- No. 2. SPITSBERGEN WATERS. 73 



This might seem to be in fair harmony with our theory. If the velo- 

 city of the current decreases downwards, its direction should be more and 

 more deflected towards the right with increasing depth; but if, for some 

 reason or other (e. g. by contrary winds), the velocities in the upper layers 

 of the current are reduced, its deflection with increasing depth is dimin- 

 ished accordingly; and if the surface- velocities be reduced to values below 

 the velocities of the deeper layers, the deflection with increasing depth may be 

 reversed. Our observations, especially in the upper layers between the 

 surface and 20 metres, seem thus to verify the correctness of our theory, 

 as well as could be expected with the imperfect methods. 



The water at 50 metres was on the whole colder at Stats. 20 and 

 19 than at Stat. 41; at Stat. 19 its temperature even sank to i.5iC. 

 and its salinity was then as low as 34.38 / 0o . This water must naturally 

 have come from the north or north-west. 



Computation of the Velocities of the Currents at Stats. 18 to 22. 



The sections through Stations 18 and 19 (Fig. 13) and through Sta- 

 tions 20 and 21 (Fig. 14) may be used for computation of the velocities 

 of the currents in this region. The direction of the former section is about 

 Si8E, and of the latter section about Si5E. The average direction 

 of the surface- current at Stat. 20 was towards S3iW according to the 

 current-measurements (Fig. 64). 



In the section through Stats. 18 and 19, all isopycnals at depths be- 

 tween the surface and about 90 metres, slope from Stat. 18 towards Stat. 

 19 (see the isopycnals of 27.70 and 27.80, Fig. 13) indicating that the cur- 

 rents at these depths had a component directed transversally to the section, 

 i.e. towards S7oW. This agrees well with the results of our current- 

 measurements. At depths greater than 90 metres the isopycnals were 

 sloping gently from Stat. 19 towards Stat. 18 (see the isopycnal of 27.90, 

 Fig. 13) indicating that the current at the these depths had a very small 

 component directed transversally to the section in the opposite direction 

 of that of the current at higher levels. 



At Stat. 19 several vertical series of observations vere taken, giving 

 somewhat different densities; but all of them were tajten a long time after 

 the observations at Stat. 18. The two most complete series are those 

 called 19 (taken on Aug. 3rd, 6.30 to 7.50 p.m.) and 19 c (taken on Aug. 

 4th, ii a. m. to 0.40 p.m.). The former series was taken about 8 hours 

 after the observations at Stat. 18, and the latter series about 24 hours 

 after these observations. 



