252 DR JAMES GEIKIE ON 
Storevatn evidently owes its origin to the grinding action of the ice that 
flowed from the heights between Qvalvig and Leinum. This ice was deflected 
by the mass of the hill called Saaten, and would necessarily erode a hollow 
where the opposition to its flow was greatest. I think it is also highly probable 
that in late glacial times this hollow, which had been excavated at a period 
when the ice was thickest, would be occupied by a local glacier. Mjavatn 
appears to be likewise due primarily to the excavating action of the ice-sheet. 
It lies quite close to the low co/ or water-parting at the head of Kolfaredal, 
which, when the ice first began to stream down the slopes of the island, was 
probably a more marked feature than it is now. The water-parting may then 
have formed a rocky barrier against which the ice that flowed across into 
Kolfaredal would press with great force. Here then another rock-basin would 
be formed, which, however, would tend to become obliterated as the rock-barrier 
continued to be lowered by the grinding of the strong ice-current that passed 
across it. In late glacial times the hollow may have been to some extent 
modified by the action of a small local glacier. 
Another interesting rock-basin is that in the valley of Saxen. The surface of 
this lake is 22 métres or 74 feet above the sea. The valley in which it lies is 
wide, and comparatively flat-bottomed—the position of the lake being shown in 
Plate XV. fig. 21, which is a diagrammatic longitudinal section of the valley. 
The present stream has cut deeply into the old valley-bottom, and now flows 
some 140 feet or so below its general level. The lake-basin appears to have 
been excavated in the bed of the old valley which rises some 60 feet or so above 
the surface of the water. I saw no trace of old water-levels, which would 
indicate a former greater height for the lake. But its borders are strewed with 
so much débris fallen from the heights above, that any such traces might well 
be obliterated. I think it most probable, however, that the lake-surface was 
never much higher than it is, but that the deepening of the outlet went on con- 
temporaneously with the excavation of the rock-basin, and consequently that the 
deep gully in which the stream now flows is not entirely of post-glacial origin. 
It is worthy of note that the water-parting of the valley in which the Saxen 
lake-basin occurs is a mere flat col like that of Kolfaredal. The glacier to which 
it owes its origin did not therefore head in lofty, steeply-inclined valleys and 
corries, but was rather a thick flattened mass of ice that gathered deeply over 
the low-lying co/, and seems to have flowed both to north-west and south-east. 
Besides these larger lakes, which occupy nearly the whole breadth of the 
valleys in which they occur, very many lakelets lie in cirques. The great cirque 
of Howe, for example, is studded with lakelets. I counted upwards of twenty, 
varying in breadth from 50 or 60 feet up to several hundred yards. They are 
all true rock-basins—the bed of the great cirque in which they lie being also 
abundantly covered with well-marked roches moutonnées. In the numerous 
