1873.] GEIKIE — GLACIAL PHENOMENA OF OUTER HEBRIDES. 541 



but all description of these is reserved until, in a future communica- 

 tion, I come to treat of the younger glacial accumulations of the Long 

 Island. 



VI. The Freshwater Lakes of Lewis. 



A glance at the general trend of these lakes is enough to convince 

 one that large and small sheets alike have a strong tendency to assume 

 a linear direction, and that by far the greater number are arranged 

 along one or the other of two lines, which strike as near as may be 

 north-west and south-east and north-east and south-west respec- 

 tively. It even happens pretty frequently that the same loch 

 exhibits in itself both lines of direction, one portion of the water 

 trending at right angles to the other. In the mountain district of 

 the south a number of important lakes occur which do not coincide 

 in direction with either of the series just referred to, but extend 

 sometimes north and south, sometimes east and west. The origin 

 of these three groups of lakes has now to be considered. 



a. North-west-and-souih-east Lakes. 



With one exception all the longest and most considerable lakes 

 range in a direction from south-east to north-west. The group 

 to which these belong is most abundantly developed over the low- 

 lying rocky and moory region that extends across the whole island, 

 from the district of Pairc to the broken tracts around Loch Eoag. 

 North of Beinn Bharabhais they occur but seldom, and none are very 

 strongly pronounced ; indeed it is not unfrequently difficult or even 

 impossible to say what is the prevailing direction of the lakes in 

 that section of the island. But over the whole of the rocky and 

 swampy low grounds that abut upon the southern mountains, the 

 north-west- and-south-east lakes are not only exceedingly plentiful, 

 but they are also well marked. They extend in long lines, often 

 for a mile or two, with an insignificant breadth ; and not unfre- 

 quently several lakes are joined together by narrow necks of water. 

 Their shores are sometimes perfectly straight, but are oftenest very 

 irregular, being indented with numerous lilliputian bays and creeks. 

 These indentations, however, are quite subordinate to the trend of 

 the lakes. I may add also that the numerous small runnels that 

 flow into these lakes, and sometimes out of one lake into another, 

 have for the most part the same trend as the lakes themselves. 

 When this is not the case it will be found that the sluggish streams 

 creep in precisely the opposite direction (that is, from south-west to 

 north-east, or vice versa) ; and those parts of the lakes into which 

 such streams enter, show very generally a similar trend. 



Most of these lakes, especially in the district under review, are 

 very shallow ; but in the rocky tract on the northern borders of Loch 

 Eoag they seem to be deeper. Many of them contain islets. 



The shores of not a few I found to consist entirely of peat-moss ; 

 and in some places the peat seemed even to extend for some distance 

 into the water. Occasionally, however, the borders of the lakes 

 would exhibit till and rocky debris, with here and there the solid 



VOL. XXIX. PART I. 2 



