842 J. GEIKIE ON THE GLACIAL PHENOMENA 



The same holds true of erratics. They are exceedingly nume- 

 rous in Forth Harris ; but in South Harris, although they are 

 plentiful enough, they only in a few places occur in particular 

 profusion. They are rather numerous, for example, in the neigh- 

 bourhood of East Stoc-cleit, where they are dotted over roches mou- 

 tonnees, till, and morainic debris. Many of the erratics in South 

 Harris, however, have been derived from the till ; and the same, of 

 course, is true to a considerable extent in North Harris also. The 

 till contains numerous boulders of precisely the same character as 

 those lying loose at the surface. Often, indeed, it is merely a coarse 

 aggregation of large and small angular and subangular blocks and 

 debris. Again, over wide areas in South Harris, particularly in the 

 bare rocky districts that slope towards the Minch, its only represen- 

 tatives are large blocks and smaller stones scattered loosely and 

 sparsely over the surface, many of them quite angular and others 

 more or less abraded and blunted. Such is the character of the 

 " till " wherever the ground opposes a more or less abrupt declivity 

 to the south-east ; it consists of a mere sprinkling of erratics and 

 small angular and blunted boulders, which here and there gather 

 into heaps on the north-west or lee side of roches moutonnees and 

 prominent crags. But the larger angular erratics and perched 

 blocks occur decidedly in greatest abundance on mountain-slopes and 

 in valleys which have contained local glaciers. 



8. Freshwater Lakes and Sea-lochs. — The freshwater lakes of 

 Harris, although numerous, do not form so striking a feature in the 

 scenery as those of Lewis and other parts of the Long Island. Ex- 

 cellent bird's-eye views of the Harris lakes are obtained from the 

 tops of the Cliseam, the Langa, Clisebhal, and other hills in Worth 

 Harris, and from Cluinisbhal and Roinebhal in South Harris. Like 

 a large number of those in Lewis, many of the Harris lakes range 

 themselves along the line of strike, those that lie across that lino 

 being chiefly confined to mountain-glens. The larger number of the 

 lakes and lakelets we visited appeared to rest either in complete 

 rock-basins or in hollows formed partly in rock and partly in till or 

 other glacial debris. Many, however, are mere shallow pans, and 

 have been encroached upon by peat. Some of the more interesting 

 of the N.W. and S.E. lakes are scattered over the rocky tract be- 

 tween Eoinebbal and Cluinisbhal, where they form a very picturesque 

 contrast to the weird, grey, glaciated, and weather-worn roches mou- 

 tonnees amongst which they lie. All these clearly belong to the 

 period of general glaciation, and have nothing whatever to do with 

 local glaciers. The strike of the gneiss in Harris not wheeling 

 about as it does in Lewis, there are no " strike-lakes " corresponding 

 in direction to the N.E. and S.W. lakes of the region lying to the 

 west and south-west of Stornoway. 



The linear lakes whose trend does not coincide with the strike of 

 the gneiss are, as I have said, confined to mountain-valleys, and owe 

 their origin to the action of local glaciers, the terminal moraines of 

 which occur in the immediate vicinity. Hence they follow the trend 

 of the valleys ; which may sometimes be at right angles to the 



