Oscillations in Straight Wires and Solenoids. 271 



coil ; but with a condenser of comparatively large capacity 

 connected to the ends of the coil, the wave-lengths of the 

 short free oscillations are not affected to any appreciable 

 extent by changing the capacity of the condenser, and are 

 thus distinguished from the harmonics of the principal 

 oscillation. 



When the latter is maintained in a circuit by means of a 

 valve, no short-wave oscillations were observed, except the 

 harmonics which changed in wave-length with the capacity. 



It thus appears that the principal oscillation and one of 

 the shorter free oscillations are not maintained simultaneously 

 by the valve, but it is possible to arrange the valve con- 

 nexions in such a way that the principal oscillation with its 

 harmonics are absent, and a short wave is maintained in one 

 of the free modes of oscillation of the coil. 



The following example shows ho a' the mode of oscillation 

 of an ordinary circuit may be changed by altering the valve 

 connexions. The circuit consisted of a condenser adjustable 

 from '2 to 2 millimicrofarads, connected to the ends of a 

 solenoid 14 centimetres long having tappings dividing the 

 coil into eight sections of 35 turns each. When the anode 

 of the valve was connected through the high-tension battery 

 to one end of the coil, the grid to the other end, and the 

 filament to the centre, an oscillation of 4000 metres wave- 

 length was maintained with a condenser of 1*4 millimicro- 

 farad capacity, the first twelve harmonics were observed 

 •directly by a wave-meter, and the wave-lengths of the 

 system changed continuously with the capacity of the con- 

 denser in the ordinary manner. No other oscillations were 

 observed. When the number of sections of the coil between 

 the valve-connexions were reduced to one section between the 

 anode and the filament and two between the grid and fila- 

 ment, no long oscillation was observed, but a wave of 318 

 metres and its first two harmonics were indicated by the 

 wave-meter which were not affected by changing the con- 

 denser in the oscillatory circuit from *2 to 2 millimicrofarads. 



7. A complete investigation of the free oscillations of a 

 coil, which would apply both to long and short solenoids, 

 would be very complicated, and the investigations which 

 have been made on this subject refer principally to some 

 definite cases. Thus the wave-lengths of short solenoids 

 in which the length does not exceed six times the diameter 

 have been investigated by Drude *. The series of free oscil- 

 lations of long solenoids has been investigated by Fleming t, 



* P. Drude, Ann. d. PJu/s. iv. 9, p. 29,'5 (1902). 

 t J. A. Fleming, Phil. Mag. (0) Oct. 1904. p. 417. 



