»li 



708 Prof. E. Taylor Jones and Mr. D. E. Roberts on 



of the inductances may be found which will serve for the 

 evaluation of the periods and amplitudes during at least a 

 portion of the time of discharge. This was borne in mind in 

 the experiments described below, and it was intended as fat- 

 as possible to determine the mean inductances corresponding 

 to the early portion of the discharge, since this portion 

 includes the maximum secondary potential which is the 

 quantity of greatest importance in an induction-coil. 



The inductances of the primary and secondary coils were 

 determined by electrical oscillation methods ; a condenser of 

 known capacity being connected to the primary (or secondary) 

 coil, and the curve of secondary potential being photographed 

 and measured by methods which have been described in 

 detail in former papers *. 



It was thought that by measuring that portion of the 

 damped oscillation-curve in which the current amplitude was 

 approximately the same as in the early part of the double- 

 period discharge, appropriate mean values of the inductances 

 would be obtained, since the iron core passed through the 

 same series of magnetic states in both cases. It was scarcely 

 to be expected that this plan would lead to very accurate 

 results, but it was hoped that it would suffice for the 

 purposes of an approximate estimation of the secondary 

 potential. 



The coupling coefficient of the two coils was determined 

 by finding the period of the oscillations of the secondary 

 circuit, connected to a condenser, with the primary open 

 and closed. The ratio of the squares of these periods is 

 then (1-P). 



The value of the expression (1) was worked out for two 

 induction-coils, a 12-inch spark coil by Max Kohl, and an 

 18-inch coil by the same maker. In the experiments with the 

 smaller coil a condenser was connected to the terminals of the 

 secondary coil ; otherwise the secondary capacity would have 

 been so small that one of the oscillations would be too rapid to 

 be accurately recorded by the oscillograph. In the case of the 

 larger coil it was unnecessary to connect a condenser to 

 the secondary, but the primary capacity was larger than is 

 generally employed with induction-coils. 



In these experiments the electrostatic oscillograph was 

 provided with a steel strip under great tension in order to 

 bring the amplitudes of the curves down to moderate 

 dimensions. The metallic parts of the instrument were well 

 sheathed in ebonite, and the terminals well separated, in 

 order to present sparking. 



* E. T. Jones, loc. cit. pp. 33-35; and Phil. Mag. August 1907. 



