The significant value of geopotential anomaly (0.01 dynamic meter) must also be con- 

 sidered in assessing the significance of computed vertical velocities. Since the flow 

 determined here is geostrophic, the variation of vertical velocity, w, with depth can be 

 approached through the expression 



dw_/3 

 dz f 



where vis the meridional component of total velocity and /3 and /are as previously defined. 

 In the present study area the horizontal flow is mainly zonal; meridional velocities are 

 usually of the same order as the level of significance determined for horizontal velocities. 



So the values of — obtained by the above expression are questionable. 



To help establish the consistency of the observations, and also to determine the validity 

 of the proposition that baroclinic velocity vanishes at the bottom, plots were made of the 

 geopotential anomaly as a function of latitude. Figures 19 and 20 show the geopotential 

 anomaly at 4,000 meters referred to 5,000 meters: figures 21 and 22 are of the anomaly 

 at 3,000 meters referred to 5,000 meters. At 4,000 meters, the greatest range of the 

 anomaly is 0.02 dynamic meter, only twice the measurement error. This does not permit 

 the assumption that baroclinic flow in the interval from 5,000 to 4,000 meters can be 

 ignored, but it does indicate the relative unimportance of this deep internal structure. 

 The geopotential anomaly at 3,000 meters referred to 5,000 meters varies more with time 

 and place than does the anomaly at 4,000 meters. Effects of water structure in the 3,000- 

 to 4,000-meter interval must certainly be included in the determinations. 



798-230 0-66— 2 



