or liaised Beaches in Scotland, Sfc. 283 



Good specimens of these banks of sand and gravel are seen 

 at Ballaichan and Grandtully, above Logierait, as well as at 

 that village, and all the way down to Dunkeld. Sometimes 

 two may be seen with the river flowing between them, divided 

 by a space a quarter of a mile in breadth, and shewing, by 

 the steep sides they present to each other, that they had once 

 been united. There is a well-marked terrace at Taymouth 

 Castle apparently about 35 feet above the stream ; the village 

 of Kenmore occupies a detached part of its western extremity. 



From Grandtully down to Dunkeld, the river runs across 

 the strata, and the sides of the valley are steep, rugged, and 

 picturesque. From Grandtully up to Kenmore, the course of 

 the river is alon^ the strata, and the valley is wider and more 

 tame. The terraces are much more conspicuous in the former 

 than the latter. 



The hypothesis which would ascribe such deposits to an an- 

 cient lake, seems to me inapplicable here. Let us consider 

 the case of the terraces at Dunkeld. If a barrier of rock ex- 

 isted there, to confine the water, and produce a lake, its posi- 

 tion must have been at the outer margin of the mountains, that 

 is, at Birnam Hill, which is about a mile and a half below Dun- 

 keld. Supposing the river to cut a passage here, through the 

 mass 'of hard quartzy slate, that passage would be of the 

 breadth of the stream (about 400 or 500 feet), or not much 

 greater. But at the level of the terraces (for it matters not what 

 it is 100 feet lower), the opening between the hills which 

 confine the valley is certainly not much less than a mile in 

 breadth. 



The hypothesis, however, which ascribes the formation to 

 the sea, is not without difficulties. The well-marked terraces 

 along the valley, as far as Grandtully, seemed to me, judging 

 by the eye, to vary in height above the river, from 30 feet to 

 100 or 120. Now, in the 12 miles from Dunkeld to Grand- 

 tully, the rise in the bed of the Tay can scarcely be less than 

 100 or 120 feet ; and if the relative levels remained the same 

 when the sea stood 300 feet higher than at present, the line 

 of terraces beginning with that of Claypotts, 120 feet high, 

 and continued up the valley, should have sunk gradually, and 

 disappeared at or near Grandtully ; while the one existing 



