Mr. Macgillivray’s Account of the Outer Hebrides. 409 
waves, we can easily imagine how the latter should ultimately form 
caves init. Most of the caves of the Outer Hebrides are in the 
form of fissures, and are very seldom perpendicular, but incline at 
all angles. They seldom penetrate far into the rock, although a 
few instances occur in which a passage may be obtained for a length 
of several hundred feet. In many of them large blocks are seen 
sticking in warious parts of the fissures, and in more than one in- 
stance I have seen natural bridges of the most picturesque appear- 
ance thus formed. The effect of unequal pressure and subsidence 
upon the rocks would be the formation in some places of cracks, and 
in others the production of masses of fragmentary and comminuted 
matter jammed into the rock. Upon these the action of the waves 
would shortly operate, so as to enlarge the fissures, and wash out 
the loose fragments from the face of the cliffs, leaving the layer 
behind. Accordingly, in all the caves that I have seen, there is a 
flooring of great blocks of rock that have fallen from the roof 
and sides. Caves formed in conglomerate or sandstone are entire- 
ly different in their appearance, being rounded, uniform, and gene- 
rally perpendicular. The manner of their formation is easily un- 
derstood. In the whole range of islands I have not seen an inland 
cave but one, although numerous cliffs exist there, especially in the 
Forest of Harris. This cave is merely a horizontal cavity, formed 
by the removal of a wedge-shaped mass from the face of a precipice 
in the Glen of Ulladil in Harris. 
The most remarkable geological features which these islands pre- 
sent, at least those which most attract the attention of the observer 
on a general glance, are the veins of granite and trap which inter- 
sect the gneiss so frequently and to so great extent. The trap 
veins consist of greenstone and basalt, and frequently contain zeo- 
_ lite and specks of caleedony. They are always perfectly distinct 
from the containing rock. It is remarkable of these trap veins 
that almost all of them which are of any size fall short of the surface, 
and thus generally form the beds of torrents in the glens. Whe- 
ther they have originally come to the surface, and have been sub- 
sequently worn down by the action of the weather, I cannot say, 
but of the large veins of this kind I know only one that projects. 
It is also remarkable that most of them occur in the hollows between 
the hills, although they may also be sometimes seen in the highest 
position, as on the summit of Ronaval. 
The granite veins are exceedingly interesting. In them the con- 
stituent minerals are always in large concretions, and generally of 
colours different from those of the inclesing rocks. They have no 
distinct walls, but gradually pass into the neighbeuring mass. 
They must therefore be either of contemporaneous formation with 
that mass, or have been injected into its fissures in a liquid or fuid 
state. It is remarkable of these veins also, that many of them pro- 
ject beyond the surface, sometimes to the height of thirty feet. 
‘Besides the ordinary ingredients of granite, ee contain titanitic 
VOL. I. F 
