ATLANT. DEEP-SEA EXPED. 1910. voi,. :] PHYSICAL OCEANOGRAPHY AND METEOROLOGY 



27 



not included, while on the other hand daily or casual 

 variations may be included which have nothing to do with 

 the tidal phenomena. It must also be emphasized that 

 tke results are very doublfiU on account of the shortness 

 of our series, as mentioned above. The discussion of the 

 observations may, however, serve to form a working 

 hypothesis for further investigations, for which reason we 

 shall go into details in spite of the fact that the obser- 

 vations are very incomplete. 



The maximum variations seemingly caused by diurnal 

 and semi-diurnal oscillations separately, appear in co- 

 lumns 4 and 5 in the table above. Upon adding together 

 for each lunar hour the two values of the temperature 

 deviations ascribed to these oscillations, we obtain a new 

 series of figures. In column 6 of the table the maximum 

 difference between the latter figures is entered (C). It is 



in many cases comparatively very great, at some depths 

 even as great as the difference between the maximum 

 and minimum temperatures observed. 



Knudsen has tabulated the values of 



B 



for 



max mm 



the Danish station and found a maximum value of this 

 quantity amounting to 0-3 at 400 and 500 metres where 

 the temperature changed most quickly with the depth. In 

 column 8 of our table above the corresponding values 

 are recorded for Stats. 115 and Sc, together with the 

 values given by Knudsen. At the Norwegian and Scottish 

 stations these values are in nearly all cases larger than 

 at the Danish station, which may be due to the shortness 

 of our series of observations at the two former stations com- 

 pared with that at the latter. The maximum values (0^6 — 

 0-7) are found at 100, 300 and 400 metres at Stat. 115, 



