Long Waves in Canals and Standing Waves in Closed Basins 239 



waves developed after each explosion, which caused the greatest devastations 

 on the shores of the Sunda Strait (see Symons, 1888). The height of the 

 waves varied, but attained at a great distance from the origin 15 m, and 

 in some places even 22 m. In the harbour of Batavia the waves were recorded as 

 a sudden wave crest of 1-8 m height, which was followed by fourteen others 

 with a long period of 122 min. At nearby localities a wave crest preceded 

 the wave trough. The waves did not enter the Pacific Ocean but crossed the 

 Indian Ocean in all directions and also could be traced up the Atlantic Ocean, 

 where they were observed, e.g. by the German Expedition on South Georgia, 

 13 h 57 min after the great explosion. They were even observed in the 

 European waters (Socoa, in the innermost corner of the Gulf of Biscay: 

 amplitude 8 cm, Rochefort 13 cm, Devonport 15 cm,' and others). These 

 waves, starting from the Sunda Strait, covered a distance of at least 20,000 km 

 in 32 h and 35 min. 



These dislocation waves are waves with a long period ranging from 10 min 

 to 2 h, the large values being quite rare. The velocity is of the order of 

 magnitude of about 180 m/sec and upwards. The wave lengths are, therefore, 

 exceedingly large, and one can apply the Lagrangian equation c = \ gh. 

 A condition for this is that the period of the waves should remain constant. 

 By applying this formula, it was expected to compute the mean depth of 

 the ocean from the travel speed of the waves obtained at various coastal 

 localities. Wharton and Evans (1883) did point out that the observed 

 values of c were always smaller than might have been expected, and Davison 

 (1897, p. 33) has proved that one obtains systematically too large values 

 of c, if the mean depth is substituted in \/gh. As the depth is variable along 

 the path of the wave, the Green-Du Boys equation (VI. 32) should be used 

 instead of the simple Lagrangian formula ; it is however, questionable whether 

 it can be used for waves starting from one point. Thorade has pointed out 

 these waves are circular waves and not canal waves, and that the Greenian 

 condition, that the depth varies slowly over a distance of a wave length, 

 is hardly fulfilled. When waves follow each other in a rapid succession, one 

 should consider the group velocity. 



The propagation of the tsunami of 1 April 1946, from its origin in the 

 Aleutians, south of Unimak island over the eastern Pacific Ocean has been 

 examined thoroughly. It caused large-spread destructions on the Hawaiian 

 islands (Macdonald, Shepard and Cox, 1947), and it was recorded by 

 a great number of gauges along the entire American coast (Green, 1946). 

 Figure 101 gives the record of the marigraph for Valparaiso, where the 

 wave arrived after a travel time of 18 h and 7 min, with an average velocity 

 of 390 nm/h. The observed travel time agrees very well with the theoretical 

 one, computed from the depth charts after Du Boys's formula. The wave 

 period of the first wave at the station nearest to the origin was 15 min 

 and increased gradually to 17-4 min at the more distant stations. At a fixed 



