542 PROCEEDINGS OF THE GEOLOGICAL SOCIETY. [June 25, 



rock appearing ; while in yet other instances the water rested in 

 complete rock-basins. Some good examples of the latter will he 

 found in the district south of Loch Carloway. 



Whatever theory we may form as to the origin of these lakes, it 

 will probably be admitted that their linear direction is directly due 

 to the configuration of the underlying rocks and till. We cannot 

 suppose it likely that their peculiar shape has been produced by the 

 gradual encroachment of the surrounding peat-mosses upon pre- 

 existing lakes, which may at one time have shown quite a different 

 outline. Again, if any of the lakes do rest throughout upon a bed 

 of peat-moss (and I am far from denying that such may be the case), 

 still we must admit, as it seems to me, that the lacustrine depres- 

 sions or hollows existed before the peat ; for the growth of a bog is 

 regulated by the supply of water received, and this, again, is obviously 

 controlled by the configuration of the ground. The hollows, there- 

 fore, cannot be in the peat, but in the surface upon which the peat 

 rests. And that surface consists either of rock or till, or of both 

 together. 



While describing the glaciation of Lewis I confined my remarks 

 to the appearance presented by hills and rocky knolls ; and these pro- 

 minences, as I showed, are distinctly ice-marked from south-east to 

 north-west. Now the hollows occupied by the lakes I take to have 

 been formed at the same time as the rocJies moutonnees. They occur, 

 as may be seen, either in solid rock, or in till, or, lastly, in both. 

 When the ice that swept across Lewis finally vanished it left as 

 marks of its power not only rounded and fluted hill-tops, but 

 hollows scooped out in the solid gneiss. The till that accumulated 

 below the ice was also at the same time found arranged in long 

 parallel banks, running in the exact direction followed by the ice- 

 striae and roclies moutonnees. The arrangement of the till into long 

 parallel mounds is a feature with which I have long been familiar. 

 It is admirably displayed in the valley of the Tweed, in Mthsdale, 

 and other wide dales and straths in the south of Scotland ; and pre- 

 cisely the same appearance characterizes the deposit in Ireland *. 



The north- west-and-south-east lakes, then, rest in true rock-basins, 

 and also in hollows between parallel banks formed wholly of till or 

 partly of till and rock. 



h. Nortli-easi-ancl-souili-xvest Lahes. 



The north-east-and-south-west lakes, while resembling in general 

 appearance those of the group just described, are, as a rule, more 

 compact in form. They are also, for the most part, sharply marked 

 off from each other, and seldom occur in those long straggling chains 

 so characteristic of the north-west-and-south-east group. They are 

 best developed in the district west and south-west from Stornoway, 

 but appear also in considerable numbers intermingled, and in many 

 places coalescing with the north-west-and-south-east lakes on the 

 low grounds between Loch Eoag and Pairc. None of them, as far 



* See ' The Glaciation of Iar-Connaught/ &c, by Messrs. Gr. H. Enahan 

 and M. H. Close: Dublin, 1872. 



