Vol. 59.] TORRIDONIAN, ETC. OF THE ISLE OF RUM. 199 



several places by abruptly-intruded igneous masses. The relations 

 are of a complicated kind, and it appears that more than one 

 overthrust occurs. The displaced strata are mostly shales in a 

 metamorphosed state, and they have high dips, often approaching 

 the vertical, in various directions, but chiefly to north and north- 

 east. The metamorphism is not always of a very advanced grade, 

 the only conspicuous new mineral being brown mica in flakes set 

 parallel to the lamination. The angular granules of quartz, which 

 are abundant in some beds, are quite unaltered [10489, 10490]. 

 Some associated sandstones, evidently metamorphosed, also show 

 the formation of brown mica, the detrital grains of quartz and 

 felspar being unchanged [10488]. In another specimen the 

 alteration is greater, the clastic structure being partly obscured. 

 Here some green hornblende has been produced, and in places 

 granules of a pale augite, apparently along a veinlet which has 

 contained a little calcareous matter [10487] . A crush-breccia, 

 which occurs in a few patches of no great size, has also undergone 

 metamorphism. 



The line of the overthrust runs eastward, passing north of Loch 

 Bealach Mhic JNeill and Loch Gainmhich, and then turns more to 

 the south-east. All along this line the sandstones below, or to the 

 north of, the main overthrust show no special sign of disturbance, 

 except a change of dip. As we approach from the north the 

 inclination of the strata becomes steeper, and takes a northerly 

 direction, and as we pass eastward the dip becomes east of north, 

 that is, still away from the overthrust. In Coire Dubh, opening 

 north-eastward towards Kinloch, these highly tilted sandstones 

 are cut off by a subsidiary surface of movement, and below this 

 minor overthrust the beds have the normal inclination (see fig 3 

 p. 198). ' 



In that portion of the displaced belt of rocks which we have 

 followed so far, sandstones are seen in several places above the 

 main surface of the movement. They occur always south of the 

 overthrust shales, which they doubtless succeed, though the ob- 

 served dips show that the actual junction is not a natural one. 

 The small areas of crush-breccia, which have been mentioned, are 

 found in like situations, and it appears that the conjunction of 

 sandstones and shales has favoured the production of the brecciated 

 structure. It has already been remarked that the undisturbed 

 succession shows a certain alternation of shales and sandstones at 

 the boundary between the two groups, and it is to be supposed that 

 the unequal yielding of the two rocks under mechanical force* 

 would conduce to the setting up of brecciation. The conditions 

 were indeed comparable, in many respects, with those which affected 

 the Manx Slates as described by Mr. Lamplugh. 1 



The most considerable mass of crush-breccia in Rum is that which 

 crosses Coire Dubh (fig. 3), and forms much of the slopes of Meall 

 Breac and Cnapan Ereaca on the two sides of the corrie. Though 



1 Quart. Journ. Geol. Soc. vol. li (1895) pp. 563-88 ; & ' The GeoWv of 

 the Isle of Man ' Mem. Geol. Surv. (1903) pp. 55-58. 6 



