44 



NA TURE 



[July 8, 1922 



effective mass depends on the dimensions of the whole, 

 and its damping on the sound radiated from the mouth. 

 It is then found that the tuning should not be such as 

 to make the representative point occur at the middle 

 of the figure, making both mistunings zero, but that 

 both mistunings should be of the same sign and a 

 certain magnitude, depending on the coefficients of 

 damping of the two degrees of freedom of the coupled 

 system. The mathematical theory is precisely that of 

 a wireless receiver. The ultimate sensitiveness depends 

 on the smallness of the damping of the plate. 



The apparatus as it was built several years ago was 

 mounted upon a heavy bronze stand, covered at the 

 back by a heavy bronze cover to keep out the sound, 

 while the three shafts turning the screws of the inter- 

 ferometer adjustment protruded through sound-tight 

 fittings. Upon the front of the instrument a properly 

 tuned resonator was attached, and at the side was a 



small incandescent lamp with a straight, horizontal 

 filament, an image of which was projected by a lens 

 upon the first mirror of the interferometer. Upon this 

 was focussed a telescope, giving in the reticule an 

 image of the horizontal, straight filament, crossed by 

 the vertical interference fringes seen with white light. 

 In order to get these the plate must be in the proper 

 position within a few hundred thousandths of an inch. 

 The objective of the tuning-fork was carried by a 

 tuning-fork which oscillated vertically, tuned to the 

 pitch of the pure tone to be examined, and this, com- 

 bined with the horizontal motion of the fringes, resulted 

 in a figure of coloured fringes in the form of an ellipse. 

 On slightly mistiming the fork, the ellipse could be 

 made to go through all its phases, and when it was 

 reduced to an inclined straight line its inclination was 

 read off on a tangent scale. The amplitude of the 

 compression of the air in the sound was then directly 

 proportional to the scale-reading. 



While the interferometer is still used for calibration, 

 the movement of the diaphragm is recorded for actual 

 measurements by a thin steel torsion strip carrying a 

 concave mirror. A lamp with a vertical, straight fila- 

 ment is viewed through a telescope into which the small 

 NO. 2749, vol. 1 10] 



mirror focusses the image of the filament on the reticule, 

 and a magnification of from 1200 to 1500 is used, so 

 that the sensitiveness is about the same as with the 

 interferometer. 



At first the only method of tuning was the clumsy one 

 of changing the mass of the diaphragm by adding 

 small pieces of wax. This was not capable of continu- 

 ous variation. Now the diaphragm has been discarded 

 and replaced by a rigid disc supported by three steel 

 wires in tension. The disc is made of mica or aluminium, 

 and is carried by a little steel spider containing three 

 clamps to hold the wire. The tension is regulated by 

 three steel pegs, one of which is controlled by a micro- 

 meter screw. The disc is placed in the circular hole 

 through which the sound enters the resonator. This 

 has the advantage of reducing damping very largely, 

 and thus of increasing the sensitiveness enormously. 

 The instrument now competes with the human ear, 

 and can be tuned over two octaves or more. 



This sensitiveness can be demonstrated by pro- 

 jecting the coloured interference fringes on a screen 

 and singing faintly in a remote part of the room, 

 when the fringes will disappear. Using the telescope 

 end of the apparatus, the instrument will indicate the 

 sound of a tuning-fork when one can scarcely hear it. 

 It is obvious that the disc may be made the diaphragm 

 of a telephone and thus increase its sensitiveness. In 

 fact, Prof. King has used with great success such a 

 telephone to record wireless messages. He has also 

 invented another sort of tunable diaphragm com- 

 posed of a stretched steel membrane with compressed 

 air behind it, which enables it to be tuned continu- 

 ously, but over a smaller range. 



I now come to the source of sound — the phone. 

 This has been reduced to a reversed form of the phono- 

 meter. The disc is driven by an interrupted or 

 alternating current by means of electromagnets, and 

 tuned like the phonometer. Its excursion is measured 

 by a powerful microscope, and the emission of sound is 

 known in absolute measure. It is now driven by a 

 triode valve tube, in the manner suggested by Prof. 

 W. H. Eccles, of Finsbury Technical College, London, 

 for a tuning-fork. This has been worked out for 

 me by Dr. Eckhardt at the Bureau of Standards in 

 Washington. 



The third part of the investigation involves a deter- 

 mination of the coefficient of reflection of the ground. 

 The phone is set at a convenient height, and the 

 phonometer at a convenient distance. Either is then 

 moved along at a constant height and the varying 

 deflections of the phonometer are read while the sound 

 remains the same. Interference sets in between the 

 direct sound and its image reflected in the ground, and 

 the existence of a minimum is obvious to the most 

 naive observer by the ear alone. The reflection of 

 either grass or gravel was found to be about 95 per 

 cent., while, with a most carefully deadened room, 

 the walls of which were covered with thick felt, there 

 was perhaps 20 per cent, reflection. The whole 

 measurement at both ends and the transmission 

 checks up with an accuracy of about 2 per cent. 



With this apparatus all sorts of acoustical experi- 

 ments may be performed. By attaching to the 

 phonometer a long glass- tube or antenna, it has been 

 possible to explore all sorts of places, such as the 



