142 DBS. GUY BARLOW AND H. B. KEENE ON THE ANALYSIS OF SOUND. 
The frequencies and amplitudes of the primaries were as follows :— 
Frequency. 
Amplitude 
(Galvanolneter response). 
A. 
23-9 
90 
B. 
64-6 
42 
C • • • • • • 
69-4 
62 
D ..... 
98-8 
68 
E. 
219-5 
55 
F. 
502 
130 
These values are sensibly the same as were obtained in the separate examination except 
for some small differences in amplitude easily accounted for by unavoidable variations 
in the amplitudes of the forks during the experiment. As the whole range explored 
extended from 3 -2000/sec. and every response found was critically examined, this 
analysis occupied a long time—about four hours. 
In the diagram the response F is shown accompanied by companions F' and F", 
forming a triplet. These companions are due to irregularities in the spacing of the 
segments of the 64 interrupter-disc. They are always given with this disc, but not with 
the others in which, presumably, the spacing is more uniform. 
When two responses occur close together the amplitude of each is reinforced by the 
other. Several such examples occur in the diagram, notably in the case of A and C/3. 
Analysis of Sound in Aim 
(«) Magnetophone Receiver. 
A Graham’s Patent “ Loud-talking Apparatus ” (iron diaphragm 10 cm. diameter, 
0-061 cm. thick) was placed in series with the Broca galvanometer and interrupter. 
This instrument, although not extremely sensitive, proved to be satisfactory for analysing 
sounds. Examples of some of the experiments made are here briefly indicated :—- 
(1) Bowed fork, n = 512, at 8 metres gave a good response at this frequency. 
(2) Voice, singing a note, n — 256, with moderate intensity gave a deflection off 
the scale. 
(3) Organ pipe, blown at a distance of a few feet. Analysis gave the fundamental 
n = 531 and also the third harmonic. The change to the octave on over-blowing 
was readily shown. 
(4) In determining the frequency of a fork it is not necessary that the vibration 
should be sustained. By tapping a 640 fork with a rubber hammer the frequency 
was determined by the interrupter to \ per cent. 
