PROCEEDINGS OP GEOLOGICAL SOCIETIES. 290 
Seen in profile, as when looked at horizontally, they resemble so many arti- 
ficial hill- side cuttings, the back of each terrace lying within the general profile 
of the mountain slope, while the front or outer edge is protuberant beyond it. 
Each is indeed a nearly level, wide, deep groove, in the easily eroded boulder 
drift, or diluvium, which to a greater or less thickness everywhere clothes the 
sides of these mountains. They vary greatly in their relative distinctness, 
being in some places vaguely discernible, while in other spots they indent the 
surface very plainly, just as they happen to be narrow and to coincide in slope 
with the hill, or to be broad and apparently level from front to back. Where 
most indistinct they are frequently not discernible at all when we stand upon 
them ; though we may in a favourable light have detected their position and 
course from the opposite side of the glen, or, better still, from the bed of the 
valley. The conditions which influence this fluctuation in distinctness promise, 
if carefully observed, to dispel much of the obscurity which has hitherto in- 
vested the origin of the terraces. The modifying circumstances seem to be all 
referrable to one general condition, that of exposure to a current or inundation, 
supi)osed by the speaker to have rushed through these glens from their mouths 
to their heads, or upper ends. Thus it would appear : 1st, With scarcely an 
exception, that each terrace or shelf is most deeply imprinted in the hill-side, 
and is broadest where the surface thus grooved has its aspect down the ylen or 
towards the Atlantic, and is faintest where the ground fronts towards the head 
of the valley on the German Ocean. 2nd, While conspicuous on the open 
sides and the westward sloping shoulders of the hills, the terraces disappear 
altogether in the recesses or deeper corries which scollop the flanks of the 
mountains. 3rd, Each shelf, or " road," grows usually more and more distinct 
as it approaches the head of its own special glen, until those of the two oppo- 
site sides meet in a round spoon-like point. 
A fact obviously material to a true theory of the origin of the terraces is, 
that each of them coincides accurately in level with some water-shed or notch 
in the hills leading out from its glen into some other glen or valley adjoining, 
a coincidence suggestive of the notion that they were formed by the grooving 
agency of a flood pouring through the glens while it was embayed at the re- 
spective levels of these natural waste weirs. In confirmation of this view that 
they were transiently caused by erosive currents held successively at the 
heights of the barriers on whose levels the terraces terminate, we have as 
another interesting general feature, a remarkable ruggedness of the bed of each 
external glen just outside the water-shed or barrier closing the glen which con- 
tains the terrace. These rough and deep ravines, contrasting strikingly with 
the smooth spoon-like terminations of the terrace-lined glens which head 
against them, strengthen the suggestion already awakened by the marks of 
horizontal erosion in the terraces themselves, that the notches or passes which 
determined the grooving of the hill-sides on their one hand were externally the 
sites of so many stupendous cataracts. 
The internal structure or disposition of the matter composing each terrace, 
affords a further and striking corroboration of this hypothesis of the passage of 
an erosive flood. It consists in an " oblique lamination," or slant bedding of 
the constituents of the shelves — viz., the layers of gravel, sand, and other 
sediment, such as geologists familiarly recognise as the result of a strong 
current pushing forward the fragmentary material which it is depositing, and 
which is held by them to indicate in the direction towards which the laminae 
dip, the direction towards which the current has moved. Now, it is a most 
suggestive peculiarity in the oblique bedding of these terraces, that the " dip," 
or downward slant, is almost invariably up the glen, or towards its head, and 
not down the glen, or towards the Atlantic, as we must suppose it would have 
been, had the glen been a bay of the sea, and these materials but portions of 
