CHAP. VIII.] TRIASSIC PERIOD. 133 



which were doubtless much steeper and higher then than 

 they are now. 



It will be seen from the map (Plate VI.) that I have 

 supposed the Scottish Triassic deposits to lie on the sites of 

 two smaller and separate lacustrine areas. It might, of 

 course, be argued that the western basin was only an arm 

 of the Anglo-Hibernian lake, and that it was also connected 

 with the north-eastern basin by a narrow channel along the 

 site of the Great Glen ; but the absence of rock-salt and 

 gypsum in the Scotch Trias, and the presence of bivalve 

 shells resembling Cyrena at Ardtornish, are facts which 

 make it probable that the water of the Scotch lakes was 

 fresh, and not salt, and consequently that the connection 

 between them and the salt lake was by means of a river 

 channel. 



The absence of the Ehaetic beds and the estuarine cha- 

 racter of the Lower Lias in Sutherland are also facts which 

 confirm the view that the Scotch lakes were at a somewhat 

 higher level than the great salt lake ; and the contrast be- 

 tween the marine and estuarine character of the Lower 

 Lias of the two districts seems to indicate that the level of 

 the eastern was higher than that of the western. Hence 

 we seem justified in concluding that in Triassic times the 

 eastern discharged itself into the western, and that the 

 overflow of the latter was conducted by a river into the 

 western arm of the great salt lake. "We may, in fact, re- 

 gard the Scotch lakes as the mountain-fed reservoirs of the 

 Anglo-Hibernian lake. 



