177 



ted with cell design and electrode behavior must be solved before an instrument 

 having the reliability and precision of the Knudsen method can be constructed. 



The use of high frequency techniques permits the measurement of the di- 

 electric properties of a solution in a cell in which the solution does not come in 

 contact with the electrodes. The measurements can be interpreted in terms of 

 the concentration of substances in solution. Jensen and Parrack (1946) de- 

 scribed an instrument and procedures for a variety of titrations. A Survey of 

 High Frequency Oscillators and Their Chemical Application by Flonn and Elving 

 (1950) describes the state of the art up to October 1950. S'brief description of 

 the commercially available Sargent Model V Chemical Oscillometer is given by 

 Muller (1952). 



The application of high frequency techniques to the direct determination 

 of chlorinity appears to be limited by the lack of sensitivity of relatively low fre- 

 quency oscillators and the lack of stability of high frequency oscillators. The 

 Texas A. and M. Research Foundation, with sponsorship from the Bureau of 

 Ships, is making an Investigation of High Frequency Oscillometric and Bridge 

 Methods of Measuring Conductivity of Solutions. 



When stable oscillators can be constructed of high enough frequency to 

 permit measurements with the desired precision, it would appear that many of 

 the troubles encountered with electrodes in the older type of conductance meas- 

 urement will be overcome. 



Some of the subjectiveness of the Knudsen titration lies in the visual de- 

 termination of the chromate end point. The substitution of a potentiometric 

 determination for the visual would, in principle at least, improve the situation. 

 Potentiometric titrations are common in analytical chemistry. The potentio- 

 metric titration of halides with silver nitrate using a silver, silver halide indica- 

 tor electrode is described in many quantitative analysis texts and references. 

 It is natural, then, to attempt to adapt these methods to improve the Knudsen ti- 

 tration. 



Lingane (1948) described an automatic potentiometric titration apparatus 

 in which the potential of a silver, silver halide electrode vs. a saturated calomel 

 electrode was made to control the flow of a silver nitrate solution during the titration 

 of a chlorideor iodide solution, and to stop the delivery of the titrant at the equival- 

 ence point. As the result of the analysis of several chloride solutions , he concluded, 

 "It is evident that the automatic titrations are fully equal in both precision and accura- 

 cy to the manual methods of titrating chloride ion with silver ion. " 



Carritt, Rakestraw, Snodgrass and Isaacs incorporated these principles 

 in the design of an automatic titrator made especially for the chlorinity titration. 

 The instrument was called the Automatic Chlorinity Titrator (ACT-I). The con- 

 struction and testing of the instrument has been carried on recently by Folsom, 

 Snodgrass and Isaacs at the Scripps Institution of Oceanography. 



Prototypes of the instrument have been in routine operation at the Scripps 

 Institution of Oceanography for the past three years. Changes in design have 

 been made to eliminate undesirable features that appeared after continued use. 

 Modifications of the basic instrument are recorded in Scripps Institution of Ocea- 

 nography. Pictures and a block diagram of the instrument are shown in figures 

 5 and 6. 



The heart of the ACT-I lies in the electrode reactions and their associated 

 potentials. If a piece of silver wire coated with silver chloride is in contact 

 with a solution in which chloride ions are being precipitated as silver chloride, 



