dividing each of the 32 spectra into bandwidths (If) of 0.025 cycle 

 per minute and then averaging the energy per bandwidth does re- 

 veal some dominance in the spectrum. Figure 10 shows the 

 average power spectrum for the two isotherms of the 16 data 

 sections between San Diego and Hawaii. If no frequency band in 

 the spectrum is dominant then the ensemble average would not 

 show any peak. This is not the case, as shown by the two peaks 

 clearly retained in the ensemble average spectrum (fig. 10). One 

 peak is at 0. 15 cycle per minute (X = 0. 67 mile) and the second at 

 0.25 cycle per minute (A = 0.4 mile) . 



The thermistor chain data of Cruise 8, from around the tip 

 of Baja California, were also synthesized into spectral ordinates 

 (LaFond and LaFond, 1966) . 18 Cross sections of up to 12-hour 

 duration of depths of isotherms were analyzed in the same manner 

 as described for the Cruise 4 data. Figure 7 shows the location 



10,000 



0.05 



FREQUENCY, CYCLES PER MINUTE 

 0.1 0.15 0.2 0.25 



0.3 



0.35 



1,000 



100 



10 

















1 















\ 















\ 















\ 















\ 















\ 















I 















\ 















V 















\ 



























































































^/A = C 



.67 MILE 











■*«- 



/* 















V 















X 















N. 















^ 







0.4MILE- 









"S 

















































Figure 10. Ensemble average spectrum for NEL Thermistor Chain Cruise 4. 



25 



