including the cleanness of the electrode surfaces, and can be expected 

 to change with time. Unless the polarization resistance is small in 

 comparison with the total cell resistance, the results obtained will not 

 be reliable. The ratio of polarization resistance to cell resistance 

 depends upon cell geometry, electrode material, frequency, tempera- 

 ture, and conductivity. For laboratory studies of electrolytic conduct- 

 ivity, the classical method of reducing the effect of polarization has 

 been to use "platinized 11 platinum electrodes (electrodes coated with a 

 layer of finely divided platinum deposited electrolytically) and to use a 

 frequency of atleast 1,000 c.p.s. "When the greatest accuracy is required, 

 the platinized electrodes must be treated very carefully: they must be 

 kept wet at all times, must not be touched, and must be kept free from 

 oil and grease and from organic contamination. These requirements 

 have made use of such electrodes infield instruments difficult. However, 

 some of these difficulties can be overcome by providing interchangeable 

 electrodes, by replatinizing the electrodes at regular intervals, and by 

 checking frequently against standards. 



Laboratory measurements of the electrical conductivity of a solution 

 usually are made with a two-electrode cell. When this type of cell is 

 immersed in a solution the measured conductance includes the conduct- 

 ances of the paths both within and external to the cell. Since the cross- 

 sectional area of the external shunt path may be very large for a cell 

 submerged in the sea, the external conductance will be large compared 

 to the conductance within the cell. A cell with an electrode at each end 

 may be effectively short-circuited when used in this way. 



To avoid this difficulty a cylindrical cell with three coaxial electrodes 

 has been designed. One electrode is at the center of the cell and one 

 at each end. The two outer electrodes are connected together, and the 

 potential is applied between them and the center electrode. The lead to 

 the center electrode is insulated, and approximately half of the current 

 in this lead flows through the sea water to each end electrode. Inasmuch 

 as the two paths from the center electrode are in parallel, the measured 

 conductance is the sum of the conductances of these paths. Since the 

 two other electrodes are short-circuited together, they are at the same 

 potential, and no current will flow between them through any external 

 shunt path. Thus, the measurements are independent of external shunt 

 conductances, and the cell constant may be determined without the need 

 for duplicating the actual conditions of use. 



Derivation of salinity or chlorinity directly from measurements of 

 conductivity requires use of either a servo system to compute chlorin- 

 ity from measured values of temperature and conductivity or compen- 

 sation for. the effects of temperature. Such a computing servo system 



IV- 3 



