Dr. Richard H. Fleming and Dr. Maurice 

 Rattray of the Department of Oceanography, 

 University of Washington, generously helped 

 plan the study. Thomas S. Austin, James W. 

 McGary, and Gunter R. Seckel of the U. S. 

 Fish and Wildlife Service, Pacific Oceanic 

 Fishery Investigations, also reviewed the 

 manuscript and made many useful suggestions. 



METHODS 



From an anchored vessel, the currents 

 were observed at three meters, the top and 

 bottom of the thermocline (where present), 

 and near the bottom, every hour for 38 hours 

 at each of four positions (ainchor stations). 

 An Ekman current meter (Ekman, 1932), re- 

 cording both current velocity and direction, 

 wjis used for the observations. No correc- 

 tion for the roll of the ship was attempted 

 because during the observations, except at 

 station B, the roll was usually nonexistent. 



Before the beginning of the observation 

 period at each station, a bathythermograph 

 (BT) trace was taken to determine the ther- 

 mocline depth, thereby locating the sampling 

 depths between the surface and the bottom 

 observed depth. (Because of isothermal con- 

 ditions on station B, only three depths were 

 observed.) The mid-depths thus found were 

 then used throughout the sampling period, 

 regardless of possible later changes in the 

 thermocline depth. 



In the analyses of the data for the 

 currents involved, portions of the methods 

 described in both the Manual of Current 

 Observation (U. S. Coast and Geodetic Sur- 

 vey, 1950) and the Admirality Manual of 

 Tides (Doodson and Warburg, 1941) were used. 

 Tables from the Manual of Current Observa- 

 tion were used to resolve the observed 

 hourly velocities for each station into 

 their respective north and eaist component 

 velocities which include both the tidal and 

 nontidal currents. (Negative north or east 

 values for the components indicate a current 

 flowing south or west, respectively.) After 

 the component velocities were obtained, the 

 method for obtaining mean sea level by use 

 of multipliers was used (Doodson and War- 

 burg, 1941, p. Ill, table 13.6) to extract 

 the average current from the hourly compo- 

 nents. This method has the advantage of 

 reducing casual error which would probably 

 be found in the hourly observations. By 

 then substracting the north and east compo- 

 nents of the average current from the 



corresponding components of the observed 

 current for each hourly observation, the 

 north and east components of the tidal 

 currents were obtained. 



TIDAL CURRENTS 



Observations of tidal currents occur- 

 ring in southeastern Bering Sea show that 

 both rotary euid reversing movements occur, 

 the rotary on the more offshore stations 

 (A and B), the usual locus for this type of 

 tidal current, and the reversing on the 

 inshore stations (C and D). The rotation 

 period for all stations was semi-diurnal, 

 accompanied by cyclic variations of the 

 current velocity, also semi-diurnal, thus 

 showing the relationship between the cur- 

 rents and the tidal cycle. Figures 2 

 through 5 are hodographs of the tidal cur- 

 rents, these showing the hourly observed 

 velocity cuid direction, the change in cur- 

 rent direction between observations and the 

 characteristic rotation occurring on each 

 station. 



In the northern hemisphere, the normal 

 rotation of a rotary current is dextral 

 unless some other interfering force modifies 

 the rotation. On station A the rotation 

 was normal at all observation depths, but 

 on B the direction of rotation varied with 

 depth. Dextral deviation occurred at the 

 surface, while at 20 meters and 40 meters 

 the pattern was confused, both dextral and 

 sinistral rotation occurring at each depth. 

 This pattern is shown in figure 3. 



To determine the principal flood and 

 ebb directions of the rotary currents, the 

 average maximum flood and ebb currents were 

 used, these showing, with slight individual 

 station variation, that the flood and ebb 

 were generally in a northeast-southwest 

 direction at all observation depths. The 

 directions of the flood or ebb and their 

 variations in average velocity and direction 

 with increased depth are shown in table 1 

 (page 7). 



On stations C and D, the tidal currents 

 were essentially reversing but with slight 

 rotation, the amount increasing with depth 

 to the bottom on C but remaining approxi- 

 mately the same at all depths on D. Because 

 of the slight rotation, apparently no slack 

 water occurred as is associated with the 

 completely reversing type of tidal current 

 movement . 



