Conductivity of some Dielectrics. 501 



In the case of mica the formula only applies to a gradient 

 of about 1*5. Above this it yields too high a value for 

 the conductivity, see curves figs. 3 and 4. 



Material. t. A. B. X m . R. 



Mica specimen A 14 4"S6 119 >3"0 14X10 15 



„ B 13 13-70 1-18 >29 l'TxlO 17 



,, 37 5-39 1-18 >2-9 4-lxlO 16 



„ 66 4-07 1-18 >3-0 SoxlO 15 



„ 109 4-88 1-18 >'2-8 1-3 xlO 15 



„ 128 3-27 1-18 >29 5'4xl0 14 



Glass (cover glass) 14 2'33 150 >0;7 4-2xl0 13 



, 29 1-18 133 >06 6-6xl0 72 



45 0-00 115 >0-6 , lOxlO 12 



58 057 105 >06 2-7X10 11 



74 1-23 091 c.0-5 5-9xl0 10 



Paraffin wax 15 — c. 0'25 c.4xl0 13 



bhellac from solutiun 15 c. 04 c. 2'7 c. 25 C.4X10 11 



„ by casting 12 — — c. 0D4 >4xl0 u 



Celluloid (Kodak film) 14 133 38 016 4-7x101° 



Summary and Conclusion. 



1. A method is described of testing the conductivity of 

 dielectrics when subjected to high continuous pressures, 

 produced with the aid of rectifying valves/and measured by 

 means of a rotating contact-maker. 



2. The previous results for mica are generally confirmed, 

 but it is shown that when the logarithm of the conductivity 

 is plotted against the potential gradient over the con- 

 siderable range now available, the isotherms exhibit distinct 

 curvature. 



3. In the case of the other dielectrics tested, the available 

 range is much less, and the isotherms are approximately right 

 lines. Grlass differs from mica, as in its case the slope of 

 the isotherms decreases with rise of temperature. 



In conclusion I wish to express my gratitude to Prof. W. E. 

 Thrift for his kind permission to continue this work in the 

 Physical Laboratory, Trinity College, Dublin, after I had 

 resigned my position on the laboratory staff. 



Eoyal Dublin Society, 

 May 23, 1921. 



