Initial runs were made solely to adjust the vane. Lying down on the 

 carriage I was able to see the towing condition and slope of the tubes at all 

 of the speeds. From the front end of the carriage I could watch how the 

 azimuthal orientation was controlled by the vertical vane. Even the lowest 

 speed of 0. 50 knot was enough to turn the tubes in the proper direction. 

 Once aligned there was no discernible change or oscillation in the azimuthal 

 angle. The vertical declination of the tubes was observed visually from a 

 low platform that was only an inch above the water. I sat on it and kept 

 my line of sight to the tow point on the tubes perpendicular to the path. The 

 tubes hanging from a cable 15 ft long were about 8 ft below the surface, and 



1 was sitting about 5 ft to the side of the tow path. Because the line of sight 

 is then at a glancing angle into water, the tubes appear curved upward at 

 their ends. Adjusting the view as I did makes the upward end curves sym- 

 metrical when the tubes are horizontal. The best procedure is to observe 

 the change that occurs with speed. At 0.50 knot or less, because of the 

 careful balance of the tubes, they always appear horizontal. With the 

 adjustment vane perfectly horizontal the tubes turned downward at 2 knots, 

 and this declination was much more marked at 3 knots. The angle needed 

 for the vane proved to be extremely sensitive in its influence on tube ele- 

 vation in the speed range 2 to 3 knots. A final adjustment had the tubes 

 moving down a few degrees between 2 and 3 knots with level flight at 2. 

 For the presently contemplated use this is probably safest since any up- 

 ward angle tends to be accentuated. The drag on the stabilizing weight and 

 strut is reduced and a lift seems to work on the tubes themselves. 



Such extremely well-mannered behavior for a towing device made it 

 possible to calibrate the velocimeter assembly in terms of speeds that 

 could in the final analysis be set and measured to 0. 02 knot or about 1 

 centimeter /sec (Lord [1964]). 



TESTS AT SEA FOR TOWED VELOCIMETER SURVEYS 



The device was used at sea for the first time in March 1964 and it is 

 possible to compare depth fluctuation data on these runs with those 

 obtained on an earlier test which used only a dead weight to hold down a 

 single velocimeter. This was done in July 1962 at a series of four latitudes 

 ranging north of the island of Bonaire in the Caribbean. The parallel tube 

 work was done at 3 locations, west of the Windward Island chain, east of 

 the same chain, and finally southwest of Bermuda. Naturally the choice of 

 these locations was relevant to our study of sound velocity variation, not to 

 towing characteristics. However, weather plays such a large role in these 

 matters that we can use the depth records accumulated as if all possible 

 variations in towing characteristics were covered with equal comprehension 

 in the two sets of runs. 



Towing depth fluctuations were assessed by calculating them as devi- 

 ations from a moving average. In all runs the towing speed was around 



2 knots or 1 meter/sec and a moving average embracing around 210 meters 

 or 3z minutes of towing was calculated for most of the run. The instan- 

 taneous depth readings then depart from this as shown in the three typical 



51 



