where S is salinity in parts per thousand and m is the X-dial reading 

 of the bridge at balance. The similar expression determined in 1954 

 is: 



200. 1+m 



Over the range of salinities encountered in the area (30.2 to 36.6°/oo) 

 the difference between these two curves may be compensated for with 

 negligible error by adjustment of the X-dial reading selected as 

 corresponding to the salinity of the Copenhagen standard water used 

 (batch PI 7). 



The Y-cell, which is filled with sea water simply to provide a bridge 

 arm liaving thermal characteristics similar to those of the arm made 

 up of the X-cell, undergoes a slow drift in its impedance as its contents 

 is gradually concentrated through evaporation. It has been surmised 

 that there also are erratic changes in its impedance as discrete drops 

 of condensate return to the solution from the neck of the cell. The 

 first change is evidenced by the slow change in the setting of the 

 Y-dials necessary to bring the bridge onto the calibration curve 

 during successive standardizations. As the range of adjustable im- 

 pedance covered by the Y-dials has a finite limit, the evaporation 

 from the Y-cell eventually reaches a point where the solution must be 

 changed. In an attempt to remedy both the slow drift and the sup- 

 posed erratic changes in the impedance of the Y-cell, a layer of oil 

 about 1 mm. thick was placed on the surfaces of the solution in the 

 cell and in the exhaust tube. No noticeable changes in the rate of 

 drift resulted during the approximately 3-month period covered by 

 the preliminary measurements ashore and the first two surveys. The 

 oil layer was then increased in thickness to about 1 cm., with the 

 result that the slow drift in impedance was very much reduced. 



During the routine measurements standardizations were made with 

 water from an oil-sealed carboy of sea water, and at least twice during 

 each run a sample of Copehnagen standard water of batch Pl7 was 

 measured as an unknown. At the end of each survey these measure- 

 ments of Copenhagen water were used to compute such corrections 

 as had to be made to the salinities measured during the survey. For 

 each of the four surveys and the postseason cruise, the indicated 

 corrections did not exceed 0.005°/oo so no corrections have been made. 

 These measurements of Copenhagen water indicate the salinity deter- 

 minations were precise to about ±0.005°/oo. The accuracy is, of 

 course, limited to that of the silver nitrate titrations involved in 

 determining the calibration curve of the salinity bridge. 



The determinations of the total phosphorus concentration were 

 carried out by personnel of the Woods Hole Oceanographic Institu- 

 tion, and the probable error is about ±10 percent. With only minor 



83 



