TIDES, WAVES 113 



(1) the shape of the lake-basin, and (2) the 

 temperature distribution. The smaller the 

 difference between the temperature of the 

 upper and bottom water the longer will be 

 the period. Mathematical formulae have been 

 obtained which take account of varying 

 breadth, depth, and temperature distribution, 

 from which the period of the temperature 

 oscillations can be calculated. 



Such oscillations are also possible in the 

 ocean wherever there is a sharp difference in 

 density some distance beneath the surface. 

 The difficulty of recording them, however, 

 is so great that there are no observations 

 which definitely show the existence of long- 

 period oscillations in the ocean. The period 

 of these oscillations must frequently be days 

 or weeks, and continuous observations in one 

 position over a lengthy period would be 

 necessary to demonstrate their existence. 



Otto Pettersson did observe an oscillation 

 in the Skaggerak with a period of fourteen 

 days, and, on the assumption that this was 

 a temperature seiche analogous to the seiches 

 observed in bays by the Japanese, E. M. 

 Wedderburn calculated the period for the 

 Skaggerak, which was found to be fourteen 

 days. It is possible therefore that what 

 Pettersson observed was a temperature seiche. 



Wherever we have two liquids of different 



