o66 Prof. E. Taylor Jones on 



theoretical and experimental curves for the coils with air- 

 core and with the core of iron-filings, fair agreement in the 

 case of the iron- wire core. 



In the present paper further tests are described of the 

 applicability of the theory to the case of an induction-coil. 

 In these, smaller primary capacities were used, and the 

 constants of the circuits were determined in actual working 

 conditions. The effect of varying the primary capacity is 

 specially considered, and a view is put forward which differs 

 from that usually held as to the function and the most 

 effective value of this capacity. 



There are two things which should be borne in mind in 

 determining the constants of the primary and secondary 

 circuits of an induction-coil. The first is that, owing to the 

 variable permeability of the core, it is necessary to use mean 

 values of the induction-coefficients, which can only be 

 expected to give correct values for the secondary potential 

 over a limited range of values of the primary current. 



The other is that the induction-coil as generally used has 

 the terminals of its secondary coil connected to bodies of 

 very small capacity, e. <j. spark-gap terminals, the electrodes 

 of an exhausted tube, or an electrometer or electrostatic 

 oscillograph of very small capacity. In these circumstances, 

 the capacity of the secondary circuit is mainly distributed 

 along the secondary coil, and when the oscillations are taking' 

 place and no discharge is passing, the current is not uniform 

 throughout the length of this coil, but is greatest at the 

 central winding and very nearly zero at the terminals *. 

 The system is in fact similar to that of the oscillation 

 transformer, or Tesla coil, the theory of which was given by 

 Drude f. 



The expression "mutual inductance " is not appropriate to 

 such a system, but if (in Drude's notation) L ]2 is the co- 

 efficient of induction of the secondary on the primary, and 

 L 21 that of the primary on the secondary, then L 12 is less 

 than L 21 . L 12 is defined as the magnetic induction through 

 the primary coil due to the secondary current divided by the 

 value of the latter in the central winding. On the other 

 hand, the terminals of the primary coil are connected to a 

 condenser of considerable capacity, so that the current in 

 this coil is uniformly distributed. The coefficient L 21 is 



* As tlie current varies (in its fundamental oscillation) as the cosine 

 of an angle proportional to the distance from the central winding, this 

 distribution of current will be referred to as a cosine distribution. 



t Drude, Ann. d. Physik, xiii. p. 512 (1904). 



