that a bathythermograph was attached just below the 150-pound weight to record depth and that the 

 wire angle and wire out were measured at 1- to 2-minute intervals during each haul. From these 

 measurements the approximate curvature of the towing wire on these four hauls has been recon- 

 structed (fig, 2) by means of a series of triangles. If we assume that the wire maintains the same 

 angle as it is lowered farther into the sea, then the sum of the vertical sides of the triangles 

 should approximate the true depth attained under uniform current and ship speed. The pertinent 

 data on the four test hauls are given in table 2. It is evident from this table that the method illus- 

 trated in figure 2 produces an estimate of the depth of hauling which is much closer to the depth 

 recorded by the bathythermograph than the estimate based on the assumption that the towing wire 

 conformed to a straight line. 



Figure 2. --Reconstruction of the curvature of the towing wire during four test 

 hauls utilizing wire-angle and wire-out observations. Depths of 

 the samplers assuming the towing wire to describe a straight line 

 are shown for comparison. 



The next step was to plot for the four test hauls the estimated depths calculated by the 

 straight-line method against the estimated depths calculated by the curved-line method. On 

 logarithmic graph paper the distribution appears to be rectilinear and a straight regression line 

 was fitted as shown in figure 3. This line was then used to convert the straight-line estimated 

 depths of cruise 16 to the equivalent curved-line estimated depths (given in table 1), using the 

 maximum wire out and an average wire angle for the 1-hour haul. For example, if the estimated 

 depth of the deep sampler on one of the hauls was 100 meters calculated by the straight-line 

 method, the converted depth for that sampler would be 120 meters as derived from the graph in 

 figure 3. When the converted depths are plotted in reference to the temperature structure, as in 

 figure 4 and the selected bathythermographs of figure 5, it is apparent that the intermediate 

 sampler was operating within the thermocline and usually between the 60 and 70 isotherms. 

 The deep sampler was below 60 F.--with one exception (station 29)--and frequently below the 

 50 F. isotherm. 



This method of estimation was necessary because wire out and wire angle were not re- 

 corded on cruise 16 while the samplers were being lowered but only during that portion of the 

 haul when the samplers were open and fishing. This precludes direct estimation by the use of 



