Mr. Brtce on the Parallel Roads of LocTidber. 101 



to within 6 or 7 miles of Fort-William. In Glen Gluoy, on the contrary, 

 the upper shelf extends farther down the glen than the lower. Each shelf 

 is on a level with some watershed, that is, with some col, or landstrait, or 

 lowest part of the ridge dividing two glens, as in the annexed sketch, 

 No. 2. 



a a Ridge dividing two glens, h Lowest part of ridge, or col. c Shelf. 



Thus, shelf No. 1 is on the level of the passage leading from Glen Gluoy 

 into Glen Toorat, which branches off Glen Roy near its upper end. No. 2 

 stops at the extreme north-eastern angle of Glen Koy, near the level of 

 the opening into Strathspey ; and so of the other shelves, as expressed 

 by the arrows on the map referred to in the next section. The only 

 exception is the second Glen Gluoy shelf, (No. 3 ',) which is not on a level 

 with any watershed. 



Up to so recent a date as 1817, the Parallel Roads were regarded as 

 works of art ; but it is now agreed on all hands that they are due to 

 natural causes. If we suppose that, in a former condition of things, the 

 sea penetrated to these glens, or that, the mouths of the glens being 

 blocked up by earthy materials, or by ice, the waters of the rivers accu- 

 mulated behind so as to form lakes, then, the action of the water on the 

 alluvial coating of the hills, and on the earth and stones which descended 

 from the heights and were arrested and re-arranged at the margin, would 

 form a beach line such as we now see upon most shores when the water 

 stands a little lower than usual. The shelves thus mark the successive 

 levels of the water as the sea retired on each upheaval of the land, 

 or as the lakes sank to successively lower levels, by the partial disrup- 

 tion of the barriers. The cols, or passages between the glens, coincident 

 with the several shelves, are, according to one theory, the channels or 

 straits between islands ; in the other they mark the levels where the 

 redundant waters flowed out from glen to glen, during the time that the 

 lakes were forming the several terraces. In confii-mation of the latter 

 view, it has been shown that there are at the cols several deserted river 

 channels, having no reference to the present drainage. 



III. — Error of tub Maps. 

 Mr. Chambers' account of the Parallel Roads, with his theory of their 

 origin, forms a portion (pp. 95-130) of hia- valuable and beautifully illus- 

 trated work on Ancient Sea Margins. A map of part of Lochaber, 

 showing the shelves in the glens, is given at the end. It has been " con- 



