408 Mr. Macgillivray’s Account of the Outer Hebrides. 
The shores of the Lewis are for the most part equally sinuous 
with those of the other islands, excepting on the north-western 
part, where they are low and sandy. From the west a large arm 
of the sea runs into the land, containing an island several miles in 
length, and others of inferior magnitude. One of its terminations, 
named Loch Rog, is remarkable, being so narrow and elongated as 
to resemble a deep river. It appears to occupy a great fissure 
formed among the hills, and was formerly the haunt of very large 
herrings, which however seem to have entirely deserted it. Of the 
other sea lochs on the west coast, the only one that deserves notice 
is Loch Resort, which forms part of the boundary between Harris 
and Lewis on the west side. The corresponding boundary on the 
east side is formed by Loch Seaforth, from its mouth to the place 
where it forms a right angle upon itself. Other lochs exist on 
the east coast, the most important of which, in a commercial point 
of view, is the Loch of Stornoway. The shores are in general 
rugged, excepting along the north-west coast, and part of the east- 
ern, where there are accumulations of sand. ‘The general rock is 
gneiss, as in the other islands, presenting the same appearances. 
Near the village of Stornoway a remarkable peninsula projects 
to the eastward, terminated by high cliffs of gneiss. The isthmus 
of this peninsula consists of a rock very different from any that oc- 
curs in the other parts of the Outer Hebrides, being a very coarse 
conglomerate, consisting of pebbles of all sizes up to a diameter of 
three feet or more, cemented by an argillaceous substance highly 
impregnated with oxide or hydrate of iron. ‘The pebbles consist of 
quartz, felspar, gneiss and granite of various kinds. It rests upon 
gneiss, and is seen forming cliffs.as far as the Chicken Head, and 
extending along the shores southward, forming a small island near 
Stornoway and the prominence called the Point of Arinish, and 
stretching northward as far as Tolsta Head. It bears a striking 
resemblance to the conglomerate which skirts the Grampians on 
their eastern side, although the pebbles of the latter are more fre- 
quently of quartz than of any other substance. 
The caves formed in this conglomerate are of a very different 
character from those formed in the gneiss cliffs. The former are 
low, and always rounded above ; whereas the latter are of all ima~. 
ginable forms, and sometimes of great height. That the action of 
the sea has greatly influenced the formation of large cliffs of gneiss 
along its shores is not admissible, for this reason, that such cliffs 
are generally unincumbered at the base by blecks, and are there 
frquently scooped into caves and fissures. Now, supposing the sea 
to have undermined a mass of gneiss, it is evident the superincum- 
bent part in falling would form a heap of fragments, which would 
effectually repel the sea, especially as the gneiss of these islands is 
of a nature almost impregnable to the elements. But if after some 
great convulsion, produced we may suppose by the upheaving of 
the gneiss by fluid or aeriform matter from below, a mass of that 
rock should settle upon the sea-shore, presenting a precipice. to its 
