354 
PHYSICS: L. L. NETTLETON 
Proc. N. a. S. 
side parts accessible for modification, replacement of burned out filaments, 
etc. Also, more power was used than by other investigators, electron 
currents as high as 300 milamperes and at voltages up to 700 being used at 
times. 
The electrical connections were made as indicated in figure 1, except 
that the filament heating circuit is not shown. The Lecher wires, LL, were 
closely coupled to, but not in metallic contact with, the grid and plate leads, 
which were brought straight in through the sides of the tube. A sliding 
bridge, B, across the Lecher wires, carried a crossed wire thermocouple 
which was connected to a sensitive galvanometer by long, twisted leads. 
This thermocouple served to measure the strength of oscillations, and by 
jgXMoeoepec-'^ To. gpT 
/WW\AAA/WV\M 
FIGURE 1 
sliding the bridge along the wires, the points of maximum deflection gave 
the loops of the standing waves, and the distance between these points 
gave the half wave-length of the oscillations. In the discussion which 
follows, by strength of oscillations is meant the deflection of the ther- 
mocouple galvanometer when the bridge position is adjusted so that this 
deflection is a maximum. 
Oscillations of the Barkhausen type, at wave-lengths of from 50 to 200 
cm., were obtained. The unusual characteristic of a tube when oscil- 
lating in this way is that there is a negative current to the plate, which 
