OF THE OUTER HEBRIDES. 857 



diffused through it. The stones vary in size from mere grit up to 

 large boulders several yards in diameter. The boulders are generally 

 irregularly aggregated, in some places gathering together so as to 

 form a coarse breccia, with a scanty gritty matrix ; in other places 

 occurring only sporadically in a matrix through which smaller 

 stones are more equally distributed. Here and there one may detect 

 rude attempts at bedding, the stones being arranged in sloping lines 

 behind projecting rocks, whose glaciated faces look towards the 

 south-east. Occasionally, too, one comes upon traces of lenticular 

 patches of earthy gravel and sand. Many of the stones are well 

 smoothed and they often show strise, some of the larger blocks 

 being beautifully polished and striated on one or more sides. In 

 general the stones are blunted and subangular, but sharply angular 

 fragments are always present in larger or smaller numbers. The 

 till occurs in mere patches, and never occupies any wide area. It 

 lies in nooks sheltered from the east and south-east, and also 

 sprinkles the surface of the wide open valley to the east of Cuithir 

 (Cuir). As a rule the deposit is very thin, the greatest depth I saw 

 being only 8 feet. 



5. Erratics. — Loose stones and large blocks are scattered over the" 

 island. Not a few of these are glacially abraded, but many are 

 quite angular. The ice-worn boulders are in some cases merely the 

 wreck of the till, the finer materials of which have been denuded 

 away. In other cases they probably represent all the "till" that 

 ever existed. Some of the angular non-glaciated erratics attain a 

 very large size ; an immense boulder that forms a conspicuous land- 

 mark on a hillside near North Harbour was roughly estimated to 

 contain 24,000 cubic feet. All these sharply angular erratics I 

 believe to have been stranded during the final dissolution of the ice- 

 sheet: 



I noticed no trace of any later local glaciation ; neither terminal 

 moraines nor mountain-valley rock-basins. Loch an Dun is pro- 

 bably excavated in rock; but it does not appear to be deep, and 

 seems to be of the same character as those shallow lakes of Lewis, 

 Harris, &c. which owe their origin to the erosive action of the ice- 

 sheet, favoured by the configuration of the ground. 



X. Islands south oe Barra. 



The only island south of Barra that we landed upon was Bearna- 

 reidh ; but being favoured with beautiful weather we boated along 

 the others sufficiently near to determine the strike and dip of the 

 gneiss and the direction of glaciation. 



1. Bhaterseidh (Vatersey). — This island is separated by a narrow 

 and shallow channel from Barra. The highest point is not much 

 more than 600 feet. The hill-face immediately opposite Barra is 

 well glaciated, and the other hills appear to be not less well 

 smoothed off. Good roches moutonnees are seen at the south-east 

 corner of the island, just as one turns in from the south to Bhater- 

 seidh Bay. The upper part of the hills is somewhat bare, but the 

 lower grounds are well clothed with grass. There is much sand at 



