only thing along this track which could be observed as indicating 

 the presence of the western Polar wedge, was a slight decrease of 

 temperature. The 8° curve of temperature, between 17° and 19° 

 West long., runs almost up to the Icelandic coast, but at about 

 15° W. long., turns to the southward, and is only met with again 

 in 6° 20' West long., and 60° 20' N. lat. 



The Arctic currents are therefore definitely stopped in the 

 south; and this also appears to be the case in the east. The 

 results derived from the observations made on board the two sealing 

 vessels on their voyage from Utsire to Jan Mayen, shew that, off 

 Stadt, the surface of the Gulf Stream is over three degrees of 

 longitude in breadth. It is narrowest in about 65° N. lat., where 

 a combination of the surface layers from the west and east cannot 

 be regarded as entirely excluded. In any case this might be due 

 to a hitherto unknown eastward-flowing side branch of the Polar 

 current. 



We do not, however, regard such a combination as likely, 

 and if such really takes place, it is of no material consequence. 

 On the one hand, namely, great bodies of water are to be met 

 with to the northward, of high salinity, and on, the other hand, the 

 Norwegian coastal layers of a like salinity as the Polar current 

 (34 — 35 %o) possess a higher temperature than the Arctic waters. 



On the coast of Nordland, the 34 water is of a warmth, through- 

 out, of 5° to 6°, even during winter, whilst the temperature of the 

 Arctic waters declines rapidly as they retire from the Gulf Stream. 



As it now may be considered proved that, in March, no branch 

 of the Polar current traverses the Gulf Stream, it is not probable 

 that this can happen at any other season of the year. The obser- 

 vations made by the Danish Meteorological Institution (Rot.he, [96, 

 97]) shew that February and March are the coldest months of the 

 year between the Shetland Islands and Iceland, the Polar current 

 having had time throughout the winter to gain in strength and 

 thickness, whilst the Gulf Stream, probably from the autunm till 

 then, has evenly diminished. 



According to the accounts of the Swedish scientists, the Gulf 

 Stream ought to be penetrated in December, but during, that 

 month, a pretty high temperature still prevails in the sea within 

 that area. 



What then becomes of the cold bodies of water whicli move 

 in a SSE direction towards the North Sea? 



As we have seen that the Arctic water wedge, west of the 

 Faroe Islands, is undoubtedly absorbed by the warm Atlantic water, 

 it is probable that the eastern branch meets with a similar fate. 

 It must become mixed with the north eastern offshoot of the Gulf 

 Stream by which its (the Gulf Stream's) temperature and salinity 

 is lowered. Thereby it is, indirectly, of great consequence, but the 

 Gulf Stream is, however, so paramount, that the mixed currents 

 are of a well defined Atlantic nature, with a salinity exceeding 

 35 %o, and a temperature above 5°. The body of water then 

 moves on, partly in a north easterly direction towards the Norwegian 

 coast, and, partly, to the southward into the North Sea. This will 



explain the drift of Ryder's bottle from Iceland to Heligoland 

 without its having been necessary for Polar water, with its Arctic 

 conditions to flow across the North Sea. 



By this means, too, Arctic organisms may well enter the North 

 Sea and Skagerak, conveyed by the Atlantic water which at that 

 time of the year is of a temperature of 5° to 6°. High Arctic 

 animals, such as Cl/ione limacina would not be killed by such 

 wai'mth, as that is not greater than the Summer temperature in 

 many parts of the Polar current (see PI. 2). It is even probable 

 that the rise of temperature would be favourable, as the Arctic 

 waters are, in winter, exeedingly destitute of organic life. 



Neither is it impossible that such sensitive organisms as the 

 Arctic diatoms (T halassiothrix, Coscinodiscus) are, in the combina- 

 tion, transferred to the Atlantic waters, but this supposition is not 

 necessary as they are nevertheless met with in small members in 

 most oceanic layers, as will be more clearly shown in our next 

 chapter. 



Tlie hydrographic results liitherto attained, give thus, in our 

 opinion, no ground for assuming tliat the Polar current at any 

 period of the year traverses the Gulf Stream itself, even by means 

 of a thin superficial layer. 



Pettbrsson and Ekman also acknowledge ([97], p. 30) that 

 "In any case, the hydrographical grounds for this assumption can- 

 not be decisive, unless otlier facts in support of the hypothesis be 

 found in the results of the biological investigations". 



The biological conclusions are, however, less tenable than the 

 hydrographical, as we shall subsequently shew. Tlie cold layers of 

 water met with on the coasts of Norway during winter must, there- 

 fore, originate in some other manner. 



A glance at the Summer Surface Chart of the Northern Ocean 

 Expedition (PI. 2), and our Winter Chart (PI. 3) will apparently 

 convince one that the changes in Norwegian coastal waters can 

 only arise from two factors, viz., the Atlantic water on the one 

 part ; the continent and atmosphere, the water from rivers, the 

 warmth of the sun in summer and the cold of the winter on the 

 other. Is it, however, possible by the aid of tliese factors to 

 explain the changes which, as al)ove indicated, are produced by 

 the seasons in the Ocean, and in the distribution of organisms? 



If one studies PI. 7, Fig. 1 a, which represents the conditions 

 of the West Fjord, Lofoten, early in the summer, one will perceive 

 that the salinity then is very great in the surface water of the 

 West Fjord, and that only in towards the JJjords (Salten Fjord) 

 are great bodies of water met with of lower salinity. Hjort's 

 investigations during the summer and autumn 1895, shew that the 

 bulk of these numerous fresher layers increased greatly during the 

 summer and autumn; see, for instance the section from Hero in 

 August, PI. 7, Fig. 6. The chief changes in the distribution of 

 salinity and temperature in deep water, occur during tlie month 

 of September, We append here two tables of observations on the 

 salinity and temperature at various depths off the Nordland coast, 

 one for July, the other for September. 



