hour before the run showed a layer depth of 70 m with no discernible ver- 

 tical gradients. 



Figure 5 shows the results of a deeper run at 800 m. The temperature 

 scale is twice as sensitive as in the previous figure. One of the interest- 

 ing features here is the departure of the tail thermistor reading from the 

 other two instruments. Subsequent investigation showed a high resistance 

 salt water short on that thermistor. This illustrates one of the problems 

 that can arise with non-redundant instrumentation. 



Figure 6 is a run at 1300 m. Once again, the three instruments track 

 very well and the smoothing caused by the longer time constant of the 

 Gulton Thermometer can be seen. 



One of the useful features of the vehicle is the ease of maneuverability 

 at depth. Figure 7 shows the result of a 180° course change half-way 

 through a two hour run. The two paths are displaced about 200 feet hori- 

 zontally with respect to each other. The top trace is the total sound veloc- 

 ity record as recorded. The bottom trace is the upper trace expanded 

 twice in length and folded at the turn. Data near the right hand portion of 

 these curves are only a few minutes apart in time while at the left they are 

 two hours apart. The large slowly varying excursions (internal waves? ) 

 are reproduced very well but the small high frequency variations are not. 

 This is not unexpected in view of the horizontal displacement in the two 

 paths and the passage of time. 



DATA ANALYSIS 



In analyzing these data, care must be taken to assure that changes in 

 run depth resulting from control system perturbations are noted. In the 

 following summaries only the data in which this source of error is negli- 

 gible are included. The criterion used is that the temperature or sound 

 velocity variations observed must be at least 10 times the product of the 

 depth variation and the local vertical gradient. 



The primary interest in taking these data has been for acoustical appli- 

 cations. One of the primary acoustical questions has been horizontal 

 gradients and their magnitudes. The results from these measurements 

 are shown in Figure 8. Here the data from the March 1965 cruise are 

 summarized. The horizontal temperature gradient has been plotted as a 

 function of depth over a run length of 6 - 12 nautical miles. The solid 

 points are the total gradient on runs in which 90° turns were made. The 

 crosses are single gradient components derived from straight runs. In 



352 



