To be discussed below is a collection of velocity profiles 

 from a specific area which, while consistent with the broad 

 description, had a variable behaviour at depths about the sound 

 channel axis. The twofold cause of the variableness of the 

 channel was the spreading of Mediterranean water and the effect 

 of the Mid-Atlantic Ridge on the general circulation of the sub- 

 surface waters. 



It is well known that the efflux of warm saline water from the 

 Mediterranean Sea is the cause of a significant distortion to the 

 sound velocity profile in a wide region about the Strait of 

 Gibraltar. Penetrating the Atlantic water column above the sound 

 channel axis, the warm saline water results in a distinctive 

 double sound-channel profile whose variableness and geographic 

 dependence is first being systematically investigated with obser- 

 vations like those of Piip (1968). To a lesser degree, the 

 spreading of this water influences the shape of the velocity pro- 

 file over an extensive zone of the Atlantic Ocean. Any process 

 which disrupts the continuous diffusion of the Mediterranean 

 water's anomalous salinity will be reflected in the profiles of 

 that region, and such is the case at the Mid-Atlantic Ridge, in at 

 least one area. There, the ridge will be shown to mark a change 

 in the sound velocity profile with a narrow transition containing 

 unpredictable, but explainable, sound velocity structure. 



During August and early September of 1968, continuous salinity 

 and temperature profiles were obtained at seventeen stations to 

 the north of the Azores. The lowerings straddled the ridge axis, 

 and their positions relative to a coarse plot of the simultane- 

 ously obtained bathymetry are mapped in Figure 1. The depth of 

 the sound channel axis from each lowering is noted. The two 

 ridge crests enclosing the median valley are identifiable by the 

 2,000 meter line, including a lateral displacement at 43°N which 

 is typical of the ridge. Crests are seen to reach within 1500 

 meters of the surface and, though not shown, some peaked at depths 

 less than 1000 meters. Thus, the ridge intermittently rises above 

 the level of the sound channel axis, a depth which itself is only 

 poorly defined for this region. 



The lowerings indicated that the movement of water at inter- 

 mediate depths was neither stationary nor continuous on a broad 

 front. Two lowerings which were repeated after two and three 

 weeks and at locations within several miles of each other (nos. 40 

 and 49, nos. 36 and 44) had basically different velocity profiles. 

 A sequence of three stations in 60 n.m. across the ridge axis, 

 completed in 36 hours, led to equally diverse results. These 

 three are shown in Figure 2 and one may note not only their dif- 

 ferences, but also how unusual two of them are. No. 44 has 

 almost no vertical velocity gradient for 500 meters about the 



415 



