THE GEOLOGY OF THE FAROE ISLANDS. ‘253 
smaller cirques that occur in Suderée and the northern islands, lakes are of 
almost invariable occurrence. Now and again, however, such lakes have 
become silted up, and are replaced either by flats of alluvial detritus and peat- 
moss, or by quantities of rough débris which have tumbled from the surround- 
ing precipices. All these cirque-lakes are unquestionably of glacial origin—the 
cirques themselves having been originally formed by the action of springs and 
frost, and subsequently deepened and excavated by small local glaciers and 
glaciers remaniés, as I shall describe more fully in the sequel. 
The rock-basins visible in the islands themselves have thus been excavated 
under varying conditions. Some, as we have seen, have been hollowed out 
when the ice was at its thickest. They belong to the period of general glacia- 
tion, while many are due to local glacial action, which has likewise in some 
cases modified the results produced by the erosion of the ice-sheet. In the 
great amphitheatric cirque of Howe, we see how innumerable small rock- 
hollows may be scooped out by the action of one and the same ice-flow. 
Doubtless, most of these little basins owe their origin to some accidental 
circumstance. In some cases, perhaps, the rock has yielded unequally, owing 
to more abundant jointing or to differences in mineralogical composition and 
petrological structure. In other cases the basins may simply indicate localities 
where the ice, owing to the form of the ground, was enabled to exercise greater 
intensity of erosion. 
VI. ORIGIN OF THE VALLEYS AND FIoRDS: SUBAERIAL AND GLACIAL EROSION, 
1. Forms of Valleys.—I have already made brief reference to the various 
forms assumed by the valleys, and must now describe these a little more fully. 
For this purpose I shall select one or two examples which may be taken as. 
typical of all the others. 
The great cirque-valley of Howe in Suderée is one of the finest to be met 
with in any of the islands. It opens upon Howe Bay on the east coast of 
Suderée with a breadth of nearly one mile. Its bottom is gently undulating, 
being studded with clustering roches moutonnées. Following the stream 
upwards we are suddenly brought to a cliff or wall of rock at a distance of 
rather more than a mile from the sea. This wall circles round the valley, and 
appears to form its head. But when it is surmounted we find a second broad 
cirque-like valley bounded in like manner by steep cliffs, above which other 
cliffs rapidly succeed, rising tier above tier to the summits of the bare rugged 
mountains. This upper cirque-valley is wider than the lower one, and its 
bottom is covered with roches moutonnées, in the hollows amongst which occur 
the numerous lakelets of which I have already spoken. What chiefly im- 
presses one is the great width of the valley relative to its length. From the 
sea to the head of the upper cirque is just some three miles ; yet the width of 
