Nansen's Continental Oscillations — Spencer. 223 
typical valleys in Kara sea, and these may have been left 
open on account of their great distance from the Siberian 
or other rivers. The soundings in Barentz and Kara seas 
are "sufficient to show that the physical features of the 
bottom bear a striking resemblance to a former sub-aerial 
region"" as there is an "extensive coherent system of broad 
submarine valleys * * * with many tributaries drain- 
ing an extensive area of the ancient continent as far east 
as Novaya Zemlia. * * * The whole bottom seems to 
have been elevated above sea level * * * at some pre- 
vious and probably not very remote geological period * * * 
No other theory would in my opinion serve to explain the 
origin and presence of such a system of valleys under the 
water of the sea. * * * Valleys may be formed by faults 
or folding * * *. But we cannot expect the depres- 
sions thus formed to assume, without the aid of meteoric 
agencies the shape of typical river valleys with numerous 
branches or tributaries or converging into a trunk valley, 
which on the whole gradually deepens towards its em- 
bouchure opening out into the oceanic abyss." (p. 26.) He 
further points out that if the ridges between the valleys in 
this region have been formed by seismic movements they 
must have been dissected by erosion of some kind, for all 
the transverse valleys could not have been formed by tec- 
tonic strains, as they exhibit typical features of river- 
formed channels, afterwards occupied by glaciers. These 
features on the shelf required an elevation of 450 metres 
above sea level and probable much more, and he regrets 
our lack of information which prevents our defining the 
amount. He also shows that the channels, as that at 
Vardo, have not been excavated by submarine glacial ero- 
sion. These valleys were formed at a time just preceding 
the Glacial epoch. Then they were occupied by glaciers, 
which deposited detritus in the valleys, as they were reced- 
ing during the subsequent subsidence of an inter-Glacial 
epoch, which culminated when the sea had reached 100 
metres above the present level. Then recurred another 
elevation, followed by the latest Glacial epoch. The val- 
leys were partially re-excavated, and according to Ramsay 
the glaciers did not reach the coast. A second submerg- 
