498 Tides of the Oceans 



for the European coasts than the wave developed in the ocean itself. Krummel 

 (1911, vol. II, p. 250) has discussed all these more qualitative viewpoints, 

 without however attempting to give a more quantitative explanation of the 

 tides in the Atlantic (see Thorade (1931, p. 136)). 



Sterneck's (1922) new presentation of the co-tidal lines has revised the 

 question of explaining the Atlantic tides. Defant simplifying considerably 

 the assumptions (1924, pp. 153, 177 and 1926, p. 133) was the first to explain 

 by longitudinal and transverse oscillations the tidal picture as shown by recent 

 maps of co-tidal lines. He assumes that the Atlantic Ocean, including the 

 Arctic Sea, can be represented by a large canal closed at one end, in which the 

 tides result from the superposition of independent tides developed in the ocean 

 itself with tides co-oscillating with those of the large Antarctic belt. These 

 longitudinal oscillations are accompanied by two kinds of transverse oscil- 

 lations: (1) those caused by the effect of the rotation of the earth on the 

 periodical water displacements resulting from the longitudal oscillations; and 

 (2) independent transverse oscillations between Europe— Africa and America. 

 The orographical configuration of the ocean was fully taken into account by 

 Defant. All these possible oscillations were calculated numerically by step- 

 wise computation of the different variables. The superposition of all these 

 oscillations gave a tidal picture which in its most essential points, was quite 

 similar to the observed one. There was never any doubt that for tides of the 

 Atlantic Ocean the co-oscillating tide with the periodical water displacements 

 in the Antarctic belt are much more important than the independent tides 

 and that the basic thought of such a hydrodynamical theory is correct. 



Sterneck (1926, p. 1) has rejected this simple theory, and he criticizes 

 especially Defant's assumption of transverse oscillations. He published an- 

 other theory in which he limits his explanations to the centre axis of the ocean. 

 He splits the longitudinal oscillation derived from observations on islands 

 into two oscillations; one with a phase of 1 h and the other one with the 

 phase 4 h. The first one agrees very well, the second one very badly with 

 the values derived from the theory. This is a point which still needs explana- 

 tion. He disagrees in general with the possibility of splitting the oscillation 

 into longitudinal and transverse oscillations. He tries to make his theory fit 

 the observations made in the N. Atlantic by assuming oscillations in triangle- 

 form which seems rather artificial. However, the introduction of these trian- 

 gular oscillations means nothing else than splitting the oscillations into separate 

 waves, although he does not use any longer rectangular coordinates, which 

 complicates the question. 



Defant (1928) has elaborated on his theory of the Atlantic tides in the 

 volume of the "Meteor" Expedition dealing also with the anchor stations. 

 Taking into consideration the more recent theory of the tides in adjacent seas 

 and canals, Defant assumed that a part of the tidal energy penetrating into the 

 Atlantic ocean from the south in the form of a progressive wave, was dissi- 



