154 Mr. C. S. Whitehead on the Effect of 



This value, while differing but little from that of p for the 

 condensers, spark-gap, and induction-coil at the oscillator, 

 differs in the direction that one might anticipate. It also 

 supports the value previously found for p, in that p though 

 greater than p is still distinctly less than unity. It will be 

 seen from the end of the article in ' The Electrician/ April 

 9th, 1897, that Mr. Heaviside scarcely expected the long- 

 wave theory to be valid for the high-frequency waves here 

 used. 



Conclusion. — The chief results arrived at may be thus 

 summarized : — 



1. The value of the attenuation constant a for the line 

 experimented upon and the waves in use is of the order 

 0'000013, where the amplitude of the waves is affected by 

 the factor e~ ax in traversing x cm. of the line. The energy 

 of the waves is thus reduced to half by traversing about 

 250 m. of the line. 



2. The reflexion coefficient applicable to the primary 

 oscillator and induction-coil as used to generate the waves is 

 of the order 0"69, or rather more than half the energy is 

 lost on reflexion. The reflexion coefficient of a precisely 

 similar arrangement of condenser-plates but devoid of spark- 

 gap and induction-coil is rather greater but still distinctly 

 less than unity. On the long-wave theory the second of 

 these reflexion coefficients would be precisely equal to unity 

 and the first of them practically so. 



I have to thank Mr. H. A. Ratcliff for help at the 

 beginning of this work, and especially Mr. It. S. Willows for 

 his valuable assistance at the later experiments and their 

 reduction. 



University College, Nottingham, 

 May, 1897. 



XXIII. The Effect of a Spherical Conducting Shell on the 

 Induction at a Point in the Dielectric outside due to an 

 Alternating Current in a Circidar Circuit in the Dielectric 

 inside, the Axis of the Conductor passing through the Centre 

 of the Shell. By C. S. Whitehead, M.A* 



(1) TT is well known that if there be an alternating cur- 

 J. rent in a primary circuit sounds will be heard in a 

 telephone in a secondary circuit, even if the latter be at a 

 considerable distance from the former. If the sounds are 

 preconcerted, signals may be, and in fact have been, trans- 



* Communicated by the Physical Society : read June 11, 1897, 



