elements has no effect and can be ignored o 



The calibration of the individual elements is independent of the 

 position of any of the other elements o Hence ^ each calibration curve 

 measures only -what takes place at the corresponding elements Further- 

 more, the calibration of each element is essentially linear » A typical 

 calibration curve is shown in Figure 2-Ao The calibration and op- 

 eration of the elements are independent of the solution's electrolytic 

 properties other than its influence on the resistance of each element « 

 The non-linearity of the calibration is slight and contributed to the 

 water to wire resistance relationship =, The calibration may be effected 

 also by progressive corrosion o Best results will be obtained in water 

 of constant temperatures 



It may be possible to eliminate the first set of transformers 

 and supply power to the set of powerstats with one transfoimer o It is 

 also possible to supply the power to the set of transformers with one 

 or more powerstat by connecting the 155-V line to the powerstat and 

 thence to the transformers » In both cases different powerstats or 

 transformers would be needed o The disadvantage to any such saving of 

 equipment is that the resulting maximum current available to the 

 galvanometer and elements is limited and the reading of any one element 

 cannot be made prorainento It is considered that the flexibility of 

 this system outweighs any saving of equipment. From an electrical 

 standpoint^, it is better to feed the transformers through the power- 

 stats hit no difference in the two methods has been noticed. 



Sources of Errors in Using Single Power Supplies for Multi-Elements 

 and Common Leads 



The possibilities of connecting a series of elements to galvano- 

 meters and installing STn.tching circuits for ease in handling in many 

 cases leads to erroneous results between calibration curves and actual 

 measurements The three common and elementary circuits found to be 

 subject to errors are presented in Figures l-Bj C, Do Ihe same equip- 

 ment is used in these figures as was used in Figure 1 unless otherwise 

 notedt, In each of the cases presented it is seen that there is more 

 than one method of completing the electrical circuit when using the 

 water as a patho This immediately induces the electrolytic character- 

 istics of the water into the problemo When the electrolytic properties 

 of the water solution are such that the resistance of the solution 

 is low and of the order of magnitude of the recording circuity then 

 errors are observed and the relative submergence of the elements 

 effects their mutual recording circuits o Figure 2-C is an example 

 of a calibration showing this effecto The error is due partly to 

 the polarization effects when the solution is a concentrated one. 

 However^ when the electrolytic properties of the water solution are 

 such that the resistance of the solution is highy (reference is made 

 to tap water as compared to laboratory storage water which contains 

 a solution to prevent fungus growth) andj, hence j, is very large com- 

 pared to the recording circuitj then the errors disappear and the 



18 



