278 KKIDTJOF NANSr.N. M.-X. Kl. 



""J'o a /^rcal cxlcnl llicsc i)lat forms liavc (Icptlis Ijclwcen 10 aiul 20 metres. 

 'J'iiey arc ImhukUmI hy fairly sleep side skjpes, flescenrjing' to depths of 

 more llian Ho nn'tres (\'\^. i(^9), and, as it seems, by well defined ed;^es, 

 wliicli aloiii^- iJKJr outer mar^i^^in appear to be between 20 and 30 metres 

 I.elow sca-le\cl ; but llie sotuidin^s arc tocj scattcrcf] û>r an exact tracii'g 

 of tlicm. 



Tlic U(/rtli-caslcrii pari of 1 loltcnxikcn is very shallow with an ex- 

 tremely flat bottom, at ilcpths mostly between 10 and 20 metres, extending 

 60 to 70 kilometres from the coast and being- bounrled b\- a steeper slope 

 outsirle, desceufling- to (le])tlis of 80 and 100 metres. There are not suf- 

 ficient soundinj^s to trace the topof,rraphical features of this flat floor; 

 but it seems to be traversed to some extent by drowned valleys and 

 channels. Some of them forming hollows 25 to 30 metres deep near the 

 coast with somewhat hif^her platforms outside. 



To what extent these submerged platforms of Pjotlenviken are built 

 up of waste and glacial drift is not easv to decide. The narrow rlrowned 

 valleys and channels, 25 to 30 kilometres long, which the chart indicates, 

 especially along the west coast of Bottenviken, seem, however, difficult 

 to exi)lain unless they are cut in rock, at least to some extent. The 

 glaciers would hardly be able to form such narrow channels by sul)marine 

 erosion in extensive terraces of loose material; and if they had been 

 formerl 1)\' fluvial erosicjn on land before the last sul)mergence. they would 

 have been more or less obliterated by the ice-ca]) unless they were cut 

 in rock. 



The existence of these narrow drowned valleys also make it probable 

 that Bottenviken cannot to anv great extent have lÆen filled up by glacial 

 drift, or l)y waste after the beginning of the last glacial period, or after 

 the last submergence of the land in this region. 



If, however, the drowned valle\s are cut in solid rock, it is also 

 obvious that the platforms between them, with their well marked edges 

 and side slopes, are to a considerable extent cut in solid rock. 



It is then a (juestion whether we here have formations which corre- 

 spond more or less to the strandfiat of the west coast of Norway. If so, 

 we may expect that the coast has still to be elevated 20 to 30 metres 

 before the strandfiat is raised above sea-level, and if the latter shall be 

 raised to levels similar to those of the Norwegian strandfiat a still 

 greater upheaval will be needed. 



In a letter A. G. Högbom has drawn my attention to the interesting 

 fact that at Hernösand (on the west coast of Bottenhafvet) interglacial 

 fresh-water deposits occur at present sea-level [A. G. Högbom, 1909, 

 pp. 578 ff.]. This also seems to prove that during the last interglaciai 

 period the land stood higher than it does now. 



On the other hand it seems p^rolxible that in the southern part of 

 Sweden the land surface has approached its level of isostasy; for the 



