204 Temperature Variation of the Conductivity of Mica. 



energy of the electrons responsible for the conduction 

 current is independent of the temperature, and attribute the 

 great increase in conduction current with temperature entirely 

 to a diminution in the energy required to escape from the 

 atom. 



Before attempting to give any theoretical justification for 

 the comparatively simple law found connecting C with X 

 and T, which may be merely an empirical formula that 

 applies within the limits of error over the range covered, it 

 seems to be desirable to extend the observations to higher 

 temperatures and other dielectrics. One of the great hin- 

 drances to accuracy in the observations is the difficulty of 

 obtaining a sufficiently steady high voltage source. For 

 mica pressures up to about 5000 volts are required if the 

 sheet is not to be inconveniently thin. For other dielectrics 

 much higher pressures will probably be required, as the 

 sheets will have to be much thicker. For higher tempe- 

 ratures the conduction-current will be much larger, and a 

 more powerful machine will be necessary. It is hoped to 

 continue the observations on the above lines in the near 

 future. 



Summary and Conclusion. 



The results confirm the formula C = aX<? 6x , already given. 

 For the specimen previously used a was approximately pro- 

 portional to the 16th power of the absolute temperature over 

 a range from 0° C. to 47° C. and b independent of the tem- 

 perature. For another specimen a was proportional to the 

 15*47th power of the absolute temperature from 50 o, 5 C. to 

 229°'5 C. and b independent of the temperature. Although 

 the values of a were widely different in the two cases, the 

 values of b were almost identical. The fact that b is inde- 

 pendent of the temperature does not agree well with the 

 suggestion that the exponential term is due to a distribution 

 of electronic energies in accordance with Maxwell's law. 



It also appears from the observations that the dielectric 

 constant does not vary much up to 230° C. 



In conclusion I must again express my indebtedness to 

 Mr. H. Thrift, F.T.C.D., for his assistance during the course 

 of the work. 



Physical Laboratory, 

 Trinity College, Dublin. 

 June 13, 1917. 



