188 Density of Water Masses in Ocean, Vertical and Horizontal Density Distribution 

 Table 78. Density and specific volume for dijferent s, t, p (ct^, j, j, and a^, <, p) 



This type of presentation was chosen in order to allow differences from the values 

 for standard ocean to stand out. The correction terms enclosed by rectangles refer 

 to the quantities already considered during the determination of o-^ and a^. It is 

 obvious that these are the main correction terms. It is, however, generally customary 

 to judge the vertical density stratification from the a^-values. This will also be done 

 here and the more correct cr^^f^j, and as,t,p will be considered again later. 



From stability considerations it is to be expected that the values of cr^ will increase 

 with depth. Apart from the surface layer down to about 50-100 m, this is always the 

 case. In the tropics and subtropics the increase is characterized by a transition layer 

 which begins just beneath the top layer, rising to a maximum gradient, then slowly 

 changing towards the deeper layers to a much smaller gradient. Towards higher 

 latitudes the intensity of the transition layer decreases more and more and beyond 

 35° N. and S. it becomes of no significance. In the Atlantic, for example, it can then 

 scarcely be regarded as a transition layer. In these regions the vertical density gradient 

 decreases steadily from the surface value downwards. In polar and subpolar regions 

 the density gradient from the surface layer down to the sea bottom becomes minimal. 

 Figure 87 shows the vertical distribution of o-j for some stations for which the T- 

 and ^-distributions were given already in Figs. 52 and 69. 



In the uppermost layer a small increase in the a^-values with depth can occasionally 

 be noted (see the first three stations of Fig. 87). This does not, however, necessarily 

 mean that the vertical density stratification of these water masses is unstable. Because 

 reduction of the a^-values to the more correct Og^f^p may remove these small differences 

 as happens in the three cases in Fig. 87. There still remains, however, a large number 

 of stations where there is undoubtedly a state of weak instability (see Chap. V, 6). 



A better insight into the nature of the vertical cr^-distribution through the entire 

 ocean is given by constructing longitudinal sections. Wiist has prepared sections of 

 this type for the Atlantic, indeed he chooses the same sections as for T and S (see 

 Fig. 62 p. 146 and 147). Figure 88 presents the o-rsection along the Western Trough of 

 the Atlantic ; the others show in principle the same picture. Although at the surface there 

 is a general slow increase of density, from the equatorial zone towards high southern 

 and northern latitudes, already at 100 m depth and below a diflferent distribution is 



