THE GEOLOGY OF THE FAROE ISLANDS. 259 
This heightened cascade would in like manner gradually retreat up the valley, 
and perhaps would merge with others before it finally reached the steeper 
slopes in which the stream originated. The retrocession of the rock-walls 
however, would not be entirely due to the action of the stream—for the under- 
mining process would be carried on at the same time across the whole breadth 
of the valley by the action of springs and frost. In short, the rock-face would 
recede up the valley in the very same manner as the loftier lateral cliffs, 
between which the water flowed. The Ferde Islands afford numerous examples 
of every stage in this kind of valley-formation. In Kolfaredal we find the 
valley-bottom rising from the sea-level with a gentle acclivity to the watershed 
—the rock-walls and cascades have disappeared. In other valleys, again, we 
have the bed rising with the same low inclination until it is hemmed in, as it 
were, by a rock-wall over which the water pours from an upper platform. This 
latter in like manner slopes gently upwards until it is suddenly cut off by a 
similar rock-wall, above which a third flat-bottomed course succeeds, terminated 
by another steep face of rock, and so on. The steeper the gradient of a valley 
the more numerous do the transverse cliffs and waterfalls become, which of 
course is only what might be expected from the comparative horizontality of 
the strata. 
Valleys, excavated in the manner described, are necessarily more or less 
cirque-shaped at the head, and similar but usually smaller lateral cirques open 
upon them at various levels throughout their course. These lateral cirques 
have been formed by the locally rapid recession of the valley-cliffs. Here and 
there some particular bed, perhaps pretty high up in the cliff, is more quickly 
undermined than the strata below it, and the upper section of the cliff therefore 
tends to recede more rapidly than the under. In this way small lateral cirques 
are formed which collect the tribute of the springs and send down their cascades 
to the main stream. 
6. Former Greater Rainfall—N ow the recession of all those rock-walls and 
cliffs results in the production of enormous quantities of débris, the accumulation 
of which, if it be permitted, must in the course of time bank-up the precipices and 
retard their waste. And this is precisely what is taking place. The streams are 
unable to carry away all the loose material which is brought down the slopes by 
the action of springs and frost. It is evident, therefore, that the time must arrive 
when these slopes will become more or less masked or concealed under their own 
ruins. Even now one may see the process far advanced in many mountain-valleys 
—great curtains of débris hanging from the higher parts of the hills, and sweep- 
ing in long trains down to the low grounds, where they gather upon the bottoms 
of the valleys, and are often left undisturbed by the streams, confined as so many 
of these are to narrow post-glacial trenches. It is clear, then, that there must 
have been a time when the rainfall in the Feerdée Islands was more considerable 
