338 



Scientific Proceedings, Royal Dublin Society. 



The value given for red fibre is only a rough representation of its specific 

 resistance at ordinary temperature. It may be much less owing to absorption 

 of moisture from the atmosphere, or greater if in a very dry condition. The 

 values given in Section 1 of Table 2 for sheet fibre are much smaller than 

 1800, which is the mean value of the two tubular samples whose results are 

 given in Tables 10 and 11. 



A test was also made with the white and black plates of Section 1 of this 

 table, and the following results obtained : — 



A is the case in which mercury electrodes were used ; in B the top electrode 

 was water, and the bottom one mercury ; while in C both were water. 



It may, therefore, be deduced that the unprepared surface of erinoid 

 gives the sample a greater specific insulation resistance than when machined? 

 as in the case of the tubes ; that the thinner samples have a higher specific 

 resistance than the thicker ones; that the water electrode gives a closer contact 

 surface than one of mercury; that the blonde erinoid has a considerably 

 higher specific resistance than the other varieties of erinoid, and this in view 

 of the results of Sections 2 and 3 of Table 2 may be attributed to a much 

 greater skin effect or inferior contact with its mercury electrodes than that 

 possessed by the other samples of erinoid, or to a combination of both causes. 

 Finally, that erinoid has a higher insulation resistance than the red fibre 

 tested. 



The insulation resistance of the samples was generally determined by the 

 megger method ; sometimes by the galvanometer method ; and the highest 

 values by the leakage method. 1 



In the latter a mica condenser of capacity two microfarads in parallel 

 with an electrostatic voltmeter was connected across the sample and the 

 arrangement charged with a direct voltage of about 100. 



"Whenever a test was made with each of the three methods a very close 



1 These methods are described in detail in Scient. Proc. Roy. Dubl. Soc, vol. xv. 

 (N.S.), No. 27, June, 1918, p. 292. 



