[7 = 6 knots = 3.1 m/sec, cos « = 0.2740 is found. Therefore, the angle between 

 the ship's direction (255°) and the direction of the internal diurnal tidal wave is 

 74°. The wave travels in either a northerly or a southeasterly direction. The 

 northerly direction has been accepted in consideration of the surface tide. 



The doppler shift up to the 5*' mode has been computed with the same 

 angle, 74° (indicated by arrows in fig. 12). The determination of the direction 

 of the semidiurnal internal waves has been carried out in the same way, assuming 

 that the 12.4- hour period is shifted to the second group of peaks. The results 

 are shown in figures 12-21. 



The figures show that a systematic shift occurs: the diurnal internal 

 tidal waves appear at periods between about 18 and 6 hours in sets 1-3, 2-4, 

 3-5, and 4-6, which represent positions north of 28° latitude. Therefore, they 

 always travel in the same direction, namely, to the north. In set 6-8 a shift 

 to 14.4 hours occurs but the waves in the following sets turn to the northwest, 

 as indicated by the peak that occurs at 24 hours in sets 7-9 and 8-10, and at 

 36 hours in sets 9-11 and 10-12. In set 11-13 the peak seems to be shifted even 

 further to the long-period range, which indicates that the wave has turned more 

 into the ship's direction. 



In regard to the semidiurnal internal tidal waves, one finds an angle of 67° 

 between the ship's direction and the direction of wave travel. Again there is an 

 uncertainty of + rr and the direction that agrees best with the direction of surface 

 tides is accepted. This leads to 45° for the direction of internal wave progress. 



The amplitudes in the semidiurnal range decrease with increasing set 

 number; they cannot be recognized in the sets following set 4-6. 



Figure 11 shows the ship's route, the analyzed direction of the diurnal 

 (25 hour) and semidiurnal (12.4 hour) internal tidal waves and the cotidal 

 lines of the semidiurnal surface tides (as given by G. Dietrich; see ref 13). 

 The results in regard to semidiurnal internal tides agree with the theoretical 

 concept given previously: "(1) in general they have the direction of the 

 surface tides as long as these can be considered to be forced by the tide- 

 generating force, and (2) their amplitude decreases toward Hawaii and their 

 direction cannot be determined for the western part of the cruise, because two 

 tidal domains seem to be present. There should be one internal tidal wave coming 

 from the northwest and another one from the southeast; the latter belongs to the 

 amphidromic system centered near 5°S, 155°W. Therefore, both waves may 

 cancel or at least merge in such a way as to greatly reduce the amplitudes. 

 The diurnal surface tides travel toward the west near Hawaii and in a north- 

 westerly direction elsewhere on the ship's track, as shown in Dietrich's phase 

 diagram for the surface tide. ^"^ This agrees with the expected directions 

 south of 28°N. North of this latitude only internal Kelvin waves are possible, 

 and since the coastline is north-south, it is expected that these waves would 

 also travel in a south-to-north direction. This is very pronounced in all 

 spectra north of 28° N. 



*See pages 21-28. 



51 



