14 15 16 17 16 19 20 21 22 



FIGURE 18. VARIATION OF THE DEPTH OF THE THERMOGLINE WITH TIME AT WEATHER SHIP DELTA, 44°N.,4I°W., 1-22 AUGUST 1950. 



layer should form. Since the wave motion observed does not con- 

 form to this description, some obher factor must be at work. 



In an attempt to discover if a 29-day lunar variation in 

 the thermocline exists, the most complete set of daily BT obser- 

 vations available for one year was plotted. These data were for 

 the year 19l|8 at the Pacific weather ship NECTAR, An attempt 

 was made to eliminate the shorter period variations by plotting 

 the mean value for each day rather than each single observation. 

 This method was not successful because too many days had only 

 one or two observations instead of the expected six; consequent- 

 ly. It was impossible to tell in vrhich phase of the daily oscil- 

 lation this single observation was taken. Since short period 

 variations still appeared on the curve, a regression curve was 

 computed to eliminate them, as shown in figure 19. In this 

 graph the seasonal variation already discussed is apparent and. 

 In addition, a suggestion of ridges and trouglis appears In the 



41 



