OE THE OUTER HEBRIDES. 835 



and roches moutonnees on hill-tops and exposed places. Nevertheless 

 there can be no doubt that the mountains of North Harris did deflect 

 the whole mass of the ice-sheet, as is proved by the changes in the 

 direction of the glaciation to the north and south of the Langa and 

 the Cliseam. These appearances, however, are best seen upon the 

 map (Plate XXXIIL). Unfortunately, owing to the small scale, I 

 have not been able to indicate all the bearings taken by us. 



4. Till or Boulder-clay. — The till of Harris presents, I need 

 hardly say, the same general character as the bottom-till of Lewis. 

 It may be described as an amorphous gritty and sandy greyish clay, 

 more or less abundantly charged with angular, subangular, rounded, 

 and, now and then, well-striated stones *. Sometimes, however, it 

 assumes the character of a tough dark-blue clay, with angular and 

 subangular stones and boulders ; in other places it contains masses 

 of laminated clay, silt, and sand, and, now and again, itself shows 

 a kind of rude bedding or, rather, arrangement of stones and boul- 

 ders ; while in yet other places it passes into a rough pell-mell mo- 

 rainic debris, made up of earth, clay, grit, angular gravel, and large 

 and small blocks of gneiss, many of which show traces of glacial 

 abrasion. 



If the glaciation of the rocks of Harris indicates clearly the direc- 

 tion from which the ice overflowed Harris, we shall find the evidence 

 supplied by the old moraine jprofonde of that ice no less decisive on 

 this point ; for, whether we have regard to the distribution of the 

 till, to its peculiar position, to the mode in which it has been, in 

 many places, gradually accumulated, or to the proofs of transport in 

 one determinate direction which its included boulders supply, we 

 shall be alike forced to conclude that the movement was from south- 

 east to north-west. 



The till nowhere forms such a broad and thick enveloping sheet 

 as the massive deposit that cloaks the Lowlands of centra] Scotland. 

 On the contrary, it is exceedingly patchy, resting for the most part 

 in valleys or on gentle declivities, and seldom reaching a greater 

 thickness than a dozen feet or so. But although it is so sparsely 

 developed, it nevertheless gathers in some bulk and spreads over 

 considerable areas along the western borders of South Harris. 

 Hardly a trace of it, however, is met with throughout the rough 

 undulating tract of bare grey rock that extends inland from the east 

 coast until it merges with the hills that overlook the western shores 

 of the island. This, of course, is what we might have anticipated ; 

 the till has gathered thickly in those places where it would be pro- 

 tected from the full grinding of the ice. This is well seen at Clife 

 Sheileboist, along the shores of Traigh Siar, at Scarrasta, and along 



* I have explained in my former paper why the stones in the till of a gneissio 

 region do not show stria so frequently as the stones in the till of the Low- 

 lands. Gneiss, as a rule, is not well adapted to receive and retain striae ; it is 

 too granular and coarse, and it is generally only the harder and finer-grained 

 varieties that show striation. Nevertheless the proportion of stones which 

 have been subjected to much abrasion is not less than in the till of many 

 upland tracts in central Scotland. The well-known blunted and subangular 

 form is always plentifully present. 



3k2 



