396 
DR. W. S. TUCKER AND MR. E. T. PARIS ON 
lie used, sucli as measuring the increase of current required to bring the grid back to its 
initial resistance, or determining the alteration in resistance when the current is main¬ 
tained at a constant value. 
It is important when using the microphone in this way that it should not be moved 
during the course of an experiment. This is one of the disadvantages of the method. 
A small alteration in the tilt of the microphone upsets the balance of the Bridge and 
renders the sound-measurements inaccurate. If 6 is the angle between the axis of 
the microphone and a vertical line (so that 0 = 0 when the microphone is held in its 
normal position with neck uppermost), then it is found that as 0 is increased the 
resistance of the grid gradually falls and reaches a minimum when 0 is about 100 degrees. 
The fall in resistance is then about 3 ohms (see fig. 13). 
The resistance of the grid also changes when the microphone is rotated about its 
own axis, except of course when the axis is vertical. Thus, if the microphone is held in 
a horizontal position and in such a way that the glass-enamel support also lies horizontally, 
then a rotation of 90 degrees, bringing the glass-enamel support into a vertical position, 
is accompanied by a fall in the resistance of the grid of about 1 ohm. 
All these effects are due to the influence of the convection currents issuing from the 
heated wire, and it appears that if the convection current from one part of the wire 
impinges on another part of the wire the resistance of the grid as a whole is always 
lowered.* As will be seen from the experiments described later, these convection currents 
play a very important part in the working of the microphone. 
Although the resistance of the grid is changed (current being constant) when the 
plane in which it lies is altered, there is but little change in the sensitivity of the microphone, 
whether it is held horizontally or vertically, provided that its initial resistance is the 
same. 
§ 4. Sharpness of Tuning of the Microphone. 
The natural pitch of a microphone can best be determined by plotting its resonance 
curve. For this purpose the microphone to be tested is set up at a distance of two or 
three feet from a siren (a modified form of Seebeck’s siren was used in the present 
experiments), and the grid is connected into one arm of a Wheatstone’s Bridge as shown 
in fig. 4. The strength of the blast of air in the siren having been adjusted to a suitable 
value, a series of readings are taken of the deflection of the galvanometer and the pitch 
of the siren note. The curve formed by plotting deflection against the interval njp 
(a/2tt being the resonant pitch of the microphone and p /2?r the pitch of the siren note) 
gives what we shall call the “ resonance curve ” of the microphone. 
* The mutual action due to convection of two electrically heated fine platinum wires is described 
by J. S. G. Thomas: “An Electrical Hot-Wire Inclinometer,” ‘ Proc. Phys. Soc. Loud.’, vol. XXXII, 
pp. 291-314 (1920). 
