The Glacial Theory of Prof. Agassiz. 351 
The ice, in consequence of its tendency to dilate, and its numerous 
fissures, accommodates itself to the sinuosities of the rocks which con- 
fine it, cutting off the smaller projections, and rounding and_ polishing 
the larger, which assume the form of domes, and were termed roches 
moutonnées by Saussure. Agassiz’s eighth plate gives some fine exam- 
ples of these rounded swells. Owing to the immense pressure, the 
included. pebbles of conglomerates, and the hardest veins inshpelndia 
rocks, are cut away to the very same level with the softer parts peas 
envelop them. 
Thus, one of the marks by which the ancient existence of glaciers 
can be detected in situations where they are no longer seen, is the pol- 
ished, striated, or grooved appearance of the rocks. Sometimes it is 
very distinct, but in many cases it is not visible, because the surface of 
most rocks wastes away by disintegration or decomposition, unless it is 
well protected by a covering of clay or turf. The most satisfactory 
specimen near Edinburgh, is in the quarry on the south side of Black- 
' ford Hill, at a place laid open a few years ago, where the rock leans 
forward, forming a sort of vault. The surface of the clinkstone here, 
for a space of ten or twelve feet in length, is smoothed, and marked 
by strie@ or scratches in a direction approximating to horizontal. We 
accompanied M. Agassiz to the spot about two months ago; he 
expressed doubts as to some other supposed marks of glacial action 
near this city, but on seeing those at Blackford Quarry, he instantly 
exclaimed—* That is the work of the ice.” On the top of Salisgury 
Crags, at a quarry about two hundred yards from their south extremity, 
the polishing is very well seen at intervals over a space of twelve or 
feet just at the edge of the precipice ; and strig, running east 
and west, will also be discovered here by an eye accustomed to observe 
them, though they are much less distinct than at Blackford Hill. In 
quarrying the Crags’at this spot, the rock had been cut back about one 
hundred and twenty feet from what was originally the edge of the pre- 
cipice, and this part, which had been well protected by the turf, was 
only exposed about 1822 or 1823. We have little doubt that similar 
appearances would be presented if other parts of the — equally 
distant from the edge of the pigeons were any laid 
north end of the Castle y polished, aa the groovings 
on the western sited of Corstorphine Hill, described many years ago by 
Sir James Hall, are well known. We have observed similar marks of 
abrasion at Craigleith Quarry, Craigmillar Hill, and elsewhere. . 
_. These marks of abrasion, both on rocks in situ, and on boulders 
found in the soil, have been usually attributed to the action of the cur- 
rents of water, rolling along stones and gravel, an explanation felt not 
to be satisfactory, but adopted for want of a better. It is admitted that 
