[m'lknnan] 



RADIOACTIVITY OF METALS 



41 



Volta eifect and of approximating to a measure of the contact differ- 

 ences of potential for a series of metals. 



With the rod D in position a difference of potential would exist 

 between the cylinders A and B. Its withdrawal would be followed by 

 a gradual equalization of these potentials which would result in a 

 charge being communicated to the connecting wire and the free 

 quadrants. This again would set up a current to the surrounding tube 

 C, and finally the free quadrants would attain a steady potential, the 

 current between the cylinders being equal to that between 'the con- 

 necting wire and its surrounding tube. The potential assumed by the 

 free quadrants under these circumstances, while approximating to, 

 would be less than the contact difference of potential for the metals 

 used in the two cylinders. 



The values obtained in this way for any two given metals, while 

 differing in sign, should be numerically the same, whichever metal was 



S 



e 



used for the inner or outer cylinder. But on trial with a number of 

 pairs of metals, it was invariably found that the readings differed 

 when an interchange was made of the metals in the cylinders. This 

 is illustrated by the curves in Fig. III., which give the results for the 

 metals, zinc and copper, the upper representing the negative potentials 

 acquired by the quadrants with an inner cylinder of zinc, and the 

 lower the positive potentials with one of copper. The final potentials, 

 it will be seen, are 527 and 304 millivolts respectively. 



This result finds its explanation in the metal effect described above. 

 We have shown that, with zinc cylinders, the inner, in virtue of this 



