66 FRIDTJOF NANSEN. M.-N. Kl. 



tion and velocity of the currents in their details, it would be necessary to 

 have observations at a great many depths, with much shorter intervals 

 vertically, than is the case in our series of observations. 



Though our observations are not sufficient to give in detail the changes 

 in the currents at the different depths, they prove at any rate that the 

 movements of the water in the currents of the sea are much more com- 

 plicated than was generally assumed. 



Computation of the Velocities of the Currents in the Region between 

 Stats. 41 and 40, by means of Dynamic Sections. 



It would be of much interest to examine to what extent the velocities 

 of the currents at different depths found by our measurements, may agree 

 with the corresponding values of velocity computed by means of dynamic 

 sections. It has, however, to be considered, on the one hand, that the 

 computations by sections should give the average velocity of the water at 

 each depth in the whole region between every two stations, while the 

 current-measurements give the velocities at one particular station, and there 

 may be many local variations in the velocities, owing to vortex move- 

 ments, etc. 



On the other hand, our current-measurements prove that the velocities 

 and directions of the currents change very quickly from hour to hour, 

 while the sections will not, as a rule be representative for any special 

 moment, because of the intervals of at least some hours, that generally 

 will exist between the observations taken at two different stations af a 

 section. 



Moreover the computation of the velocities by means of dynamic 

 sections is based upon the supposition that there is perfect lateral equilib- 

 rium in the sections; but with such rapidly changing currents it is not 

 probable that such an equilibrium can ever be attained. 



In addition to this, there may also be other agents producing tempo- 

 rary oscillations of the layers at the stations. Hence it follows that we 

 cannot, even under favourable circumstances, expect to find a ver}' close 

 agreemeent between the values of velocity found by measurements and 

 those found by dynamic computation. 



In our case we have the two Stations 40 and 41 a, the latter situated 

 perhaps a little more than 3.5 miles, or about 7 kilometres, towards the 

 north from the former. The observations at the two stations were taken 

 at an interval of 3 hours and 30 minutes. At Stat. 40 no observations 



