142 Proceedings of the Boyal Society 
The gentleman who expresses this opinion, and reports on the 
Foula boulders, affirms unhesitatingly, that these boulders came 
from the mainland of Shetland, separated by a deep sea of from 
16 to 18 miles in breadth ; and he even specifies the hill from 
which the boulders have come. 
So also the report from the Lewis is distinct, that there are 
boulders on the east shores of the island which must have come 
a distance of at least 35 miles across the sea from the mainland of 
Sutherland. 
The important bearing of these cases of island boulders on the 
nature of the transporting agent, is evident. It is difficult to see 
how glaciers could have been instrumental in carrying them . 
4. The information recently obtained by the Committee throws 
additional light on the direction from which the boulders have 
come. 
(1.) As already mentioned, some boulders appear to have come 
down valleys, brought apparently on glaciers. Two localities 
are mentioned, viz., in Lochaber and in Nairn Valley, where such 
explanation may be accepted. 
There are also probably places in Perthshire, Forfarshire, and 
Aberdeenshire, where boulders lie in valleys, or at the mouths of 
valleys, which belong to this class of cases. The nature of the 
rock composing the boulders being found to be the same as the 
rocks in situ at the head or along the sides of the valleys, the 
birth-place of the boulders may be readily and correctly assumed 
to have been there. 
(2.) But there are hundreds of localities with boulders, to which 
this explanation is inapplicable. Not only boulders situated on>. 
islands, but boulders perched high up on hill-sides and near 
mountain-tops, seem to require a different explanation. 
On the mainland of Scotland, at least in its eastern half, from 
which the Committee have received the fullest information, the 
angular boulders, and also many of the rounded boulders, appear 
to have come over a wide extent of country in one and the same 
direction, viz., from the north-west, crossing valleys and ranges of 
hills. 
This inference had been drawn, years ago, from such facts as 
the finding of granite and mica-schist boulders in the counties of 
