I92I. No. II. THE STRAXDKLAT AND ISOSTASY. 277 



the Finnish coasts and round the .Mand Islands have certainly a great 

 resemblance to the "Skjærgard" of the Norwegian strandfiat, but it has 

 obviously not the same kind of horizontal planes, and is more like a low 

 flat lanrl wh.ich has been submerged. As has been pointed out by A. G. 

 Högbom [1910a], this low plain probably represents a precambrian plain 

 of denudation which is widely extended in Sweden, and may be seen 

 continuing in under cambrian-silurian deposits. 



As long as we do not know the level of the strandfiat we have hardly 

 any means of determining to what extent the rising land surface of this 

 region has approached the level which it had bel'ore the last glacial sub- 

 mergence. 



As the land is still rising, the strandfiat. if there is one, may not yet 

 have been elevated above the sea. 



In my opinion there may have been quite favoural)le conditions for 

 the planing of a strandfiat in solid rock, especially along the coast of 

 Bottenviken, during the cold periods of the interglacial epochs. It is true 

 that the wave action cannot have been very effective in this enclosed sea, 

 and there was practically no tide. There was, however, a verv severe 

 climate in this northern region, favouring an active shore-erosion by frost, 

 which also was much increased by the fact that the surface lavers of the 

 sea consisted very nearly of fresh water, which bv freezing has a much 

 greater disintegrating effect on the rocks than sea-water. The winds would 

 also cause frequent changes in the level of the sea, which to some extent 

 woulrl make up for the lack of tide, and the wave-action would be sufficient 

 to render substantial help in transporting the débris of the shore-erosion. 



There is, therefore, good reason to expect that a strandfiat, cut in 

 solifl rock, may actually exist at some level below present sea-level. In 

 that case it might be possible to trace it on the sea-floor. 



Through the kindness of Prof. A. Högbom I have obtained a chart 

 with soundings of Bottenviken (Chart No. 20 of the Swedish geodetic 

 Survey). Tlie soundings are not numerous enough for a studv of the 

 detailed topography of the sea-bottom. 



By drawing contour lines for every ten metres of depth, a great 

 number of sharply defined submerged valle\s and channels become con- 

 spicuous. They form, as a rule, continuations of the bavs and of the 

 valleys on land. Some of them are quite narrow, with fairly steep side 

 slopes. In Fig. 169 a small portion of this chart is reproduced as an 

 example. They often seem to descend to depths of 60 and even 70 metres; 

 but as they have probably been deepened by glacial erosion it is hardly 

 possible to decide what their base-level may have been. In several places 

 there are hollows with somewhat higher ridges outside indicatijig ex- 

 cavation by glaciers. 



Between these submerged valleys there are shelves or platforms, 

 extending as much as 40 kilometres, or even more, from the coast. 



