174 
PHYSICS: A. G. WEBSTER 
a 'phonometer;' third, in order to check the theory of the two instruments a 
series of experiments on the propagation of sound from one to the other, free 
from the effect of disturbing bodies. Such experiments must be made either 
in a room so padded as to absorb all sound, which is at present impossible, 
or else out of doors at a distance from all objects except the fiat surface of the 
earth, of which the coefficient of reflection is measured. This has been the 
method here pursued. While many observers have attacked one or both of 
the instrumental problems, it is not known that any has previously attacked 
all three parts above mentioned. 
1. Phonometer. — In order to attain the required sensitiveness for the small 
amount of energy concerned, it is necessary to use the principle of resonance, 
and a coupled system of two degrees of freedom is used, the first consisting 
of a tunable air resonator, the second, of a tuned diaphragm which is to make 
the motion visible. Inasmuch as the sensitiveness will ultimately depend 
on the damping intrinsically residing in the diaphragm, it is desirable to reduce 
this as much as possible. For many years the best thing found was a glass 
diaphragm, tuned by adding weights. Finally it was found that mica dia- 
phragms could be had having a smaller damping than glass. Finally, how- 
ever, after an attempt to measure the loudness of some of the fog-signals on 
the Maine coast, all of which have different pitches, it was found necessary 
to have a diaphragm susceptible of gradual tuning, which could not be obtained 
by variation of mass. Accordingly the instrument was entirely redesigned, 
and the tuning was done by varying the potential, rather than the kinetic 
energy of the diaphragm, by stiffening it with a string tuned by tension. And 
now the chief improvement was made of abolishing the diaphragm, and since 
the resonator has a hole for the sound to enter, it was determined to replace 
the diaphragm by a rigid piston placed centrally in the hole, leaving an annu- 
lar aperture for the sound to enter, and at the same time freeing us from the 
necessity of using the calculated equivalent area of the diaphragm. These 
improvements are embodied in the instrument shown in Figs. 1 and 2. 
The phonometer is shown in elevation in Fig. 1, where 1 is the cylindrical 
resonator, sliding in the cylindrical piece screwed to the main casting A, rest- 
ing on two feet and the cylinder 8. The end of the resonator is closed by a 
glass window 15. For years all measurements were made by the Michelson 
interferometer viewed stroboscopically, as it was desired to have the dia- 
phragm absolutely free to move. For purposes of portability and conven- 
ience this has now been given up, and replaced by observing in a telescope 
the displacement of a mirror, as in Max Wien's phonometer. The inter- 
ferometer, 14, (not shown) is attached temporarily at the front, or receiving 
end, to check the indications of the mirror. In the micrometer eyepiece, 2, 
adjustable with five degrees of freedom, is observed the image of the filament 
of the small lamp 4, fed through a key 5 by three dry cells, 10, placed in the 
cylindrical holder upon which the tube 7 is carried by a bayonet joint. The 
