at the Ends of a Circular Resonator. 653 
thin brass wire 0°04 em. thick, the width of the rectangles 
being 31 cms. The experiments were made by the first 
method, described on p. 647, in which the length of the 
resonator is kept constant, the length of the parallel wires 
being altered by winding the wire into the little boxes. In 
an experiment the length of the parallel wires was altered by 
steps of 5 or 10 cms., and readings taken of the disturbance 
in the resonator, with a Rutherford solenoidal detector, on 
discharging the condenser. The observations with the de- 
tector were plotted against lengths of rectangular circuit and 
a curve drawn freely through the points obtained. The 
crest of the curve has been taken as marking the length of 
the rectangular circuit when in tune with the open circle. 
The sparks discharging the condenser passed between 
aluminium spheres immersed in paraffin-oil. In spite of 
Fig. 6. 
20 
‘9 
8 
(A CRIS 
I7 
as 
15 
Detlechioars of maznetometer 

3 a5 4 45 
Resonator from boxes «metres 
the fact that the discharge-surfaces were kept well polished, 
the sparks were not uniform in character; this accounts 
for the variations in the readings. At least two sets of 
observations were taken in determining the position of each 
crest. A specimen curve is given in fig. 6. 
