water masses. Each water mass, as it enters the circvilation pat- 

 tern either by influx from another basin such as the Mediterranean 

 water or by sinking from the surface at a particular site, can be 

 unambiguously identified on a T-S plot and then traced until dif- 

 fusion erodes its distinctiveness. Mavement within the ocean is for 

 the most part isentropic, and the Jfediterranean water, originating 

 from a warm saline sea, is traceable as a warm saline core spread- 

 ing across the Atlantic Ocean. 



The higher soxmd velocities at intermediate depths to the east 

 of the ridge axis and the change in profile in a relatively short 

 distance to the west indicate a rapid change in the properties of 

 the core layer. Examination of the T-S plots for either side show 

 this discontinuity clearly. If one next compares the T-S diagram 

 of a transitional station there is a possibility, if excessive 

 mixing has not already occurred, of dividing the water column into 

 layers of known origin. One could thereby reconstruct the process 

 by which a transition profile was formed. This was attempted for 

 the lowerings of Figure 5- 



Each reconstruction is somewhat different in detail and an ex- 

 ample of one is shown in Figure 6. On the left-hand side of the 

 figure, the T-S diagrams of two stations representing the east and 

 west groups are compared, a data point being plotted every ten 

 meters of water. The change in the salinity of the core layer 

 from 55*5 o/oo to 35«3 o/oo was typical of the present observa- 

 tions and can be demonstrated as being generally valid from the 

 data of this region contained in the NODC files. The Iso-density 

 lines of =27.^ and 27. 6I are indicated because it was at 

 those densities that the discontinuities in the profiles (temper- 

 at\ire, salinity and sound velocity) of the lowering to be recon- 

 structed. Wo. 50, were observed. 



On the right-hand side of Figure 6, two T-S diagrams are com- 

 pared. The dots are data points from No. 50 (one every two meters), 

 while the crosses define an artificial' composition of the fresher 

 (west) column, excluding the layer between the density limits, and 

 that part of the saline layer of the east column which was between 

 the density limits. The comparison is thought to be remarkably good; 

 the temperature and salinity discontinuities across the upper inter- 

 face are almost exactly reproduced. Lowering of Nos. ^5 and k$ can 

 be similarly reconstructed, and in this manner one group of transi- 

 tional profiles can be inferred to have evolved from the penetra- 

 tion of a saline water layer into a fresher water column. A sche- 

 matic of this interpretation is shown in Figure 7» 



Stem (1967) has hypothesized that the lateral encoimter of 

 two water masses such as described here will restilt in the mutual 

 penetration of each one into the other. The velocity profile 

 (Figure 2) and the T-S diagram of No. kh give some indication 



417 



