112 CARRITT [CHAP. 5 



The original definition of chlorinity required a knowledge of the chlorine 

 equivalent of bromine and iodine. Chlorinity values so defined required revision 

 with each redetermination of the atomic weights of the elements involved. A 

 more recent, and presently accepted, definition which makes chlorinity in- 

 dependent of atomic weight values, given by Jacobsen and Knudsen (1940), is : 

 "The number giving the chlorinity in per mille of a sea- water sample is by 

 definition identical with the number giving the mass with unit gram of Atomge- 

 wicht Silber just necessary to precipitate the halogens in 0.3285234 kilograms 

 of sea- water sample". 



Thus, the accuracy and precision with which chlorinity can now be deter- 

 mined depends upon our ability to prepare atomic-weight silver, to weigh 

 solids and solutions, and to carry out rather simple chemical operations. 



The measurement of chlorinity on routine samples has been done generally 

 by the Knudsen titration (for details see Oxner, 1920) using Copenhagen 

 Standard Sea- Water as the primary standard. The accuracy and precision of 

 measurements made by this procedure are of the order of + 0.01 % in chlorinity. 



Replacement of the colorimetric and point detection of the Knudsen titration 

 by potentiometric methods has been reported by several investigators. 



An automatic chlorinity titration, described in NAS-NRC (1952) used 

 the potential of the cell AgCl(s) : Cl~ : : Ag + : Ag for end-point detection and 

 control of titrant delivery. Although satisfactory from the standpoint of 

 precision and accuracy of results, excessive maintenance requirements caused 

 abandonment of the device after several years' use. 



A procedure described by Bather and Riley (1953) utilizes potentiometric 

 end-point detection, employs relatively simple equipment and gives results of 

 higher precision than the Knudsen titration. These authors recommend a weight 

 titration procedure for results of high precision (standard deviation 0.001 % in 

 chlorinity) or a volumetric titration for somewhat lower precision (standard 

 deviation 0.005 % in chlorinity). With the weight titration, up to forty-five 

 samples per day can be processed and with the volumetric procedures up to 

 ninety samples per day. 



Two methods have been used to establish the chlorinity of Primary Standard 

 Sea-Water by the International Council for the Study of the Sea (Jacobsen and 

 Knudsen, 1940). From 1902 to 1938 the chlorinity of new batches of Standard 

 Water (20 different preparations were made) was established by comparison 

 of each new batch with a previous batch. The comparisons were made using the 

 Volhard titration as modified by S. P. L. Sorensen. In 1938 a new batch of 

 Standard* Sea- Water (labeled "Urnormal 1937") was analyzed by the usual 

 comparison with the previous batch, and also by direct comparison with silver, 

 according to the new definition of chlorinity. 



The development of new high -precision instruments for the indirect measure- 

 ment of chlorinity suggests that attention should now be paid to the methods 

 that provide a direct measure of chlorinity, for these measurements must 

 provide the basis for calibration and standardization of the indirect methods. 

 The precision of the widely used Knudsen titration, ± 0.01 Cl% , is not adequate 



