of the Electrical Conductivity of Mica. 203 



the readings at 16° CL which were obtained after the mica 

 had been heated to 227°*5 C, were rather too high. As, 

 however, these very small readings are liable to very large 

 relative errors if the voltage is not quite constant, this cannot 

 be regarded as definite evidence of any permanent change 

 in the mica, especially as the readings at 166° C, which were 

 the next in chronological order, were quite normal. 



The variations in the values of corresponding to the 

 observations plotted in fig. 3 are very large. For the lowest 

 point plotted for 50°*5 0. the value of C was 0'058 micro- 

 ampere per sq. cm., while for the highest point plotted for 

 227°*5 C. it was 4*57 microamperes per sq. cm. This was 

 about the largest current that could be obtained, and to do so 

 necessitated driving the Wimshurst machine up to the limit 

 of safety. 



Discussion of Results. 



On comparison of the figures given above with those given 

 in the previous paper for another sample of mica, it will be 

 noticed that there is a very great difference in the value of A 

 in the two cases. The value previously given for a tempe- 

 rature 9° C. was 'd'286, that found for the present specimen 

 for this temperature from the formula is only 5*245; so that 

 for a given field the current through the former specimen 

 was about 110 times that through the present one. If this 

 ratio holds for weak fields, the specific resistance of the 

 present system in such would be about 1*5 x 10 17 ohm cms. 

 Although this enormous difference in the resistivities of the 

 two: specimens is very remarkable, it does not appear quite so 

 striking when the very great effect of temperature is taken 

 into account. We find from the formula that the present 

 specimen would have the same resistivity at about 109° C. 

 that the former one had at 9° 0. The relative effect of rise 

 of temperature was about the same in the two cases. The 

 two specimens came from different sources, the former being 

 taken from a mica gas-chimney, while the latter was taken 

 from a packet of mica sheets. The original sources in the 

 two cases are not known. 



On the other hand, the close agreement in the two values 

 of B is remarkable, the figures being 0*794 and 0*787. This 

 seems to suggest that this coefficient may have some funda- 

 mental significance, though the fact that it is independent of 

 the temperature is opposed to the suggestion, previously 

 made, that it was connected with Maxwell's law, unless we 

 make the apparently improbable assumption that the mean 



P2 



