92 FRIDTJOF NANSEN. M.-N. KI. 



If the shortest distance between Bennett Island and the edge of the con- 

 tinental shelf to the north of it, is much shorter than the distance between 

 the island and the region of the Fram in the autumn of 1893, ^he tidal 

 wave may spend less time c. g. 2 hours less in reaching the island. Its 

 tidal hour might then be 22h 50m, or loh 25m. But this would be about 

 ^h 48™ later than found by the observations of 1881 on the island, — 

 or also would it be 8h 37m earlier, but in that case we would have to 

 assume that the tidal wave reaches Bennett Island one semidiurnal period 

 later than given by Harris's chart, and that seems hardly possible. 



At present it seems difficult to find a feasible explanation of these 

 great discrepancies. 



The Extension and Shape of the North Polar Basin. 



As the results of the investigations described in this paper, especially 

 the salinity of the deep-water of the North Polar Basin, may have some 

 modifying effect upon the views previously held by the writer, as to the 

 extension and shape of the deep North Polar Basin, it might be appropriate 

 to take up this subject for new discussion here. 



In 1902 and 1904 the writer held the view that the deep North Polar 

 Basin had a wide extension, and probably covered the greater part of the 

 still unknown North Polar region [1902, pp. 399 et seq.; 1904 pp. 227 etseq]. 

 This view was chiell}' based upon the nature of the cold deep-water or 

 bottom-water of this basin. As, according to the investigations during the 

 Fram-Expedition, this water should differ entirely from the deep-water of 

 the Norwegian Sea, it was thought to be formed in the North Polar Basin 

 itself; but if so, this formation seemed impossible to account for unless the 

 basin had a wide e.xtension. 



After having found by new investigations that, by a special process, 

 cold bottom-water with a comparatively very high salinity may be formed 

 in the Barents Sea [1906], the writers views as to the extension of the 

 North Polar Basin were much modified [1907], and he no more considered 

 it necessary to assume that the North Polar Basin had such a wide exten- 

 sion as he had thought before. His view was now, "that the known geo- 

 graphical and geomorphological features of the Arctic regions do not 

 exclude the possibility of a wide extension of the continental shelf be^'ond 

 the northernmost known islands of the American Arctic archipelago and 

 Greenland, and there may be lands on this shelf in the Unknown North. It 

 is also possible that the Siberian continental shelf may have northward exten- 

 sions with land in the region between the New Siberian islands and Alaska". 



