﻿318 Dr. P. E. Shaw on the Disruptive Voltage of Thin 



Przibram (Akad. Wiss. Wien, Sitz.-Ber. Nov. 1904), and 

 Voege (Elektrotechn. Zeitschr. Dec. 1904) have tested the 

 relationship for spark-lengths greater than -^ mm. 



In the following paper the P.D. varies from 25 volts to 

 400 volts, and the corresponding spark-lengths vary from 

 jJ) micron (micron = yq&q mm to about 10 microns. 



Recently a book on the subject has been published giving 

 useful information ('The Insulation of Electric Machines/ 

 by H. W. Turner and H. M. Hobart, Whitaker & Co., 

 Paternoster Sq. 5 E.G.). It is made clear in this book that 

 water, acids, and alkalies are very destructive of the insulating- 

 powers of oils ; thus Skinner has shown that ~ r } Q per cent, of 

 water reduces the dielectric strength of transformer oil to half 

 its value when quite dry. 



In the present work all the commercial liquids used were 

 shaken up and allowed to remain over potassium carbonate, 

 filtered, and then heated to 110° for several hours, then kept 

 in a glass-stoppered bottle till used. 



Method, 



In a research by the present writer on the sparking- 

 distance in air (loc. eit.), the electric micrometer with 

 levers was used to measure spark-length ; in the present case 

 the electric micrometer, without levers, is used. It has been 

 already described (Proc. Roy. Soc. April 1906) for measuring 

 gauges; it has proved very convenient for the present 

 experiments. 



The only parts of the apparatus which need be here men- 

 tioned can be understood from fig. 1. Two spindles, A and 

 B, are in one line, and can be worked to or from one another 



Fio-. 1. 



by micrometer-screws not shown, and their distance apart 

 can be read and controlled to ^ yu, (i. 



•J 100,000 



mm.). The 

 divisions on the micrometer-head show movements of \fi } and 

 these, when viewed by a microscope, can be each subdivided 

 into 100 parts. 



