THE EXPERIMENT 
The experiment was conducted from the R/V GERDA during the period 
August 11 to 14, 1957, approximately 15 miles off the coast of Eleuthera 
Island, BWI. Figures 1 (a-c) show the tracks of the five buoys which sup- 
ported the thermometers. Each buoy consisted of a spar buoy from which 
were suspended four thermometer units and one temperature-pressure unit. 
Figure 2 illustrates the geometry employed. 
Each thermometer unit consisted of an open-aperture camera which 
photographed a mercury-in-glass thermometer illuminated at approximately 
two-minute intervals. The motive power for the film drive as well as the 
timing of the illumination was supplied by a 24-hour clock. Standard ther- 
mometers calibrated to 0.1°C were enclosed in a well to protect them from 
hydrostatic pressure. Good thermal contact between the thermometer and the 
well was secured by filling the well with water. The time constant for the 
complete unit was of the order of 1 to 2 minutes. 
On the first morning, the five buoys were launched six miles apart 
and allowed to drift. The next morning, the thermometer units were retrieved 
and replaced with another set of thermometer units and similarly on the 
third day. Hence the final result is a set of one-day thermometer records 
from five buoys at four depths. Due to various difficulties with the appa- 
ratus, usable data was obtained from only 36 of the 60 units launched. Of 
these, 29 records were processed for autocorrelation and power spectrum 
analysis. 
RESULTS 
A. Temperature Records 
The temperature records are typically fairly irregular functions of 
time with some short sections being oscillatory, with a vague hint of a daily 
or semidaily period, and with root-mean-square fluctuations in the order of 
0.3°C. No obvious correlations of the temperature fluctuations exist with 
