10 ON THE PARALLEL ROADS 



of the instrument around all these different points, and the re- 

 sults were perfectly satisfactory. In short, from all our expe- 

 riments, and from all our other observations, there did not re- 

 main a shadow of doubt in our minds, that the whole of these 

 shelves were perfectly horizontal in themselves, and that eve- 

 ry part of any shelf on one side of a glen, was decidedly of the 

 same level with the corresponding portion of it on the other 

 side. 



The breadth, or depth of these shelves, on the steep sides of 

 the hills, is very various, and is evidently much modified by 

 circumstances, particularly by the nature of the ground. That 

 part of the shelf is generally deepest, and most strongly mark- 

 ed, where the face of the mountain forms an acute angle, or 

 rounded promontory ; and this is more particularly the case, 

 where the promontory is of comparatively soft materials. In 

 all other places, whether bay or projection, where the surface 

 of the hill is soft, and easily worn away, the indentation is al- 

 most uniformly better defined than where a harder soil occurs. 

 Where rock manifests itself, little more generally appears than 

 a slight tracing on its surface, merely enabling the eye to fol- 

 low out the line with difficulty ; but on the fronts of many of 

 the rocks, all appearance of it is lost for a space, until it 

 again manifests itself on the softer surface. The indentation 

 of shelf 4th, on the rocks at the entrance to Loch Treig, is pe- 

 culiarly strongly marked; and the same shelf is also well defi- 

 ned in its circle round the top of Tom-na-Fersit, a small isola- 

 ted hill near the same point ; but these form, perhaps, the 

 only striking exceptions to what may certainly be considered 

 as a general fact. 



These remarks, apply rather to the degree of distinctness 

 with which the shelves are traced along the length of the glens 

 on the steep sides of the flanking mountains, than to their ac- 

 tual breadth. For when any one shelf approaches the point, 

 where, (by the rising of the level of the bottom of the valley) 



it 



