416 
DR. W. S. TUCKER AND MR, E. T. PARIS ON 
experiments A and B respectively. The curve A' (fig. 12) is calculated from d 
146 -3 
It will be seen that the observed curve A lies along A' when r is greater than about 
28 feet, that is, when the sound has fallen below a certain amplitude. When r is less 
than 28 feet, the amount of deflection is less than it would have been had the simple 
inverse first-power rule continued to hold. This is indeed what might be expected 
from a consideration of the action of the microphone grid, as pointed out at the end of 
§ 6, but the observed falling off in deflection is too great to be accounted for in this 
way alone. 
In experiment B (curve B, fig. 12), the agreement between the observations and 
the relation d oc - is better. 
r 
The broken curve B' is calculated from d = 
185-4 
Although the deflection corresponding to a given value of r is greater in the B curve 
than the A curve, it must not be inferred that the sound was louder in the former 
case. The deflection obtained depends not only on the loudness of the sound but also 
on the sensitivity to which the amplifier is adjusted and which can be varied over a very 
wide range. As a matter of fact the sound was louder in the A than in the B 
experiment. It may also be noted that in both cases the sound was louder than in 
the Wheatstone’s Bridge experiment. 
To sum it up, it may be said that the available evidence points to the fact that, when 
the microphone is employed with an amplifier and vibration galvanometer and only very 
faint sounds are observed, the deflection shown by the galvanometer is approximately 
proportional to the amplitude of the sound. 
§ 8. The Effect of Tilting the Microphone. 
It has already been mentioned that, when taking sound measurements with a Hot- 
Wire Microphone, it is necessary to keep the axis of the microphone inclined at some 
fixed angle to the horizontal. In fig. 13 a curve has been given showing the relation 
between the resistance of the grid and the angle of inclination of the axis to the vertical, 
the electric current carried by the grid being maintained at a constant value throughout 
the measurements. It is obvious from this curve that the tilt of the axis must not be 
altered while measurements are being taken by the Wheatstone’s Bridge method. 
The two most noticeable features about the curve are : 
9 
(i) That the resistance of the grid is less when the microphone is held upside down 
(0 = t r) than when it is in its normal upright position (6 = 0), although in both 
cases the plane in which the grid lies is approximately horizontal; and 
(ii) That the resistance is least—that is, the rate of cooling is greatest—when 6 is 
somewhat greater than T. 
These two experimental results are curious and difficult to explain. They are both, 
