578 



Prof. E. Taylor Jones on 



was then unusually large, and these large amplitudes were 

 alone selected for measurement. The iron-steel interruptor 

 gave much more variable results than the other, but at its 

 best was quite equal to the mercury interruptor. 



A number of curves was first taken for different values of 

 the primary current, in order to see whether the amplitude 

 of the curve is proportional to the square of the current. 

 Table VI. contains a set of results for Case I., the current 

 being given in amperes and the amplitude A of the curve in 

 millimetres. 



Table YI. 



V 



A. 



VX/i . 



1-83 

 2-81 

 3-75 

 4-91 



3-1 



7-8 

 14-0 

 23-3 



•962 

 •994 



•998 

 •983 



4-91 

 6-41 



10o 

 169 



•660 

 •641 



The last two plates were taken after the sensitiveness of 

 the electrometer had been reduced by tightening the strip. 

 In cases where the deflexion on the plate is over 3 cm. the 

 photograph is generally very faint, and such large angular 

 deflexions of the reflected ray probably cease to be propor- 

 tional to the deflecting force. 



The numbers in the third column of Table YI. show that 

 over the range of this set of readings the maximum potential 

 Y m at the electrometer is within 4 per cent, proportional to 

 the current interrupted in the primary coil. Similar results 

 were obtained, within the same limits, from the plates taken 

 for Cases II. and III. The numbers given in the table also 

 point to a maximum value of the ratio Y m /i , in the neigh- 

 bourhood of 4 amperes, a result which might be expected 

 from the manner in which L 9i varies with z' , as shown in 

 Table I. 



A set of three plates was next taken for the three cases 

 with the same primary current, in order to see whether the 

 maximum potentials are proportional to the calculated values- 

 The results are given in Table VII. in which Y c represents 

 the calculated value in volts of the maximum potential. In 

 each case i was 3*0 amperes. 



