Potts — Rowland's New Method. 



Wi 



ordinates. It appears that at ordinary laboratory temperatures 

 a variation of *1° C. will cause a change of about 1 per cent in 

 the value of A/T and consequently in A for a given T. From 

 this it appears that the slight variation noted above in the con- 

 stancy of A/T would be easily accounted for by the changes 

 of temperature due to the current or the gradual changes due 

 to changes in the temperature of the room, since a series of 

 observations usually occupied three hours or longer. 



Table YII gives the results for the two \ microfarad con- 

 densers described above connected in parallel ; and in fig. 5 the 

 results are plotted as before. 









Table VII. 















R 4 = 303-3 



r = 



4811 

 Calculated. 











R 2 



R>3 



Bi 



B'i 



A 



T 



Date, etc. 













(R'i-Ri) 





11-29-99 



1 



409-7 



2984- 



34-07 



41-66 



7-59 



•0271 



105 



2 



409-7 



2723- 



34-07 



45-59 



11-52 



•0345 





3 



409-7 



1807- 



34-07 



68-79 



34-72 



•0752 





4 



409-7 



1604- 



34-07 



77-49 



43-42 



•0893 





5 



409-7 



2454- 



34-07 



50-66 



16-59 



•0428 





6 



409-7 



3010- 



34-07 



41-30 



7-23 



•0239 





7 



410-6 



3157- 



34-04 



39-38 



534 



•0187 



It was necessary to use the two in parallel in order to get 

 sufficient current through the fixed coils of the electro-dyna- 

 mometer. This was especially true for long periods, as then 

 the impedance of the condenser increased and at the same time 

 the available electromotive force from the dynamo decreased. 



Capacity of a Condenser which shows Electric Absorption. 



From the theory of electric absorption as based on the hete- 

 rogeneous nature of the dielectric it appears that there should 

 be a variable value of the capacity of such a condenser depend- 



5 



40 



30 



20 



10 ■ 



in seconds. 



Ordi nates represent /f in ohms. 



o-o -01 



•02 -03 -04 -05 -06 '07 08 



•09 



