254 DR JAMES GEIKIE ON 
the valley at its upper termination, at the base of Borgaknappan, is nearly one 
mile and a half (see Plate XIII. fig. 1). 
The geologist, crossing from Saxen to Tidrnevig in Stromée, will traverse 
two very characteristic valleys. The stream which comes from the north 
falls into Saxenfiord down a succession of steep cliffs. When we ascend these 
cliffs, which are perhaps 400 or 500 feet high, we find ourselves in a flat- 
bottomed valley bounded by encircling cliffs of dolerite, above and beyond which 
another similar but shorter cirque-valley appears. Between this upper plateau 
and the valley of Tidrnevig the water-parting is reached at a height of 1693 
feet (516 métres). From this point the ground descends rapidly towards the 
‘north-east into another cirque-shaped valley, the sides of which consist of a 
succession of narrow plateaux, so that the stream descends by leaps from one 
stage to another till it reaches the sea at Tidrnevig (see Plate XIII. fig. 2). 
Similar features characterise nearly all the valleys. They descend in a 
series of platforms, which vary in number, breadth, and relative height, but 
invariably present the same features. Those I have now described agree in 
this respect that they all have well-defined water-partings ; to get from one 
valley over into the other we must ascend and descend several hundred feet. 
But there are other valleys, the heads of which coalesce, as it were, so that we 
pass from the one into the other over a low flattened co/ or water-parting. 
Such valleys form the great hollows which I described in a previous section 
as crossing some of the islands from sea to sea. The lofty rock-barriers which 
must at one time have separated the heads of these valleys have been demo- 
lished. The measurements we obtained in Kolfaredal will illustrate the 
general character of those remarkable hollows. From Kollefiord we found 
that the valley rose to the water-parting (259 feet) with a mean inclination of 
13° or 14°. The water-parting itself is low and flat, and it was hard to dis- 
tinguish any culminating point. The descent on the other side of the water- 
parting is very gentle, the fall being only some 50 feet or so for the first two 
miles. After this the sea is reached at Leinum in less than a mile—the fall 
being of course more than 200 feet in this short distance. 
2. Fiords.—The soundings upon the chart prove that the long fiord which 
separates Stromée from Osterée, occupies the bed of two submerged valleys 
with a low separating co/, over which there is shallow water. This col occurs 
in the narrow part of the sound between Nordskaale and Ore, and the sound- 
ings show that from this point the water deepens both towards north-west and 
south-east. The fiord is shallower at its mouth near Eide, where there are 54 
and 9 fathoms of water, than it is at and above Haldervig, where we get depths 
of 18 to 30 fathoms. The southern part of the fiord has not been sounded, but 
it is probably the deeper of the two sections. Many of the other sounds 
between the islands are apparently of the same nature as that just described— 
