- No. 2. SPITSBERGEN WATERS. 55 



the whole. At the greatest depths, between 300 and 470 metres, 31 ob- 

 servations were taken in 40 hours; and the curve for these depths should 

 consequently be the most trustworthy. At TO metres there are 12 obser- 

 vations in 34 hours, but they show great irregularities, and the curve of 

 the vector-diagram has a very improbable shape. The vector-diagram for 

 200 metres, based on only 3 observations, is naturally very uncertain, but 

 may give an idea. 



On the whole, the curves of these vector-diagrams show a tendency 

 towards forming more or less irregular ellipses, which, e. g. at 20 metres, 

 approach circles. The directions of the long axes of the ellipses may shift, 

 but seem to have a tendency towards NW and SE or WNW and ESE. 



By means of the directions and velocities of the relative movements 

 for each hour, given by our curves in Figs. 36 & 37, the progressive vector- 

 diagrams, Fig. 44, have been constructed. They show the directions 

 towards which the water at each depth of observation was moving in 

 relation to the ship or the ice-floe. 



If we were to assume that the actual movements of the water at the 

 greatest depths of observation were so insignificant, that they might be 

 left out of consideration, the indications of the current-meter at these depths 

 give the direction and velocity of the drift of the ice-floe and the ship. 

 The progressive vector-diagram representing the movement of the deepest 

 water in relation to the ship, consequently gives the drift of the ship 

 or the ice-floe, if we turn it round so that North becomes South and 

 East West. 



During the period of observation the ship should thus have drifted 

 from the end of the progressive vector-diagram to the right to its starting 

 point to the left, i.e. towards S66W (S78W magnetic) with an aver- 

 age velocity of about 12.7 cm. per second (taking the observations between 

 Aug. i8th 2 a. m. and Aug. igth noon). 



But during the same time the water at the other depths would have 

 moved along their progressive vector-diagrams in relation to the ship. 

 If, therefore, we draw lines from any point of the first vector-diagram to 

 the corresponding points (/. e. corresponding to the same hours) of the 

 others, these lines give the distance and direction of the active movement 

 at the different depths during the period that ha* elapsed from the be- 

 ginning of the observations till the moment chosen. 



We thus find that from August i8th 2 a. m. till August iQth at noon, 

 the water at 20 metres should have moved towards N6W (N6E mag- 

 netic) with an average velocity of about 7.8 cm. per second. 



