills of Skye. In each case, there is a structureless 

 central region, where the rocks are more coarsely crystalline, and an outer marginal belt, 

 where they assume a bedded character and become finer in grain. But, unfortunately, in 

 Rum denudation has gone so far that no connection can be traced on the ground between 

 the gabbros and the plateau-basalts. As already stated, the latter rocks have been almost 

 entirely stripped off from the platform of sandstones and schists which they undoubtedly 

 at one time covered, and the few outliers of them that remain lie at some little distance 

 from the margin of the gabbro area. Nevertheless, we are not without some indications 

 of them underneath the gabbros. I have alluded to the basalts that lie at the base of 

 the eastern cones. As we follow the bottom of the gabbro southward round the flanks of 

 the hills, dull compact black shattery basalts, with a white crust, appear from under the 



