536 PROCEEDINGS OF THE GEOLOGICAL SOCIETY. [June 25, 



of this district certainly does not seem to promise much for the in- 

 terior ; and one who shall limit his investigations of the glacial 

 phenomena of Lewis to the neighbourhood of Stornoway will perhaps 

 think the evidence there does not go a long way in support of the 

 conclusions arrived at in this paper. In that district, where, as I 

 have said, the direction of dip is right against that of glaciation, the 

 latter can hardly be so striking as it would have been had the gneiss 

 dipped in some other direction. But fortunately in other parts of 

 the island the direction and angle of dip vary, while the trend of 

 the glaciation never does. 



The first unmistakably glaciated rock-face I met with in Lewis 

 was on the side of the road leading from Stornoway to Gearaidh nah 

 Aimhne, about a half-mile beyond Loch a Chlachain. It is a broad 

 flat surface of gneiss, planed and smoothed, from which, however, the 

 striae have disappeared. A little further on, at Beinn a Bhuna, the 

 watershed is reached amidst ruinous roches moutonnees. Here the 

 glaciated and mammillated aspect of the rocky knolls is still more 

 apparent, and the smoothed and glistening domes of gneiss seem to 

 give promise of a plentiful- crop of striae. But, although I searched 

 long, I found none ; the treacherous gneiss had failed to preserve 

 them. The rock at this place is for the most part coarsely crystalline. 

 It occurs in masses irregularly jointed ; and the bedding is either 

 very obscure or altogether lost. The direction of the glaciation, 

 however, as evinced by roches moutonnees, cannot be mistaken. It 

 is as near as may be from south-east to north-west. 



But perhaps the best display of roches moutonnees in Lewis, and 

 certainly the most easily accessible, occurs along the borders of 

 Loch Roag. Shortly after leaving Callernish, with its weirdlike 

 Tursachan, the road enters a rough hilly tract in which every dome 

 and boss of rock presents a finely mammillated surface. Here the 

 direction of the glaciation is distinctly south-east and north-west, 

 or the same as upon the watershed at Beinn a Bhuna and in the 

 district of Arnish moor near Stornoway. The dip of the gneiss is 

 moreover very variable, being sometimes north and north-east, the 

 strata at the same time appearing highly crumpled and contorted. 

 Occasionally, indeed, the bedding even vanishes altogether. In 

 this district therefore the direction of the roches moutonnees cannot 

 be ascribed to the structure of the gneiss. The slopes of the ground 

 are evidently quite independent of the dip and strike of the rocks ; 

 and even were there no dome-shaped hillocks and mammillated 

 knolls, we could hardly ascribe the peculiar conformation of the 

 ground to any other agency than that of land-ice. Rock-faces 

 innumerable were scanned for striae ; but although something like 

 faint ice-markings occasionally appeared, none of these could be 

 considered sufficiently pronounced. I was beginning to despair of 

 finding any, when at last I stumbled upon a beautifully striated rock- • 

 surface from which the till had been recently removed. As this 

 was the only distinctly scratched rock-surface I met with in the 

 low grounds of Lewis, it may be as well to particularize the locality. 

 It occurred in a little opening or quarry-hole close to the roadside 



