General Theory of Ocean Currents in a Homogeneous Sea 



447 



hours. This period is only 13 min shorter than the diurnal moon period K^ of 23-94 h 

 which presumably was also present. This difference is, however, sufficiently large to 

 decide out of six wave trains which of the waves is present. Figure 194 shows the diurnal 

 regular oscillation after elimination of the semi-diurnal tide. While during the first 

 three days there was a regular damping of the waves, at the end of the series there 



Fig. 194. North Atlantic Ocean: anchor station of the "Armauer Hansen" 30° 13' N., 

 13° 57' W. Current measurements in 5 m depth after elimination of the semi-diurnal tide. 

 Full line, north component; dashed line, east component; velocity scale in mm/sec. At the 

 upper rim moon hours. The distance between two vertical lines is very nearly 6 pendulum 

 hours (according to Helland-Hansen and Ekman). 



appeared to be a phase shift in the meridional component due to a new disturbance ; 

 the oscillations then lose rapidly in regularity. Harmonic analysis for the first three 

 and then for the following three days gave (cm/sec, / in pendulum hours) : 



cos (27r/12. t - 112^) , N = 1-58 cos (27r/12. t - 102") 

 115^) ^"^E 



N = l 



E = 1-51 sin (27r/12. t 



1-29 sin (277/12. / - 97°) 



These oscillations are pictured by the full and dotted sine curves in Fig. 194. The 

 good agreement led Helland-Hansen and Ekman to interpret these waves as inertia 

 movements. The phase difference between the two components was 12 min more for 

 the first days and for the second three days 20 min less than the theoretical required 

 value of 6 h. The average ratio of the amplitudes was 1-23 as compared with a 

 theoretical value of 1 . The oscillations were thus of the elliptic type with a ratio of 

 5 : 4. Considering that besides the inertia oscillations presumably the diurnal tide was 

 also present, the results obtained are very satisfactory. 



An unambiguous proof of the occurrence of inertia oscillations was provided by the 

 current measurements organized by H. Pettersson in the Baltic. As an adjacent sea with- 

 out any significant tides this is particularly suitable for such an investigation. Gustaf- 

 SON and Otterstedt (1932) and Gustafson and Kullenberg (1933, 1936) have 

 made a detailed analysis of the suitable current measurements in the Baltic; in many 



