MoHS (1 87], p. 91) reports observations of temperature from 

 Lodingen, which show that at a depth of 100 fathoms the tempe- 

 rature does not vary throughout the year (6". 4 to 6°. 5). 



There are thus along the entire Norwegian seaboard of the 

 North Sea and Northern Ocean, 3 deep regions of a very well 

 marked hydrographical character, viz., 



1. The Eegion of periodical changes, to a depth of 200 to 

 250 metres. 



2. The Atlantic Eegion, to a depth of about .500 metres. 

 2. The Arctic Eegion. 



Of these regions, it is chiefly the uppermost that is of interest 

 to us, as it is our main purpose to unravel all the conditions which 

 may influence the niigration of fishes. And it may well be pre- 

 sumed that the great changes produced by currents, by summer 

 warmth and winter cold, and the variations from year to year of 

 the different factors, may be of the greatest importance to the 

 periodical fisheries. 



In order to investigate these conditions, Hjoht, in 1895, made 

 5 sections of the sea off the coast between Stavanger and Lofoten, 

 and through the assistance of the authorities of Bergens Museum, 

 NoEDGAABD was enabled to go over the same section during the 

 winter of 1896. 



The results of these investigations are given on Pis. 6 and 7. 

 The salinity is represented in similar tones to those used by Hjobt 

 in his former work [95]. 



Water, the salinity of which exceeds 



35 %() is denoted by dark blue colouring. 

 34 to 35 - - „ - hght „ — 



33 - 34 - - „ - dark green — 

 32 - 33 - - „ - light „ 



less than 32 - - „ - yellow — 



This method of representation was founded by the Swedish 

 hydrographers Pettersson and Ekman [91]. That this is right is 

 shown by the fact that the lines of salinity, in many instances, 

 represent the natural boundaries between layers of different origin 

 and character; each of such layers generally possessing its own 

 characteristic conditions of temperature. 



This method of division has been tested during several years 

 investigations in the Skagerak, where Petteksson and Ekman separate 

 and recognise the following layers of water : 



Atlantic Water, Salinity 35 "/oo and upwards. 



North Sea „ — 34 to 35 - 



Bank „ — 32 - 34 - 



Baltic „ less than 32 - 



On the Norwegian seaboard of the Northern Ocean, the waters 

 of the Gulf Stream can be discerned from the coastal layers by 

 the 35 7oo line of salinity. As will be seen from the tables, there 

 is often a great variation in temperature along this line. But 

 naturally all kinds of changes are to be met with, and in many 

 instances it would, possibly, have been more, or equally correct to 



fix the boundary, for instance, at 34.8 %o. But, practically, this 

 would make but little difference, and therefore we deem it prefer- 

 able to use single lines throughout, in order that the figures may 

 retain their value independent of theories and individual opinions. 



We regard it as doubtful whether the coastal water can, in 

 the Northern Ocean, be divided into such naturally confined layers 

 as in the Skagerak, but this must be more accurately investigated. 

 On practical grounds we have retained the curves and the same 

 colouring as for the Skagerak waters, but without intending thereby 

 to indicate the various sources of the coastal waters. 



For the sake of brevity, we will in the following few words 

 denote those waters whose saline contents amount from 34 to 35 "/oo, 

 as "34 water", and, in like manner, describe the other degrees of 

 salinity. 



The sections given on Pis. 6 and 7, confirm the correctness 

 of the results given on PL 5, Figs. 3 and 4. 



In summer the 35 water rises well up towards the surface, 

 being surmounted by coastal layers of various degrees of salinity. 

 On the Norwegian west coast (PI. 6, 1 a and 2 a) there still remains 

 a ([uantity of water of less salinity than 32 "/(lo? probably of Baltic 

 origin. J^irther North, coastal water is met with, consisting almost 

 exclusively of 33 and 34 water. The temperature in the coastal 

 layers decreases, evenly, from the surface downwards, but the At- 

 lantic water occasionally exhibits a rise of temperature towards the 

 bottom. The coastal layers are thus in summer of little thickness, 

 but their salinity is often somewhat low. 



If we examine the winter sections we find the conditions en- 

 tirely altered. The line representing 35 %o salinity has on the 

 shores of the Nortliern Ocean (Table 6, Figs. 3 b and 4 b. Table 7, 

 Fig. 1 b) retired to over 100 metres of depth, in parts to more 

 than 200 metres. 



On the shores of tlie Northern Ocean the 35 water will be 

 found somewhat nearer the surface (PI. 6, Figs. 1 b and 2 b). 



Along the entire coast at that season, there will be found a 

 thick layer of 34 water, and also of 33 water, which partly cover 

 the coastal banks. 



The temperature of the 34 water is not much lower than that 

 of the Atlantic water, generally from 5" to 7°, falling seldom to 

 4", but, on the other hand, the temperature of the 33 water may 

 be 3", or less than that. 



It is necessary now to discover the origin of these thick layers 

 of water, and there are two theories respecting that, which may be 

 taken into consideration, viz., that they are either formed at the 

 spot by a mixture of river and Atlantic water, or that they are 

 brought by currents from other parts. 



The Swedish hydrographers are of opinion that, a combination 

 of layers of water of different specific gravities does not occur, to 

 any great extent, except in shallow seas. 



In that case, the 34 water on the Norwegian coast could only 

 originate from places where a vast quantity of water, of a salinity 

 of 34 to 35 'Voo, exists. The Sw-edish hydrographers therefore 

 assume that the 34 water must proceed from the Arctic Ocean 

 itself, or the north western portion of the Atlantic Ocean. In 

 summer there is not any connection between the Arctic Ocean and 

 the shores of Northern Europa, but PeoDteesson and Ekman have 



