82 FRIDTJOF NANSEN. M.-N. KL 



According to our vector diagrams Figs. 46 49, the movements of the 

 water may have been approximately the following: 



at 20 Metres at 50 Metres At 100 Metres 



Aug. 18, 1.15 a.m. 5cm./sec.(?) 15 cm./sec.(?) 15 cm./sec.(?) 



towards S 20 W (?) towards N i o E towards N (?) 



Aug. 19. 10. oo a.m. 2 cm. /sec. 7 cm./sec. 3 cm. /sec. 



towards N 4 W towards S 24 E towards N 44 W 



Aug. 19, 3.20 p.m. 10 cm./sec. 3 cm.sec./ 10 cm./sec. (?) 



towards N 80 W (?) towards S 30 W towards S 60 E (?) 



Let us assume that the prevailing current of the layers between 20 and 

 50 metres, at Stat. 41, was running on the average in a north-easterly 

 direction more or less parallel to the average direction of the edge of the 

 continental shelf, as indicated by our current charts, Figs. 66 and 67, and 

 that Stat. 41 was situated nearest the right hand side of that current. It 

 is then clear that the movements found at 10 a. m. on Aug. i9th, would 

 have a tendency to lift the water-layers at 50 metres etc. because the 

 water at this depth was then moving towards the right hand side (towards 

 S24E) of the prevailing current while the water at 20 metres was moving 

 nearly in the opposite direction. 



At 1.15 a. m., on Aug. i8th, the water-layers at 50 metres ought to 

 have been depressed if the water at 20 metres was flowing to wards S 20 W, 

 or to some extent towards the right hand side of the prevailing current, 

 while the water at 50 and 100 metres was flowing in the opposite direc- 

 tion with comparatively great velocities. 



If at 3.20 p. m v on Aug. i9th, the water at 20 and 50 metres had 

 westerly movements with greater velocities at 20 than at 50 metres and 

 the water at 100 metres had an easterly movement, these movements would 

 naturally tend to lift the layers, but not as much as the movements at 

 10 a. m. on the same day. 



Hence it appears that the vertical movements of the water-strata, indi- 

 cated by our observations at Stat. 41, agree with the vertical oscillations 

 that presumably would be caused by the simultaneous horizontal move- 

 ments, at 20, 50, and 100 metres, computed from our current-measurements. 

 Of course, it has to be considered that the ship was drifting with the ice 

 and the curface-current, on the average towards WSW, and changes in, 

 the water strata may also arise in this manner. 



At Stations 19 and 20 a number of vertical series of temperatures 

 and salinities were taken, but the current-measurements are so imperfect 

 that an analysis of the relation between the vertical oscillations and the 



