51 



to present here figure 54, which is a dynamic current chart sho\ving 

 the horizontal motion at the 800-decibar surface rehitive to the 2,000- 

 decibar surface. In the manner in which Wiist has shown liis inter- 

 mediate water graphically, it should be moving southward at about 

 this level (800 meters) in the southern part of the area covered by 

 figure 54, However, it is evident from figure 54 that in this region 

 the general movement of the water at 800 meters is in a direction simi- 

 lar to that of the Atlantic Current and has a major northeasterly 

 component. The intermediate water of Wiist is identified by the 

 criterion of minimum salinity and such water has been found, not 

 only by others, but also in tliis sm-vey and in this particular region. 

 However, as pointed out by Smith, Soule, and Alosby in United States 

 Coast Guard Bulletin No. 19, Part 2, it should be looked upon not as a 

 dhect southward sinking from near the surface in the vicinity of the 

 "polar front", but as the product of mixing, cabbeling and gradual 

 sinldng taking place along the borders of the Labrador Current, the 

 resulting mixed water following the same general directions as the 

 parent cm-rents. This view does not deny the possibility of lateral 

 transfer by mixing. On the contrary, the very observations which 

 demonstrate, in figure 54, the direction of horizontal motion of this 

 water may help to clarify the mechanism of such lateral transfer. 

 Rossby,'* in liis work on m ake streams, and Parr,^ in his work on the 

 Caribbean, have concluded from independent considerations that 

 horizontal mixing is a maximum where vertical stability is a maximum. 

 This principle can be applied to the low-salinity water in the vicinity 

 of the fiftieth paraDel. In figure 55 is shown a vertical section of 

 salinity approximately at right angles to the direction of flow. The 

 geographical location of the section is indicated by the line AA in 

 figure 54. Here we see a tongue of minimum salinity whose axis lies 

 at about 800 meters at the left hand (southern) end of the section and 

 at successively higher levels to the right (north). In figure 56 the 

 vertical stabihty,^ expressed as the depth rate of change of ct, has been 

 shown for the same section. The axis of maximimi stability is shown 

 to approximate with remarkable closeness the axis of minimum sahnity 

 shown in figure 55. Above tliis axis of maximum stability is a layer 



* C. G. Rossby. Dynamics of steady ocean currents in the light of experimental fluid mechanics. Papers 

 in Phys. Oceanog. and Met. M. I. T. and W. H. O. I., vdl. V, no. 1. Aug. 1936. Cambridge. 



» A. E. Parr. A contribution to the hydrography of the Caribbean and Cayman Seas. Bull. Bingham 

 Oceanog. Collection, vol. V, art. 4, 1936. New Haven. 



• Strict adherence to the definition of stability permits only the determination of the average stability of a 

 layer so that in order to determine the stability at a point one must estimate the course of a smooth curve of 

 vertical distribution drawn with regard to a succession of computed averages for a series of layers. The 

 accuracy of the resulting value of stability at a point will therefore depend on the assumed course of the 

 smooth curve. Such individual values of stability are therefore approximations. Another method for 

 determining good approximations to point-stabilities is to graphically determine tangents to a carefully 

 drawn vertical distribution curve of potential densities ((Jt^). As the potential density, ct^, is not signifi- 

 cantly different from the corresponding crt in depths less than 1,000 meters, and as the phenomena consid- 

 ered occur in these shallow depths, the stabilities made use of here are the graphically determined tangents 

 to vertical distribution curves of crt; the use of true stabilities being regarded as an unnecessary refinement. 



