The above conclusion helps to explain hourly temperature fluctua- 

 tions at 60 feet. The two regular wave cycles between sunset and sunrise 

 in Figure 5 are apparently caused by internal waves. Two fluctuations 

 during the daylight hours are partly masked by diurnal heating. If the 

 curve for the 60- to 100-foot layer in Figure h is correct, heat changes 

 at 60 feet were dix'ectly reflected at 100 feet. The maximum temperature 

 at I70OZ at 60 feet (Figure 5) was about 0.2'-' F higher than the secondary 

 maximum at OOOOZ. This difference may be interpreted as the mean diurnal 

 temperature range at 60 feet. Rough correction for this value in Figure 

 5 fails to eliminate the two fluctuations during the daylight hours. 

 Thus, four internal waves are apparently revealed by the hourly analysis. 

 The analysis automatically eliminates waves with frequencies of less than 



2 hours. Since tidal effects have been largely eliminated by averaging 

 the data over a period of 20 days, these internal waves must be station- 

 ary and are probably caused by some other phenomena. One theory of a 

 possible cause is isostasy; that is, the waves may be caused by convec- 

 tion during which the water particles seek, equilibrium after vertical 

 transfer from the surface. Inasmuch as the convective process required 



3 hoiirs to progress from 20 to Uo feet and 5 hours to progress from Uo 

 to 60 feet, this theory appears to be an unsatisfactory explanation of 

 the internal waves. 



Figures 6, "J, and 8 show the results of the convective process. 

 Figure 6 illustrates depth of convection and time of maximum heating. 

 Figure 7 shows mean BT's for each hour during the night; Figure 8 shows 

 the same for daylight hours. 



SENSIBLE HEAT CHANGES 



Temperature fluctuations in the upper 60 feet are analogous to, but 

 not identical to, changes in sensible heat content, which depend on the 

 density changes. Salinity observations were obtained from the 93 Nansen 

 casts. Possible diurnal changes in salinity could not be determined be- 

 cause these observations were not /evenly distributed throughout the day. 

 The majority were taken at 6-hourly intervals (03OOZ, O9OOZ, I5OOZ, and 

 2100Z). Mean salinity values in Figure 9 show little diurnal vertical 

 variation in the upper 20 meters; whereas the diurnal range of salinity 

 was considerable at ^i^O meters. This range is probably due to fluctua- 

 tions of the seasonal halocline. A mean salinity value of 36.25 *^/oo 

 was used with standard formulas for converting temperature into sensi- 

 ble heat. 



Figiure 10 shows hourly changes of total sensible heat content (Q) 

 in the three 20-foot layers of the ocean. The curves closely resemble 

 those of Figure 2, owing to use of the constant salinity value in the 

 formulas. In order to outline the difference in heat content between 

 layers more clearly, the differences between the first and second layers 

 and between the second and third layers have been plotted in Figure 11, 

 The relationship between heat content changes caused by solar heating 

 and convection is also shown. The fact that increases in heat content 

 due to solar radiation were practically simultaneous in all layers should 

 be noted in this Figiire. This simultaneity is due to high transparency 

 of the water. Cooling due to heat loss at the surface progressed from 



