568 



Prof. E. Taylor Jones on 



deflexion method. The value so found was 20*4 henries for 

 a primary current of 5 amperes. By comparison with this, 

 the same method being used, values of L 21 were then obtained 

 for other values of the primary current i Q . The results are 



given in Table I. 



Table I. 



*0* 



i L 21 , 



2 0' 



1*21. 



amperes. 



henries. 



amperes. 



henries. 



0195 



12-9 



! 2-49 



20-0 



0-395 



147 



3-00 



202 



0-57 



160 



3-52 



20-3 



0-75 



169 



4-00 



204 



1-00 



17-8 



5-00 



20-4 



1-25 



18-6 



6-86 



20-3 



1-50 



19-0 



8-01 



199 



1-75 



19-2 



10-06 



190 



200 



19-6 



1104 



18-4 



It will be seen from the table that L 21 has its maximum 

 value at 4 or 5 amperes, and that over the range 3 to 

 7 amperes it does not vary by more than 1 per cent. As 

 most of the experiments described below were within this 

 range, the value 204 was adopted for L 21 in working out 

 the theoretical expression (1) for the secondary potential. 



The other method of determining L 21 is to take the value 

 of k 2 measured with a condenser of considerable capacity in 

 the secondary circuit, as described in the next section, in 

 which case 



V = WIIj 1 L 2 = 'L 2 21 I 1 L 1 L 2 . 



The values of L l3 L 2 being found from the periods of oscilla- 

 tion of the primary and secondary circuits when connected 

 to condensers of known capacity, L 21 can then be calculated 

 from 



L 21 = vPL 1 L 2 . 



The values found were 



P=-839, L a =-194 henry, L 2 = 2540 henries. 



Hence L 21 = 20*3 henries, 



which is almost identical with the maximum value found by 

 the other method. It may be concluded that up to the 



