change and their driving mechanisms. Valuable understanding can be 

 achieved by continuing to work from ships with lowered instruments, but 

 it seems that anchored buoys and towed instruments could be especially 

 advantageous for this work. 



The occurrence of the depressed shallow sound channel has long been 

 known, and sound transmission studies have been made along it. Never- 

 theless it seems worthwhile to remind ourselves that this is a gcod sound 

 channel, near the surface, and near home, where existing variable depth 

 sonar can be used to considerable advantage. Sonars which can be towed 

 at 600 feet will find great advantage there throughout the year. But it does 

 have rather distinct limits about which we know little. These limits 

 almost certainly change, but we now have little idea how fast and only a 

 vague idea how far. It would be worthwhile to find out. 



Conclusion 



Sound velocity data from two cruises to the area between Bermuda 

 and the Antilles in 1962 and 1964 have closely the same range of velocities 

 at any depth, differing at most by only a few meters per second. This 

 difference is the same as the change which took place between two 

 measurements at the same location during a single cruise. A general 

 decrease in sound velocity from north to south was found at all depths 

 from 300 m. to 1000 m. in both studies. A maximum of sound velocity 

 in the western half of the area with an accompanying minimum in the east 

 found in the 1962 results have become reversed in 1964. Evidence is 

 inadequate to distinguish between recurrent internal waves and circulation 

 transients as the explanation for these large scale characteristics of the 

 water. 



The whole area is divisible into three regions depending on the sound 

 velocity structure abcve 500 meters. Near Bermuda this part of the water 

 column is partially a sound channel. Farther south and west the same depth 

 has nearly constant sound velocity as a function of depth. Still further 

 south the sound velocity gradient is continuously negative. The secondary 

 sound channel which appears between the surface and 500 m. in the 

 northern profiles in 1964 was absent in 1962, the corresponding structure 

 being isovelocity. This and other evidence of change in the position 

 of the boundaries between these water masses indicate changes in sound 

 transmission. The time scale and magnitude of these changes is not known. 



445 



