TRANSACTIONS OF SECTION A. G93 



SS'. Similar discs 30 cm. distant from the former. 



SADS'A'D'. Tile wires along which the electrical waves are propagated. 



BCB'C. The abnormal part of the wires, 



M. Middle point of same. 



SAM = 101m. 



MD = 63m. 



EE'. Electrometer, situated a quarter-wave length from — 



DD', the bridge forming the end of the wires. 



AA' = EE' = DD' = 8cm. 



Fig. 2. — Diagram of apparatus. 



BMC 



B' M' C' 



The needle of the electrometer is uncharged, and therefore turns in the same 

 direction whether E or E' is positive, and is thus able to give a deflection undis- 

 turbed by the high frequency used, viz., about 33 million per second. 



Although the interference phenomena under consideration care essentially 

 analogous to those of light in thin plates, yet the mathematical theory of the 

 latter will not suffice for our case. Because in the optical phenomena the ampli- 

 tude of the incident light may be assumed constant, whereas in the electrical 

 analogue the primary oscillator emits waves each of which is feebler than its 

 predecessor. It thus induces in the long wires a continually diminishing series of 

 waves which advance along the wires in the form of a damped train with its large 

 end leading. 



I have already published an elementary mathematical theory^ dealing with 

 the interference phenomena of such a wave train. The results of this theory when 



Fig. 3. — Enlarged view of the ' abnormal part ' of the wires. 



graphically exhibited, the lengths of the abnormal part being abscissje, and the 

 intensities of the transmitted waves being ordiuates, yield a damped wavy curve. 

 For the constants involved in the experiment hereafter described this theory gives 



' Proc. Roy. Soc, vol. 54, 1893, p. 9G. 



