254 
source, which can accelerate production and improve 
the product. This is, and always has been, the 
policy of the Medical Research Council. But such 
a policy has only been made possible, in the existing 
state of the law, by the Toronto application for patent 
and the Council’s consent to accept assignment. 
(4) The most surprising, and the most serious, 
misunderstanding of the position is revealed by Sir 
William Bayliss’s statement, that “‘ the necessity 
for any laboratory being unable to do this,” 7.e. 
to carry out research on insulin—‘ except by arrange- 
ment with the patentees does not seem desirable.”’ 
I agree that it is not desirable; but the alleged in- 
ability is quite imaginary. No such permission is 
required, but many have sought and have been 
given help, and many more have received help 
without seeking. Sir William Bayliss will be familiar 
with the custom current among scientific workers, of 
maintaining a certain reticence with regard to work 
which is unfinished or results which are still in doubt, 
and even of asking others to keep clear of a certain 
problem for a time, to avoid duplication. I can 
state with confidence that, even in that restricted 
and legitimate sense, there has been no attempt on 
the part of the Council, or of those working for them, 
to keep any kind of secrecy or monopoly in this 
field, so far as pure research is concerned, un- 
complicated by questions of personal gain. 
In my own department, the whole of our know- 
ledge of this subject has been put freely at the 
disposal of other pure research workers—not only 
what has come to us in connexion with the patent, 
but what has resulted from our own investigations. 
Sir William Bayliss acquits the Council of “ any 
desire whatever to obstruct such research.” I think 
he might safely allow himself to go a little further, 
and recognise that their policy in this matter, and 
its interpretation by those working for them, has 
resulted in a quite unusual freedom of assistance to 
research, with both information and material. 
H. H. DALE. 
The National Institute for Medical Research, 
Hampstead, February 12. 

Multiple Resonance. 
Str RIcHARD PaGEt’s skilfully demonstrated 
lecture at University College on October 18, 1922, the 
substance of which is given in NATuRE of January 6, is 
not less interesting as giving a very simple account 
of the nature and formation of speech sounds, than 
as showing how far-reaching and diverse are the 
applications of “‘ multiple resonance ’’ in acoustics. 
Boys (Nature, vol. 42, p. 604, 1890) made use of 
a special kind of double resonator in constructing a 
very sensitive form of ‘‘ Rayleigh disc,”’ and Rayleigh 
extended his theoretical consideration of double 
resonators given in his “‘ Theory of Sound,” vol. ii. 
p. 190, to show that ‘‘ the condensation in the second 
resonator may be made to exceed to any extent that 
in the first, by making the second resonator small 
enough ”’ (Rayleigh, Phil. Mag. xxxvi. 231-4, 1918). 
Following some preliminary experiments by Prof. 
Callendar and Major Tucker in 1918, Capt. E. T. 
Paris employed the double resonator for the purpose 
of increasing the sensitivity of the hot-wire micro- 
phone originally devised by Major Tucker and also 
of extending the range (in pitch) of maximum 
sensitivity. 
For a single resonator the response curve is a sharp 
peak, but with a proper design of double resonator 
the two peaks (characteristic of such resonators in 
on ‘ 
NATURE 



[FeBRUARY 24, 1923 
practically one broad flat-topped peak. In the 
accompanying diagram (Fig. 1) the dotted line gives 
a typical response curve for a single resonator, and 
the full line is typical of the curve obtained when a 
resonator of suitable proportions is added, to form a 
double resonator. 
Paris has shown how the form of the resonance curve 
can be varied with tuning of the component resonators, 
from equal response to each of the two natural 
tones, to the case of much stronger response to one 
than the other. The double resonator of Boys 
appears to have been of the latter kind. ‘‘ A doubly- 
resonated microphone of the type described may be 
u077027/0q 

Piteh 
Fic. 1. 
more than a hundred times as sensitive (to one of 
its resonant tones) as a suitably tuned singly resonated © 
microphone with an aperture similar to that of the 
inner resonator’? (E, T. Paris, Proc. Roy. Soc. A, 
vol. IoI, 1922). 
The present writer, independently, developed a 
form of multiple resonator to provide a recording 
system for the Gramophone Co., Ltd., Hayes, 
Middlesex, having a uniform response over a very 
wide range—namely four octaves. For so wide a 
range (129 to 2069 vibs./sec.) the source of sound was 
a series of stopped diapason, wooden organ pipes, 
giving fairly pure tones of uniform intensity as 
judged by the trained ear. The amplitude of vibration 
of the diaphragm—or of the point of the wax-cutting 
needle—was observed by means of exceedingly small 
mirror attachments, the best of which was similar 
to D. C. Miller’s ‘‘ phonodeik ’’ (Miller, “ Science of 
Musical Sounds,’’ Macmillan). It was possible so 
to adjust the ten to twenty components of the 
multiple resonator that the response curve (plotted 
amplitudes against pitch) would conform very closely 
to a curve such as that determined by the theoretical 
constant intensity relation, a? n*=constant, without 
a serious loss of sensitivity. 
A resonator of simple form with yielding walls is 
a double resonator. Rayleigh, on the suggestion of 
Clerk Maxwell, considered the case of a sphere with 
non-rigid walls (Phil. Trans., 1871, p. 87) and showed 
that it has two natural periods, being a system of — 
two degrees of freedom, like an ordinary double- 
resonator. Similarly, a rigid resonator of conical 
form with a diaphragm mounted at the narrow end 
is a double resonator, and so we get an explanation 
of the fact that the octave interval between the first 
and second partials for the conical resonator alone, 
is increased by as much as a tone if the natural’ 
frequency of an attached diaphragm falls between the 
frequencies of the partials. 
Multiple resonance will perhaps account for the 
remarkable evenness of response and the character- 
istic quality of tone given by the resonating body 
of good violins. Helmholtz (“ Sensations of Tone,” 
1885 edition, p. 87, Ellis translation) found two tones 
of greatest resonance on a violin by Bausch, one 
between 264 and 280, and the other between 440 
general) may coalesce, so to speak, into what is | and 466 vibs./sec., when he tested it by placing the 
NO. 2782, VOL. 111] 

