The sequence of observations on C and 

 D indicated the rotation to be both clock- 

 wise and counterclockwise. Above the ther- 

 mocline on both stations the rotation was 

 counterclockwise while below the thermocline 

 the rotation vacillated between clockwise 

 and counterclockwise. 



The directions of principal flood and 

 ebb were approximately northeast and south- 

 west, respectively. The variations of each 

 with respect to one another and individually 

 with respect to velocity and direction with 

 greater depth are presented in table 1. 



The rotation period of the tidal cur- 

 rents is significant in that it was semi- 

 diurnal at all observed stations and depths, 

 thereby showing relevance to the semi- 

 diurnal tidal cycle. From the hodographs 

 one can see that the time required for one 

 complete revolution was approximately 13 

 hours . 



The velocities of the currents were 

 less regular. Probably because of shallow 

 depths and a smaller cross-sectional area 

 at the inner stations (B and C) , the greater 

 velocities occurred on these two stations. 

 The maximum, minimum, and average velocities 

 of the tidal currents (tidal phase consid- 

 ered) and the variation of these with depth 

 for the above stations as well as stations 

 A and D are shown in figures 6 to 9. 



AVERAGE CURRENTS 



In the analysis of the original obser- 

 vation data, the average currents were 

 separated from the tidal currents by the 

 method previously described. The results 

 of the analysis indicated a counterclockwise 

 circulation through southeastern Bering Sea 

 (figure 10). On the inshore stations (C 

 and D), the average current tended to flow 

 into southeastern Bering Sea, the flow being 

 more definite on the inner station (C), at 

 average velocities of 0.09 (C) and 0.05 (D) 

 knots with directions of 072"! and OOa^T, 

 respectively. On station D, the average 

 current varied, with depth, between north 

 and east. Because very light winds pre- 

 vailed during the observations, the proba- 

 bility of the currents being caused by the 

 winds is somewhat decreased. The velocities 

 and flow direction of the average current 

 on stations C and D and their variation with 

 depth are illustrated both in table 2 aind 

 in figure 10 (page 9). 



Because of strong winds during the 

 observations on station B, the offshore 

 stations (A and B) had to be considered 

 separately. On station A, the average cur- 

 rents flowed essentially southwest at an 

 average velocity of 0.05 knot and an aver- 

 age direction of 223''T. The variations in 

 the current velocity and direction with 

 respect to depth for stations A and B are 

 also shown in table 2. 



On Station B, where the wind tended to 

 set up the average current, the wind current 

 was effective throughout the water column. 

 The average current flow was generally be- 

 tween east and southeast: the flow at the 

 surface, 153°T; at 20 meters, 146''T; and at 

 the bottom (40 meters), 112°T. The whole 

 column considered, the mean flow direction 

 of the average current was 137°T at a velo- 

 city of 0.11 knot. 



To determine whether this station (B) 

 was consistent with the counterclockwise 

 circulation in southeastern Bering Sea, an 

 attempt was made to remove the wind effects 

 in order to determine the average current 

 less the wind current. The wind current 

 velocity was considered to be 2 percent of 

 the wind velocity and flowing 42° to the 

 right of the wind (Rossby and Montgomery, 

 1935). Subtracting the vectors of the wind 

 current from the observed average current 

 results in an average current (only the 



