Tides of the Oceans 



499 



pated partly in the north on the vast shelf of the Arctic Sea and by its ice 

 cover, and has not returned in the form of a reflected wave into the Atlantic 

 Ocean. As mentioned in Ch. XI. 3c, p. 345 the form of the co-oscillaiing tide 

 in such an ocean is no longer a standing wave, but must be represented as 

 a superposition of waves which are displaced with reference to each other. 

 This explains why in the south Atlantic, the tide wave is almost a pure 

 progressive wave. We would then have in the Atlantic Ocean a similar case 

 as in the North Sea, where the wave coming in from the north loses so much 

 energy in the south by friction that although the reflected wave is sufficient 

 to develop an amphidromy in the Deutsche Bucht superposition in the north- 

 ern North Sea produces only a progressive wave (see p. 368). The co-oscil- 

 lating tide in the Atlantic has the phase 1 h in the south (see Chart I), the 

 corresponding orthogonal oscillation the phase 4 h, and these can be com- 

 puted numerically for the centre axis of the ocean, taking into account the 

 effect of the tide generating forces. The boundary conditions are for the wave 

 with the phase 1 h that at the cross-section zero in the north (Iceland) ^ = 

 and r\ x is given an arbitrary value ; for the wave with a phase of 4 h, r) 4 = 



Table 85. Semi-diurnal tidal constants at islands in the Atlantic 



and | 4 is given an arbitrary value. The boundary conditions at the other 

 end (co-oscillation with the tides of the Antarctic belt determine the free con- 

 stants. These constants can also be computed by means of the tidal ob- 

 servations on some islands which are near the centre axis, for instance the 

 Azores and Tristan da Cunha, which has proved to be more practical. 

 In Table 85 are given the amplitude and phase as observed on these islands 

 and also the component values for both waves with the phase 1 and 4 h. 



32* 



