vith the mean thermocline depth computed from a similar number of the 

 first BT's taken by a relief ship. The BT observations must be made by 

 both ships in a 2- to 3-day period, when the weather and other conditions 

 remain constant for several days before and after relief. Wot many cases 

 satisfy these requirements; examination of 7 years of data from station 

 ECHO for the months of July, August, and September reveals only six cases 

 satisfactory for comparison. The mean thermocline depths (n") were com- 

 puted for departing and relief ships, and the mean difference ^h) of 

 these six observations was 6.8 feet. The degree of accuracy seems to be 

 of the same order as the approximation of depths on a bathythermogram. 

 Severe inconsistencies, as given in the example above (SAN PABLO and 

 REHOBOTH), are probably not too numerous. 



A more serious drawback is irregularity and non- continuity of BT 

 observations. In addition, BT's are not always taken at the same loca- 

 tion; positions may vary constantly within approximately a 30-mile radius. 

 It is difficult to apply such observations to a detailed study of internal 

 waves or other similar problems, even though they were taken regularly and 

 at sufficiently short intervals of time. 



Mean Level 



The interface between the mixed layer and the thermocline constantly 

 oscillates vertically owing to internal waves. The mean level about 

 which this oscillation occurs can be compared to the mean level of sur- 

 face waves. 



One BT observation gives an instantaneous thermocline depth, h - h*-te, 

 where h is the mean thermocline depth (Figure 2), and z is the elevation 

 or depression with respect to h at the instant of observation. The mean 

 periods of internal wave spectra are longer than the mean periods of 

 surface 'wind-wave spectra; therefore, a group of BT's covering a period 

 of at least 2k hours is required for computation of the mean thermocline 

 depth. A group of BT's is considered satisfactory: (l) if the thermo- 

 cline is in a steady state, that is, not in the process of changing during 

 the period covered by the group; (2) when complete weather data and weather 

 charts are available for a period of at least 15 days preceding the BT 

 group; and (3) if the BT's in the group occur at regular intervals of 

 time. 



About 356 groups in the BT data of station CHARLIE approximated or 

 satisfied these requirements at the time of this study. Some of these 

 groups consisted of more than 30 BT's covering periods of several days; 

 some were limited to 3 BT's in one day. 



Internal Waves 



A systematic study of internal waves was not attempted after analysis 

 of the selected BT groups, because of deficiencies previously indicated in 

 the data. However, the range of oscillation about the mean level of the 

 interface was recorded for all groups of more than 3 BT's. These oscilla- 

 tion ranges are assumed to represent the approximate amplitudes of internal 



