THE NATURAL PERIOD OF LINEAR CONDUCTORS 



409 



and some time was wasted trying to get consistent results. Finally 

 a control meter was placed on the generator and resonance curves 

 run, observing by small steps wave length, rod meter deflection, and 

 control meter deflection. Then by reducing the rod meter deflection 

 to a standard control meter value satisfactory observations were 

 obtained. As it turned out, the averaged value of all the unsatis- 

 factory observations checked the resonance curve value very closely, 

 but the individual observations scattered all over the rather broad 

 resonance curve top. To avoid the reaction of antennas upon each 

 other they must be separated by at least a wave length, and such a 



Fig. 3 



separation here resulted in too low field strengths unless the antennas 

 were off the ground sufficiently to get an additive combination of the 

 direct and earth reflected radiations. The resonant rod should in 

 any case be well off ground to make certain that its period is not 

 affected by the ground. Check tests showed that at 4 meters distance 

 the ground did not affect the free period markedly. It would be 

 much simpler to observe the resonance curve of a variable length 

 rod, at constant wave length, but it would not be permissible to use 

 a rod of variable diameter and a telescoping arrangement is the only 

 practical method of obtaining coUapsibility. 



The generator used was an UX852 tube connected as in Fig. 2. 

 By means of the variable condenser shown a wave length range of 

 4.24 to 8.44 meters was obtained. This condenser is a cut down 

 "Remmler," the oscillating coil is a three turn center-tapped unit 



