SECT. 1] CHEMICAL INSTRUMENTATION 113 



for purposes of calibrating the new conductivity bridge salinometers which 

 have a precision equivalent to ±0.001 Cl% and better. The Bather and Riley 

 (loc. cit) potentiometric method has a precision comparable to that of the 

 salinometer and is being used in a fundamental study of the conductivity- 

 chlorinity relationship of sea-water (Cox, 1960). 



If we accept the idea that reference and calibration methods should have a 

 precision and accuracy an order of magnitude greater than that in the system 

 being studied, only the Volhard titration and the direct comparison with silver 

 following classical atomic-weight techniques are adequate for calibration 

 purposes and for the direct comparison with silver for establishing the value of 

 Standard Sea-Water. 



In summary, the uncertainty in chlorinity for a sample of 19.000% o associated 

 with various methods are: Knudsen titration ±0.01% o CI; Bather and Riley 

 ± 0.001 % CI; Volhard ± 0.0004 % CI; and direct comparison with silver 

 ±0.0002% CI. Laboratory-type conductivity salinometers show an un- 

 certainty equivalent to approximately ± 0.0005 to 0.002% CI. 



In recent years the use of so-called "salinometers," in which electrical 

 conductivity is used to obtain a measure of chlorinity, has become widespread, 

 and, in fact, appears now to be the preferred method in most laboratories in 

 which large numbers of samples must be analyzed both at sea and ashore. In 

 principle, the measurement of electrical conductivity to obtain chlorinity has a 

 distinct advantage over the classical titration method in that conductivity 

 measurements can be instrumented to provide a continuous measure with 

 depth, whereas titration gives discrete values only at sampling depths. In 

 practice, two types of salinometers have been used, one to provide continuous 

 in situ measures, the other to measure samples obtained from Nansen bottles. 



The Wenner Bridge (Wenner et al., 1930) appears to have been the first 

 conductivity device having the stability, sensitivity and accuracy required for 

 routine measurements of chlorinity. The reliability of this instrument, as well 

 as of several more recent "salinometers" of similar design, depends to a large 

 extent on a unique feature in the arrangement of the bridge circuit. One arm 

 of the bridge consists of a conductivity cell into which the sample to be measured 

 is placed. Another arm of the bridge consists of a conductivity cell, as nearly 

 identical to the sample cell as is practical, which contains either Copenhagen 

 Standard Sea-Water or a sub-standard sea-water of known chlorinity. The con- 

 ductivity cells and critical circuit components are in a thermostated bath. The 

 associated electrical circuits provide the means of measuring the difference in 

 conductivity between unknown and standard. Calibration is by measurement 

 of samples of known chlorinity. 



This method of comparing a sample to a standard eliminates the need for 

 extreme precision in temperature control. For example, Paquette (1958) noted 

 that as long as reference and unknown salinities are all at the same tempera- 

 ture, departures from the temperature at which calibration was performed 

 introduce an error of approximately 0.005% o in salinity per degree Celsius 

 change in temperature per 1% difference between measured and reference 



