On the Pagan Monuments of the Outer Hebrides. 357 
of these can be approached on foot at low water, and the rest 
are only separated by shallow channels. On going round the 
coast I began to notice that the peat banks formed frequently 
the sea-margin at high water ; and feeling that it was a point 
of some geological importance, I noted the places on my sketch 
of the coast, till at last finding that the phenomenon was ex- 
tremely common, occurring perhaps twenty times in a day's 
walk, it was no longer necessary to record its occurrence. 
After examining hundreds of examples, I came to the conclu- 
sion that, in these instances, there was no other way of 
accounting for peat banks having that position with regard to 
the sea-margin, than by the subsidence of the land. In many 
of the cases referred to, the peat rests at once upon the naked 
gneiss. Now, the surface of the gneiss is as rugged as, and 
not very unlike what, the surface of the ocean would be if it 
were suddenly solidified in the middle of a violent storm. The 
peat has grown over the undulatory surface, but, as was to be 
expected, is deepest in the depressions, where I have sometimes 
seen it in section more than ten feet in height. Let it be sup- 
posed that the relative levels of land and sea remained the same 
until the formation of peat had occurred, and then that the 
land began to subside. The effect would be that when the sea 
reached the foot of the peat banks small cliff's would form, 
identical in their features with those of clay or gravel, or any 
other soft material. But from the extreme softness of the peat 
the sea would quickly eat it away, and Avork itself into all the 
sinuosities of the surface gneiss that were below the level of 
high water. This is what has happened around Benbecula ; 
the sea at high water flowing in amongst the knolls, at first 
forming a simple gap, then branching out like the arms of a 
tree ; at a further stage in the destruction of the peat, a group 
of islands is formed, till at last the whole vegetable soil is 
washed away, and a few bare rocks serve to point out where but 
one generation ago the cattle have pastured and corn has 
grown. Some of these rocks, which are now completely bare, 
still bear the names of particular plants that once grew upon 
them. It is difficult in words to convey an idea of the 
appearances and of the facts ; but an inspection of the coast 
would show that the peat could only have been brought into 
