FRIDTJOF NANSEN. 



M.-N. Kl. 



changing direction shows a striking disagreement with the directions obser- 

 ved both before and after that hour. It is obvious that under such circum- 

 stances it is impossible to construct the probable curve of movement with- 

 out having numerous and frequent observations. Our curve in Fig. 33 must 

 thus be considered as quite unrealiable, and it this therefore difficult to 

 discuss the motion at this depth. 



20 TT 



N 



200 cm./ sec-. 



Fig. 45. Progressive Vector Diagrams of the Relative Movements at 10 Metres and the other 

 depths of Fig. 44, between Aug. i8th 3 p. m. and ipth 2 p. m. Stat. 41. 



By using those parts of the curves of velocity and direction which 

 should be most reliable, owing to a greater number of observations, I have 

 attempted to construct a progressive vector diagram for the relative move- 

 ments at 10 metres, between August i8th 3.00 p. m., and August igth 2.00 

 p. m., and have compared it with the progressive vector diagram for the 

 deepest water-layers (/'. e. presumably representing the drift of the ship) 

 during the same period (see Fig. 45). 



The progressive vector diagrams of the relative movements give us 

 (in the same manner as described above p. 55) the following values of the 

 real currents between August i8th 3 p. m. and August igth 2 p. m. : 



