Conductivity of some Dielectrics. 49£> 



heating the upper side of the plate with a Bimsen burner. 

 When the film was quite cold the upper mercury was poured 

 iu and connected to the high-pressure source, the central rod 

 being connected to the galvanometer. 



Tests were carried out at 15° C, the pressure being raised 

 until the film was pierced. The film was then cut out and 

 its thickness measured in various places with a micrometer 

 screw. This w T as not uniform as the lower mercury stood 

 higher in the centre than round the edge, owing to surface 

 tension, during ihe casting of the wax. Over a disk 1*5 cm. 

 in diameter the thickness varied from about 033 mm. near 

 the fracture, which w r as not quite central, up to 0*37 mm. 

 Further out the thickness increased rapidly up to several 

 millimetres at the edge of the lower mercury cup, which 

 was 2*54 cm. in diameter. 



The specimen was pierced by a pressure of 7600 volts 

 giving a maximum gradient of about 0*25 megavolt per cm. 

 The current passing through G, just before the failure of 

 the specimen, was about 4 x 10~ 4 microampere. As, however, 

 the ionization current through the air, when the specimen 

 was removed, was about 1/5 X 10" 4 the current through the 

 specimen was about 2 , 5xl0 -4 . Assuming the effective 

 area of the specimen to be 2 sq. cm., and its mean thickness 

 0*35 mm., the conductivity works out as 6xl0 -4 under a 

 gradient of 02. The conductivity is so small that there is little 

 chance of observing its rise under the very limited gradient 

 that the wax will stand, as the pressure of the source 

 employed cannot be maintained constant enough to allow 

 of an electrometer being used for measuring the conduction 

 current. 



Shellac. 



The same apparatus was used, a little shellac dissolved in 

 absolute alcohol being poured on the surface of the lower 

 mercury and the whole allowed to stand for about two 

 months. Unfortunate^, the alcohol on the surface evaporates 

 very rapidly, leaving a solid crust through which the alcohol 

 from lower layers diffuses very slowly, so that it is probable 

 that the film when tested contained appreciable quantities of 

 alcohol, though when examined after fracture it was quite 

 hard throughout. 



The film was unfortunately pierced by a pressure of about 

 3000 volts before very many readings had been obtained. 

 The average thickness was found to be 0-125 mm., so failure 

 occurred between 0'2 and 0*3 megavolt per cm. The tew 

 readings obtained for the conduction current at 14 'I (. . 



