September 15, 192 1] 



NATURE 



83 



Mr. Preston Edwards makes the hypothesis that a 

 weak larynx tone arouses a complex of . resonance 

 tones in the vocal cavities, and that these are then 

 heard with a strong lower tone corresponding to the 

 larynx tone. We can sing all the vowels with a single 

 strong larynx tone ; that all the varied resonance 

 tones should coincide in producing this single strong 

 larynx tone is quite beyond imagination. It is 

 scarcely necessary to follow this thought out ; the fact 

 remains that the larynx tone is by far the strongest 

 tone in the vowel, both physically and for the ear. 



A comparison of vowels with tones from reed in- 

 struments is inadmissible. The resonating cavities 

 in the latter have hard walls and can force their 

 periods on the reeds, whereas the vocal cavities have 

 soft walls and cannot do so. 



Helmholtz's theor\- of the vowels rests on three 

 assumptions : — (i) The vibrations from the lar\nx are 

 of the form of a simple sinusoid or of the sum of a har- 

 monic series of simple sinusoids. (2) The vocal cavities 

 act like resonators with hard walls. (3) The lan,nx tone 

 is maintained at constant pitch. This is the condi- 

 tion exemplified in Helmholtz's vowel apparatus, in 

 which tuning-forks (representing the larynx tone) are 

 maintained electrically in vibration before brass 

 resonators (representing the vocal cavities). Such 

 resonators can respond only to harmonics of the fork 

 tone. If the three suppositions are true, the harmonic 

 theory necessarily follows. 



In an interesting apparatus devised by Dr. H. 

 Hartridge a series of weights is hung by threads 

 of different lengths from a bar that can be set in 

 sinusoid vibration. When the bar is started all the 

 pendulums begin to swing, but after a few vibrations 

 all become still except the one that corresponds to a 

 harmonic of the period of the bar. It takes, there- 

 fore, several vibrations before the period of the bar 

 is forced on the appropriate pendulum, and the others 

 are forced to become still. Remembering the first 

 fact — that a single movement of the bar starts all 

 the pendulums — let us now drop the third supposition 

 above and conform to the truth that the voice tone 

 is never of a constant pitch for a single instant. In 

 speech it changes so rapidly up and down that two 

 successive vibrations are rarely of the same pitch. 

 Each single vibration must start all the pendulums — 

 that is, all the tones of the resonators — and there is 

 no possibility that the tone of the voice can force 

 itself on the resonators. Even retaining the first two 

 suppositions, we are forced to conclude that any har- 

 monic adjustment of the vocal resonators to the voice 

 tone is impossible, because it takes at least two like 

 periods of the voice tone to force its frequency on a 

 resonator, whereas in actual speech the voice tone 

 does not have two like periods in succession. The 

 Helmholtz theory is thus impossible, because the third 

 supposition is contrary to fact. 



Let us now examine the effect of retaining the first 

 and third suppositions while modifying the second 

 to fit the facts. The resonators have soft walls, 

 and any vibration aroused in them will die away 

 rapidly. Even if the lar\ngeal vibration were of the 

 sinusoid form, the resonance vibration aroused by one 

 positive (or negative) phase would die awav before 

 the corresponding phase occurs again. To a simple 

 sinusoid or a tuning-fork tone there is at the best 

 only a weak resjx>nse. With more complicated sinu- 

 soids or sharper tones there may be rapidly repeated 

 overtones that evoke some response. The result is 

 utterly different from the loud responses evoked bv 

 forks held in front of resonators with hard walls. 



The moment we give up the first assumption and 

 accept the fact that the tone from the larynx consists 

 of a series of sharp blows, and not of continuous 

 NO. 2707, VOL. 108] 



vibrations, all possibility of the harmonic hypothesis 

 vanishes. A sharp blow can arouse only the free 

 vibration of the cavity, whether hard or soft. The 

 tones aroused are those of the cavities themselves. 

 Their periods have no relation to the period of the 

 blow, simply because the blow itself has no period. 

 Even in a cavity with hard walls the aroused vibra- 

 tion dies away so quickly that the blow would have 

 to be repeated rapidly in order to catch any of the 

 dying motion. 



.As explained in my account in Nature (January 13 

 and 20, pp. 632 and 664), every one of the three 

 assumptions is contrary- to fact. If Helmholtz had 

 had the data we possess to-day the Helmholtz theory 

 of the vowels would assert that the cavity tones may 

 bear any relation to the lar\nx tone. The whole 

 thing follows so easily and naturally from the facts 

 that he would have drawn the unavoidable conclusion, 

 and not have become involved in the difficulties and 

 impossibilities caused by asserting that this relation 

 must be harmonic. 



Of course, this has nothing to do with the Fourier 

 analysis. This method gives us a plot of what har- 

 monics might be used to reconstruct a curve. The 

 profile of a face can be resolved into a series of har- 

 monics. The blunder occurs when someone supposes 

 that this mathematical analysis proves anything 

 physical. The analysis of the profile does not prove 

 that the face was constituted of harmonics. As ex- 

 plained in my first article, the harmonic plot of a vowel 

 wave represents a mathematical formula. As soon 

 as we attempt to give a physical interjM-etaticwi to this 

 formula we are forced to reject the harmonic thec«y 

 and to assert that the vowel tones may bear anv 

 relation whatever to the fundamental. The Fourier 

 analysis still remains the only way of analysing vowel 

 curA-es, and its results prove the independent theory of 

 the vowel tones. E. W. Scripture. 



The Generation of Heath-fires. 



Is Mr. Martin (Nature, August 25, p. 811) certain 

 that the fire seen by him had not spread through the 

 peat from a neighbouring area recently ablaze? That 

 such subterranean combustion may persist for manv 

 days and spread a long distance is common know- 

 ledge. 



Experience shows that the surest way of prevent- 

 ing fires is to cut away the heather; whereas, on Mr. 

 Martin's hypothesis, by exposing the loose soil to 

 the direct rays of the sun this procedure should add 

 a new risk. Henry Bury. 



Mayfield House, Farnham, Surrey, August 26. 



Life and Mind. 



During the past six months Nature has published 

 many articles and reviews about life and mind, but 

 there has been no clear and precise definition of these 

 two terms ; so permit me to submit for the serious 

 consideration of your scientific readers a brief state- 

 ment of what appears to be the true position of 

 aflairs : — 



(i) W^e depend only upon evidence of what is called 

 "matter " and its various states. 



(2) Life is a state of certain kinds of matter when 

 physical conditions are favourable. 



(3) Mind is the state of some portion of a living 

 organism — in man the cerebral cortex, where 

 thousands of nerve-endings are concentrated. 



This is no attempt to explain the how and why ; 

 but can the facts, if facts they be, be expressed more 

 simply or concisely and with greater precision ? The 

 question may appear trivial, but it has an important 

 bearing on other subjects which are alluded to from 

 time to time in Nature. Sesamy. 



