Zambra, some by the GM Manufacturing Co., and some by the Kahl 

 Scientific Instrument Corp. The depths of observation were based 

 on Richter and Wiese unprotected thermometers. 



The thermometers were used in pairs and a program was followed 

 of periodically shifting the thermometers comprising the pairs. 

 This provided a series of comparisons of each thermometer with 

 several other thermometers and served not only to identify 

 thermometers which were not functioning properly but also to de- 

 termine consistent errors in the scale corrections in use and to 

 ascertain the degree of precision of the temperature measure- 

 ments. In judging this precision, 1,847 comparisons were con- 

 sidered. These gave a probable difference between the corrected 

 readings of a pair of thermometers of ± 0.013° C. As many of 

 the thermometers had recent laboratory comparisons with 

 thermometers tested by the National Bureau of Standards and, as 

 in most cases, the temperatures are the means of the corrected 

 readings of a pair of thermometers, it is considered that the tabu- 

 lated observed temperatures have a probable error of ± 0.01° C. 



As in previous years, the routine salinity determinations were 

 made with a Wenner salinity bridge which had been calibrated by 

 the measurement of a series of samples whose salinities were as- 

 sumed to be knovv'n from comparisons with Copenhagen standard 

 water by means of silver nitrate titration.^ Thus the accuracy 

 of the tabulated observed salinities is only that permitted by 

 silver nitrate titration although their precision is much better 

 and probably about ± O.OOSo/oq. For the most part, the salinities 

 are tabulated to the nearest 0.01o/qq. Each cell of the bridge was 

 standardized at the beginning and end of each run with a sub- 

 standard of salinity from a carboy of sea water kept under an oil 

 seal. Intermediate standardizations were made every 10 to 15 

 measurements. In addition, Copenhagen water was measured as 

 an unknown one or more times during the run. These measure- 

 ments of Copenhagen standard water were used to determine 

 minor corrections to the salinities determined for each survey and 

 to adjust the assumed value of the salinity of the carboy of sea 

 water used as a substandard. None of the surveys required 

 salinity corrections of as much as O.OOSo/oo and the figures illus- 

 trating the dynamic topography are in accordance with the tabu- 

 lated temperatures and salinities. 



- Since a substitution method is used, only relative conductivity is involved in the bridge 

 calibration. If C is conductivity and A and B are constants and S is the salinity corresponding 

 to conductivity C. then for the range of salinity from 30 to 40 0/,,,, the isothermal conductivity- 

 salinity relationship may be approximated by C = C (A + BS) . Early work by E. Ruppin 

 (O. Kriimrnel, Handbuch der Ozeanographie, Band I, p. 291, Stuttgart, 1907) and A. L. Thuras 

 (J. Acad. Sci., 8, p. 685, Washington, 1918) resulted in mean values such that C = C->-, 

 (0.10925 + 0.02545 S) . More recent work by Thomas, Thompson, and Utterback (1934), as 

 digested by Lafond in U. S. Navy Hydrographic Office Publication No. 614, results in a similar 

 expression for this range of salinity at a temperature of 25° C. of C = C35 (0.11055 + 0.025361 S) . 

 From the 1950 calibration of the salinity bridge, made at a temperature of about 28.9°C., the 

 similar expression derived was C = C33 (0.11436 + 0.0:25304 S) . ■■ 



20 



