ALTERNATING 

 FILTER FUNNELS 



COD END- g| 



WATERMETER 



WINCH 



ROLLER 



FIREHOSE WITH 

 ENCLOSED CABLES 



CONNECTOR 

 HOSE 



PUMP INLET 



Figure 1. — Schematic drawing of the plankton pump system. The drawing is not to scale. 



Three rings of rods and plates support each 

 shell- -one at each end and one in the center 

 of the shell. When the first layer of hose has 

 been unwound, the quarter sections of the outer 

 shell are unbolted in sequence to expose the 

 layer beneath. The second shell is similarly 

 removed after the second layer is unwound. 

 The winch is chain-driven at 2 r.p.m. by a 

 0.75-hp. motor with built-in brake (fig. 2d). 

 The disc brake at the end of the axle was 

 added as a safety measure. 



Launching the pump is relatively simple. 

 The pump, which weighs slightly less than 

 20 kg., is placed over the open roller at the 

 stern and lowered by its own weight as the 

 hose is unreeled. The hose has paint marks 

 at 5-m. intervals. Three men required about 

 45 minutes to unwind and retrieve 105 m. of 

 hose, including time for removal and replace- 

 ment of shells, but excluding sampling time. 



Four trials yielded some information on the 

 relation between length of hose out and depth 

 attained. The first trial was made while the 

 ship was drifting in a calm sea. A bathykymo- 

 graph, which records depth against time, was 



attached to the pump and 105 m. of hose was 

 unreeled. During descent, the winch was stopped 

 at 15-m. intervals to obtaindepth readings. The 

 tracing showed that when 60 m. of hose was sub- 

 merged, the pump was at a depth of 60 m.; 

 therefore the hose was vertical. The bathy- 

 kymograph malfunctioned beyond this depth. 



In the second and third trials the seas were 

 moderate. The ship headed into the swell as 

 hose was unreeled but was then allowed to drift. 

 In these trials, a BT (bathythermograph), which 

 records temperature against depth, was at- 

 tached to the pump. The winch was stopped, 

 and the temperature of the stream of water 

 delivered by the pump was measured after 

 each 15-m. length of hose was unreeled. By 

 later comparing these temperatures with the 

 temperature-depth profile on a BT trace, we 

 could estimate the depth of the pump at each 

 15-m. interval. With 105 m. of hose unreeled, 

 the pump was at depths of 55 and 60 m. at 

 a vessel speed of 1 knot and 100 m. when the 

 vessel was adrift. 



In the fourth trial, 35 m. of hose was sub- 

 merged while the vessel was anchored in 45 m. 



