572 



Prof. E. Taylor Jones on 

 Table III. 



Condenser. 



Frequency of secondary. 



Sparking Sbort-circuiting 

 method. method. 



A 



247-3 

 5190 

 630 

 989 

 1246 



245-9 

 517-8 

 636 



1026 



1266 



B 



C 



F 



none 



It will be seen that there is no great difference between 

 the results obtained by the two methods. Such differences 

 as exist are to be accounted for by the difference in the 

 circumstances oil the experiments. In the sparking method 

 the iron core is initially unmagnetized, and the oscillations 

 are accompanied by reversals o£ the induced magnetism. In 

 the short-circuiting method the core is initially strongly 

 magnetized, and as the magnetization decays the oscillations 

 produce small fluctuations in its intensity. The difference 

 between the two cases may be illustrated by saying that, if 

 N is the total flux of induction in the core and i the current 

 in the secondary coil, then in the sparking method the fre- 

 quency depends upon the coefficient N/f, whereas in the 

 short-circuiting method it depends upon dN/di, and the value 

 of this latter coefficient will be modified to some extent by 

 the hysteresis of the core. 



The circumstances of the sparking method are in closer 

 accord with those of the actual case of an induction-coil, and 

 the result obtained by that method, viz. F = *768, is therefore 

 adopted in the calculations of potential described below. 



Determination qfLiCi, L 2 C 2 , Ri/L 1; R 2 /L 2 . 



The expression (1) for the secondary potential was worked 

 out for three cases in which mica condensers, of capacities 

 stated to be 3, 2, and 1 microfarad respectively, were con- 

 nected across the primary coil. These values were, however, 

 not assumed, but rather the value of the product I^Ci was 

 determined for each case. This was done by coupling the 

 primary (with the condenser across its terminals) loosely with 

 a secondary of many windings to the terminals of which the 

 electrometer was connected. The most sensitive electrometer 



