Wave-trains through Layers of Electrolyte, 533 



to the spark-gap, S, by wires, AS, A'S, and was worked by a 

 Ruhinkorff coil, R. Opposite the plates A A' stood similar 

 plates B B', from which a pair of long wires BFD were led 

 off, stretched parallel to each other. These wires serve 

 simply as guides for the radiation, which travels straight up 

 the space between them and through the electrolyte x u x 2 . 



The principal dimensions of the oscillator &c. were as 

 follows : — 



Diameter of plates A A', B B' . . . 40 centim. 



Distance A B, A' B / 30 „ 



Length of wire A S A' (diam. 2 millim.) 200 „ 

 Wave-length \ 900 „ 



The wires BFD (about 1 millim. diam.) were spanned 

 6 centim. apart. If these wires be made too short, a wave- 

 train emitted from B W may be reflected at the electrolyte x 1 

 or the bridge D and arrive back at B before the primary has 

 finished oscillating. If this occur, alterations in the state of 

 the secondary may react on the primary somewhat as in an 

 alternate-current transformer, and inconvenient interference 

 takes place. But if Bx x be made longer than half the effective 

 length of the wave-train, the reflected Avaves will not reach B 

 until the primary has practically finished oscillating, and the 

 latter can then know nothing of any alterations in the second- 

 ary at or beyond x l% Such a reaction of the secondary on the 

 primary was noticed to a very serious extent by Herr von 

 Geitler* with a Blondlot oscillator f. 



In my apparatus the wires were led out at Fi in a circuit 

 of about 50 metres circumference round a garden next the 

 laboratory. They re-entered the room at F 2 , and were then 

 run vertically through the vessel containing the electrolyte 

 x u x 2 . A second length of 50 metres, F 3 F 4 , round the 

 garden, completed the circuit, the wires re-entering the room 

 at F 4 and being bridged at D. The electrometer was con- 

 nected at E a quarter wave-length (2*25 metres) from the 

 end of the wires, at the crest of the standing waves formed 

 by interference of the direct and reflected trains. If we 

 accept, at least as an approximation, the value of the primary 

 damping determined by Bjerknes (loc. cit.) for his apparatus, 

 these dimensions should suffice to exclude any sensible reaction. 



The electrometer was the one employed by Bjerknes for 

 his researches in the same laboratory. It is a simple 

 quadrant-electrometer with only one pair of quadrants and an 



* Doctor-Dissertation : Bonn, Jan. 1893, p. 22. 

 t Comptes Rendus, cxiv. p. 283 (1892). 



