18 U. S. COAST AND GEODETIC SURVEY 



little or no tendency to rotate. The reduction of weak reversing cur- 

 rents will be taken up in connection with rotary currents in the next 

 section. The observed directions were reduced to magnetic directions 

 by applying to each pelorus reading the proper corrections for the 

 ship's head and the deviation of the ship's compass. 



The observed velocities were plotted on cross-section paper, the 

 times of observations being taken as abscissae and the velocities 

 plotted as ordinates, the flood velocities above and the ebb velocities 

 below the horizontal line representing zero velocity. Smooth curves 

 were drawn following the general trend of the plotted velocities and 

 from these curves the times of slack waters and the times and velocities 

 of the strengths of flood and ebb were taken. These times and veloci- 

 ties, together with the magnetic direction of each strength of flood and 

 ebb were tabulated, usually in monthly groups, on forms prepared for 

 the purpose. The times of slack water and of strength of current were 

 then compared with the times of high and low water at a tidal refer- 

 ence station and average time difl'erences computed for each of the 

 four phases of current — namely, slack before flood, strength of flood, 

 slack before ebb, and strength of ebb. Average magnetic directions 

 of flood and ebb were obtained for each series of observations and the 

 average velocities of flood strength and ebb strength were computed. 



Finally the average time differences for the four phases were referred 

 to Greenwich transits of the moon by means of Icnown time relations 

 between the Greenwich transits and the high and low waters at the 

 reference station, and the magnetic directions of the flood and ebb 

 strengths were changed to true directions by applying the magnetic 

 variation. 



NONHARMONIC REDUCTION OF ROTARY TIDAL CURRENTS 



Rotary currents and weak reversing currents were, in general, 

 treated identically. As a preliminary step, the observed hourly 

 velocities were resolved into their north and east component velocities. 

 The component velocities were then tabulated in hourly groups 

 arranged according to the times of the high- and low-water phases at 

 a tidal reference station. The north and east components for each 

 group were summed and averaged separately, usually by months. 

 These averages included both tidal and nontidal currents. In order 

 to separate the two, the north and east components of the nontidal 

 current were obtained by averaging separately all the north com- 

 ponents and all the east components of the observed hourly velocities. 

 This process eliminated the cyclic or tidal current, leaving only the 

 north and east components of the average nontidal current. These 

 north and east components of the nontidal current were then sub- 

 tracted algebraically from the corresponding components of the 

 observed current for each hourly group. The resulting north and 

 east components for each hour represented the tidal current for that 

 hour. From these components the velocities and directions of the 

 tidal current for each hour of the tidal cycle were obtained. These 

 hourly velocities and directions were plotted on cross-section paper, 

 separate graphs with the same time scale being made for velocity and 

 direction. 



From the known time relationships between the Greenwich transit 

 of the moon and the high and low waters at the tidal reference station, 

 a second time scale reckoned from the Greenwich transit of the moon 



