1865.] BRYCE AERAN DEIFI-BEDS. 211 



of Arran adown all the glens to the outer margin of the island, on 

 all the coasts. Some of these, even at remote points, are of enor- 

 mous magnitude, such as water never could have borne ; and they 

 have passed down slopes, across glens, up steep inclines, again and 

 again, to reach the points where they are now lodged. A few masses 

 of slate and Old Eed Sandstone have also travelled; but the majority 

 of the travelled blocks are of granite of the two kinds which constitute 

 the northern nucleus. They are in two streams : one, on the east 

 side of the dividing ridge of the island, is wholly of the coarse- 

 grained granite constituting the GoatfeU group ; while another, on 

 the west, and emanating from Glen lorsa, consists partly of the fine- 

 grained variety which forms the bottom and sides of that glen, and 

 partly of the coarse-grained which rises over it in high ridges. In 

 many places striated and grooved rocks and " roches moutonnees " 

 are found, indicating the action of ice as high as 1200 feet. There 

 is a wide enough space near the centre of the mountain-group to 

 maintain a considerable snow-field as a feeder of glaciers. Now, as 

 supporting the view towards which these facts point, it is most 

 worthy of note that, as regards these appearances, Arran is perfectly 

 isolated: the transported blocks are, undoubtedly, all of its own 

 granites and other rocks ; nothing has been contributed by the ad- 

 joining mainland*. Terraces on the sides of the valleys, composed 

 of alluvial matter, not of rocks, and great mounds at the mouths of 

 the glens, such as those at lorsa waterfoot, which are the admiration 

 of all visitors, point in the same direction. We thus seem almost 

 shut up to the conclusion that the whole island, raised as now, or 

 perhaps to higher levels, may have been wrapped in sheets of ice, 

 across which the granite blocks falKng from the precipices where 

 snow could not lie, or torn off from the sides of the glens, would be 

 carried forward in all directions on the glaciers filling the valleys. 

 The glens being filled up with ice to the level of the bounding ridge, 

 this might thus be passed over, and a new descent begun, and thus 

 the load of blocks might surmount successive obstacles, and so come 

 to crowd the southern plateau, and strew the southern shores. In 

 this state of the surface the true Boulder-clay or lower TiU may have 

 been formed by the joint action of ice and streams of water exist- 

 ing among the glaciers, after the manner of the beds now laid down 

 on land by ice. The absence of fossils, the unstratified character, 

 the pell-mell admixture of striated stones of all sizes, unite in proving 

 for the Boulder- clay an origin on land, with which ice, acting with 

 powerful force, was somehow connected. The angular form of many 

 of the blocks, and the preservation of the striation and polishing on 

 all alike, clearly show that there was neither an after readjustment 

 by currents nor a lengthened transport ; for such an action would 

 have obliterated the markings, as in the case of the alpine streams. 

 Now the distribution of the shell-bed marks the Kmit of a former 



* This interesting and important fact was first published by Dr. MaccuUoch, 

 in 1819, but has not been alluded to by later writers. He showed that the 

 transported blocks all belonged to Arran ; without forming any theory, he 

 ascribed the fact in a general way to revolutions of the surface. 



