of precise deteriniiiations of chloriiiity b}- the Volliaiclt method, with 

 checks against silver of high purity; and of electrical impedance by 

 means of the Coast Guard's Wenner salinity bridge. The procedures 

 followed and the results obtained in the precise chlorinity determina- 

 tions are given in appendix 1. Other details regarding location of 

 origin and date of collection of the large samples are given in 

 appendix 2. 



Three measurements of each sample were mach' on the Coast Guard 

 bridge during each of two I'uns on 25 and 26 February 1960. The 

 samples w^ere also measui-ed on the Woods Hole Oceanographic 

 Institution's bridges No. 1 and Xo. 2 on 16 March 1960. Schleicher 

 and Bradsliaw of WHOI considered, from an examination of Thomas, 

 Thompson, and lUterback's work on the electrical conductivity of sea 

 water, ^ that the relationship between conductivity and sahnity of sea 

 water could be expressed b}' a second degree relationship over the 

 salinity range from 30 to 40°/oo- Then* took actual high salinity 

 Atlantic water (ca. 36°/oo) and concentrated it to about 40%o by 

 slow evaporation. From tliis they prepared liv(» samples, distributed 

 in salinity from 30 to 40°/oo, by dilution witli conductivity water. 

 Fi'oni tlie second degree relationshi]) and measurements of tlie ratios 

 of the salinities of tlie prepared samples they determined the shape of 

 theu- calibration curve and fixed it laterally by the measurement of 

 Copenhagen water. 



The Coast Guard has always used an assunu'd linear relationship 

 between conductivity and salinity over a lesser range of about 31 to 

 36}^°/oo but has avoided the assumption that Copenhagen water has 

 tlie same chlorinity-conductivity relationship as actual newly col- 

 lected sea water. The February 1960 measurements present an op- 

 ])oit unity of examining the magnitude of the errors involved in assum- 

 ing a linear relationship between conductivity and salinity of sea 

 water over this range, of determining the corrections to salinities pub- 

 lished in the Ice Patrol bulletins for the postwar years, and the varia- 

 tions between the chlorinity-conductivity relationsliip of sea water 

 and of several batches of C^openhagen water. 



In assuming a linear I'elationship between conchictivity and salin- 

 ity, the calibration curve of the Coast Guard's bridge lias the form 



S=-^ C where m is the reading of the variable impedance in the 



B-\-m 



X-dial arm of the bridge at balance corresi)on(ling to a salinity S; 

 and A, B, and Care constants which depend in part on the arl)itraiy 

 selection of one ])oint on the curve. In the ]9r)7 calibration curve 

 this point was an X-dial reading of 49.91 1 corresponding to *S'=35°/oo 

 and the numerical exj^ression was 



■ Thomas, liertram D., Thomas G. Thompson, and Clinton L. Ilttorback, The ek-cfrical conductivity of 

 sea water. Jour. Conseil, v. 9, pp. 28-35, 1934. 



34 



