SECT. 1J CHEMICAL INSTRUMENTATION 115 



in studies in which waters ranging in conductivity from that of river-water to 

 that of sea-water were used (Pritchard, 1959). This experience together with 

 repeated recalibration and standardization indicates that salinometers of this 

 type are extremely stable over long periods. 



A laboratory model salinometer with a transformer-type conductivity cell 

 has been constructed by Brown and Hamon (1961). This instrument has several 

 features not found in other salinometers. In addition to high long-term 

 stability, temperature-measuring and compensating circuits eliminate the 

 need for a thermostated bath for the cell, and permit samples to be measured 

 as quickly as they can be introduced into the cell- and bridge-balancing units 

 adjusted. The instrument has a sensitivity of 0.0004% o in chlorinity and a 

 precision of 0.001 % . The electronics have a low power requirement and can be 

 readily adapted to battery operation. The device is about the size of an ordinary 

 brief case and weighs approximately 25 lb. The instrument is to be manufac- 

 tured commercially and will sell for less than $1500. 



Three of the laboratory-type salinometers mentioned above, the Schleicher, 

 the Brown and Hamon and the Paquette instruments, were compared in a 

 recent test in which twenty-four sea-water samples with a salinity range of 

 33.88% to 34.78% were analyzed by each instrument. The means of the 

 twenty-four samples were 34.393% , 34.394% and 34.39% respectively. The 

 Paquette instrument will provide results significant to the third decimal place ; 

 however, because of a convention in reporting salinity by the organization that 

 performed the analyses, the third figures were omitted in their results. The 

 maximum deviation of the Schleicher and the Brown and Hamon results was 

 0.007% o and the standard deviation 0.003% . 



The high precision capable with the conductivity- bridge salinometer provides 

 the means of measuring a property of sea-water much more precisely than has 

 been possible in the past. The accuracy of corresponding chlorinity measures 

 may, however, not be as high as the precision in the relative conductivity 

 measures. This will depend upon the constancy of the chlorinity-conductivity 

 relation in various batches of Copenhagen Standard Sea-Water, and the method 

 of calibration with sub-standard sea-water. Furthermore, as should be apparent 

 from the discussion below, increased precision and accuracy in the measurement 

 of chlorinity will not increase the accuracy of computed salinity values. 



Forch, Knudsen and Sorensen (1902, p. 116) give the definition of salinity as : 

 "Salinity (Salzmenge) is defined as follows : The weight of dissolved solid 

 material found in 1 kilo of sea water, after the bromine has been replaced by an 

 equivalent quantity of chlorine, all the carbonate converted to oxide, and all of 

 the organic matter destroyed". 



Forch, Knudsen and Sorensen (loc. cit.) measured both the chlorinity and 

 salinity of nine samples [selected from those used to provide the primary 

 information upon which the Knudsen Hydrographic Tables (1901) were con- 

 structed]. From the results they computed the regression equation, which is 

 the familiar relationship : 



8°L = 0.03+1.805 01%. (1) 



