On the Vibrations and Tones produced by the Contact of Bodies having different Temperatures .
By JOHN TYNDALL , Ph. JD .
, F.R.S. , Member of the Royal Society of Haarlem , and Professor of Natural Philosophy in the Royal Institution .
Received January 15 , -Read January 26 , 1854 .
IN the year 1805 , M. SCHWARTZ , inspector of one of the smelting-works of Saxony , having a quantity of silver in a ladle which had just solidified after melting , and wishing to hasten its cooling placed it upon a cold anvil , when to his astonishment sounds , which he compared to those of an organ , proceeded from the mass .
The rumour of this discovery excited the curiosity of Professor GILBERT , the editor of GILBERT'S Annalen , and in the autumn of the same year he paid a visit to the smelting-works in question .
He there learned that the piece of silver from which the sounds proceeded was cup-shaped , had a diameter of 3 or 4 inches and a depth of half an inch .
GILBERT himself , under the direction of M. SCHWARTZ , repeated the experiment .
He heard a distinct tone , although nothing that he could coinpare to the tone of an organ .
He also found that the sound was accompanied by the quivering of the mass of metal , and that when the vibrations of the mass ceased , the sound ceased likewise .
The Professor limited himself to the description of the phenomenon and made no attempt to explain it .
In the year 1829 Mr. ARTHUR TREVELYAN was engaged in spreading pitch with a hot plastering iron , and observing in one instance that the iron was too hot , he laid it slantingly against a block of lead which happened to be at hand .
Shortly afterwards he heard a shrill note , resembling that produced on the chanter of the smaller Northumberland pipes , an instrument played by his father 's gamekeeper .
Not knowing the cause of the sound he thought that this person might be practising out of doors , but on going out the tone ceased to be heard , while on his return he heard it as shrill as before .
His attention was at length attracted to the hot iron , which he found to be in a state of vibration , and thus discovered the origin of this strange music .
In 1830 he came to Edinburgh and mentioned the fact to Dr. REID ; the latter , not knowing what SCHWARTZ and GILBERT had observed previously , regarded the phenomenon as new and recommended Mr. TREVELYAN to investigate it more fully .
Mr. TREVELYAN did so ; among other things he discovered the form to be given to the vibrating mass ( the rocker ) in order to obtain the effect with ease and certainty .
The results of his numerous and well-contrived experiments were cormmuinicated to the Royal Society of Edinburgh , and were subsequently printed in the Society 's Transactions .
On the Ist of April 1831 these vibrations and tones constituted the subject of a Friday evening 's lecture by Professor FARADAY at the Royal Institution .
The following extract from the Journal of the Institution , vol. ii .
p. 120 , informs us of the views of the philosopher last mentioned with respect to the cause of the tones .
" As the sounds were evidently due to the rapid blows of the rocker , the only difficulty was to discover the true cause of the sustaining power by which the rocker was kept in motion , whilst any considerable difference of temperature existed between it and the block of lead underneath .
This power Professor FARADAY referred to expansion and contraction , as Professor LESLIE and Mr. TREVELYAN had done generally .
But he gave a minute account of the manner in which , according to his views , such expansions and contractions could produce the effect ... . .
The superiority of lead , as a cold metal , he referred to its great expansibility by heat , combined with its deficient conducting power , which is not a fifth of that of copper , silver , or gold ; so that the heat accumulates much more at the point of contact in it than it could do in the latter metals , and produces an expansion proportionably greater .
" Professor J. D. FORBES was present at this lecture , and by it , apparently , he was induced to undertake the further examination of the subject .
On the 18th of March and on the 1st of April , 1833 , the results of his inquiries were communicated to the Royal Society of Edinburgh .
He dissents from the explanation supported by Professor FARADAY .
The vibrations , he urges , are dependent for their existence on the difference of temperature of the two surfaces in contact ; if then the heat accumulate at the surface of the cold metal , its effect will be to bring both surfaces to a common temperature and thus to stop the vibrations , instead of exalting them , as supposed by Professor FARADAY .
Again , if the phenomenon be due to expansion , the greater the expansion the greater ought to be the effect ; but the expansion depends upon the quantity of heat transmitted from the hot rocker to the cold block during their contact , and this again upon the conductivity of the block ; so that instead of being a bad conductor , the block , to produce the greatest effect , ought to be the best conductor possible .
The idea of an accumulation of heat at the surface being more favourable to the action than a rapid communication with the interior , Professor FORBES regards as an " obvious oversight* .
" Having thus , to all appearance , overturned the views previously entertained , Professor FORBES proceeds to found a theory of his own .
His experiments have led him to the enunciation of certain '"general laws , " and these converge upon the still more general conclusion , -- " that there is a repulsive action exercised in the transmission of heat from one body into another which has a less power of conducting it .
" This repulsion Professor FORBES considers to be " a new species of mechanical agency in heat , " and he cites the remarkable experiments of FRESNEL , on the mutual repulsion of heated bodies in vacuo , as bearing directly upon the subject .
Such , apparently , was the unsettled state of the question when my attention was drawn towards it last summer .
The possibility of the explanation offered by Professor FORBES , affording , as it seemed to do , a chance of becoming more nearly acquainted with the intimate nature of heat itself , was a strong stimulus to inquiry .
I was not aware , until informed of it by my friend Professor MAGNUS , that SEEBECK had further examined the question , and substantiated the conclusions arrived at by FARADAY .
On reading SEEBECK'S interesting paper I found that he had already obtained many of the results which it was my intention to seek ; nevertheless the portion of the subject which still remained untouched presented sufficient interest to induce me to prosecute my original idea .
I purpose in the present memoir to examine the experimental basis of those laws which Professor FORBES regards as establishing the existence of 'a new mechanical agency in heat ' ; and as I am anxious to place it within the power of every experimenter to test the results to be communicated , I shall connect with each series of experiments a sufficiently exact description of the instruments made use of .
The first general law enunciated by Professor FORBES is as follows : " The vibrations never take place between substances of the same nature .
" Let us see whether this law will bear the test of experiment .
I. Iron Rocker .
Fig. I represents a sketch of the rocker ; the length AC is 5'1 inches ; the width AB 1'85 , and the length of the stem OF is 12 inches .
Fig. 2 is a transverse section Fig. 1 .
Fig. 2 , Fig. 3 .
Fig. 4 .
IIv ' !
I 1r 1111 !
- !
^.v of the rocker , showing the groove underneath ; the deptli SN is 0-5 'of an inch .
B2 iiiiiiliii !
iiiilliiiiiiiiiiiiliiiiiiji ; illiiiiiliiiiiiiiiiii/ ; !
iiii/ iiilii ` irrrtii rtrtrtitif irrirf tntrrlrftiriririiritl rtirrtirtrtf to rif Irtrrtrtrrrrr 3 The distance rt is 035 of an inch , and it is divided into three equal spaces , the central one of which is occupied by the groove , and the other two are the surfaces which rest upon the bearer .
Fig. 3 is a plan of the rocker turned upside down .
The spaces k , fig. 1 , and mnop , fig. 3 , will , for the present , be left out of consideration .
1 .
The blade of a dinner knife was fixed in a vice so that the edge was horizontal .
The rocker was laid upon the edge , and the stem suitably suppotrted .
On stirring the rocker a loud and musical sound commenced , and continued for a considerable time .
2 .
The knife was removed and a plate of sheet iron 4-1th of an inch in thickness was fixed in the vice ; the hot rocker was caused to rest upon the edge of the plate .
On stirring the rocker , vibrations , accompanied by a musical tone , were set up as before .
3 .
The experiment was repeated with a second plate of iron -oIth of an inch in thickness , and a still better tone was obtained:-when the rocker rested on a block of iron the vibrations were not permanent .
II .
Copper Rocker .
The pieces k and mnop , figs. 1 and 3 , are plates of copper , screwed tightly on to the surface of the iron .
In this way a single rocker is made to do the duty of two .
1 .
A plate of copper " th of an inch in thickness was fixed in the vice , as in the former cases ; and the copper portion of the rocker was caused to rest upon it .
A slight shock , imparted to the rocker , immediately excited a strong and durable tone .
2 .
A bit of copper foil was fixed in the vice ; it was almost as flexible as stout foolscap paper , but to give it rigidity the height of it which projected above the vice was very minute .
With a little care I obtained tones stronger and more musical than in the foregoing instance .
3 .
When the rocker was laid upon a block of copper no tone was obtained , and it was found that the difficulty of obtaining a tone increased as the plate made use of became thicker .
4 .
Instead of the plate , two wires of copper -1-th of an inch thick , and pointed with a file , were fixed in the vice at about one-eighth of an inch apart .
The rocker was turned upside down , so that the flat surface of the copper k rested on the wires as in fig. 4:-forcible vibrations were obtained in this way .
III .
Brass Rocker .
1 .
A piece of brass tube was fixed in the vice and its cylindrical surface rendered clean by a fine file .
A brass rocker of the same dimensions as that represented in fig. 1 , was caused to rest upon the tube ; on stirring the rocker continuous vibrations succeeded .
2 .
A plate of brass 1~oth of an inch in thickness was fixed in the vice .
The rocker being laid upon the edge of the plate and stirred , stronger and more durable vibrations were obtained than in the case of the tube .
4 The experiment was repeated with a plate twice the thickness of the former ; distinct vibrations were obtained .
It was found in this case also that the thinner the plate , within its limits of rigidity , the more decided were the effects:-when the rocker was laid upon a block of brass there was no permanent vibration .
4 .
The rocker was turned upside down , its flat surface resting upon the points of two common brass pins ; a constant rocking was the consequence .
IV .
Silver Rocker .
This instrument was formed by attaching a piece of silver to the brass rocker used in the last experiments , exactly as the piece of copper , mnop , was attached to the iron rocker , fig. 1 .
The silver partook of the general shape of the under surface of the rocker , being bevelled off on both sides of the groove passing through its centre .
1 .
A strip of silver about -rToth of an inch in thickness was fixed in the vice , and the silver portion of the rocker was caused to rest upon the edge of the strip .
On shaking the rocker a fine mellow musical tone was obtained .
2 .
A new half-crown was fixed in the vice and the rocker caused to rest upon the milled edge ; no permanent vibrations were obtained .
A similar difficulty was encountered with the edge of a shilling .
On the edge of a sixpence , a feeble , though distinct vibration was obtained .
3 .
When the edges of the coins were beaten out with a hammer , and thus rendered thin , distinct vibrations were obtained with all of them .
I do not assert the impossibility of obtaining vibrations on the edge of a half-crown , but merely state that with the same rocker vibrations were obtained upon a thin edge of silver , and not upon a thick one .
4 .
The rocker was placed against a block of silver weighing about ten ounces ; no permanent vibrations were obtained .
V. Zinc Rocker .
The instrument is of the same size and shape as the iron rocker , fig. 1 , except that the depth , SN fig. 2 , is less and the mass therefore lighter* .
1 .
Placed upon the edge of the thinnest sheet zinc , the edge having been sharpened by a file , distinct musical tones were obtained:-on a block of zinc the rocker refused to vibrate per'manently .
Tin Rocker .
1 .
A cake of tin , formed by pouring the molten metal upon a smooth flat surface , was heated and balanced upon two small protuberances of a second piece of the same metal .
Continuous rockings were immediately set up .
I met the mass by accident in the laboratory , and having obtained the vibrations without changing its shape , I did not think it necessary to strengthen the action by forming it into a regular rocker .
The number of metals capable of this action might , I doubt not , be greatly extended .
Thus far we have obtained vibrations with Iron upon iron , Copper on copper , Brass on brass , Silver on silver , Zinc on zinc , Tin on tin , and these , I think , are sufficient to show that thefirst general law of Professor FORBES does not stand the test of experiment .
SEEBECK indeed had already proved the untenableness of this law .
His method of experimenting has been followed in one or two of the cases above described .
The placing of the heated rocker upon pointed wires is his idea .
Rockings are very readily obtained in this way ; but when tones are r equired , the sharp edge will , I think , in general be found preferable .
The second general law of action stated by Professor FORBES is as follows : " Both substances must be metallic .
" This is the case which first excited my attention ; for even granting the final explanation given by Professor FORBES to be the true one , the necessity of the law before us does not at all follow .
Previous to entering upon the present subject I had found that rock-crystal and rock-salt possessed conducting powers not much , if at all , inferior to some of the metals ; and this led me to suppose that either , or both of these substances , might possibly be made to exhibit the action which the above law restricts to metals .
My first attempts failed through want of delicacy , as first attempts generally do .
But a little practice suggested the means of imparting to the rocker the requisite degree of mobility .
Crystals of quartz were cut in such a manner that when the rocker was laid upon them a very slight force was sufficient to cause it to oscillate .
By this means I had the satisfaction of obtaining distinct vibrations from a brass rocker placed upon rock-crystal .
I refrain from entering into a more exact statement of the manner in which the crystals were cut ; for subsequent experience proved that there is no difficulty in obtaining the effect , without any artificial preparation whatever .
I shall now proceed to describe the results obtained with non-metallic bodies .
1 .
Rock-crystal.-The brass rocker already described was heated and placed upon the natural edge of the prism ; the stem was supported by a knife edge , so that the rocker lay nearly horizontal : a strong tone was thus obtained .
Vibrations also followed when the rocker was laid upon the edge of the pyramid which caps the hexagonal prism .
6 The experiments were repeated with fumy quartz , and the same result was obtained .
2 .
Fluor-spar.-A smaller brass rocker than that , last used was found to answer best with this crystal .
The dimensions corresponding to AC , AB , fig. 1 , were 3*8 inches and 1*25 inch respectively ; while the depth was the same , or nearly so .
This rocker having been placed upon the natural edge of the crystalline cube , a clear and melodious note was instantly produced .
Forcible vibrations were also obtained with the larger rocker , but not so clear a tone .
The angle of the cube was cloven off so as to expose the edge of the octahedron on this edge also vibrations were obtained .
Fortjfication Agate.-Distinct vibrations and tones were obtained with the large brass rocker .
I found it sometimes convenient to increase the time of oscillation by laying a thin brass bar with smtall knobs at the ends , across the rocker-a mode of experiment due , I believe , to Mr. TREVELYAN .
With this precaution , in the case before us , the rocker continued swinging for nearly half an hour , and when it ceased it was under the tempelature of boiling water .
Rock-salt.-The rocker used in the foregoing experiment was laid aside and a piece of rock-salt was prepared for trial .
The mass was cloven so as to exhibit the surfaces of the primitive cube ; and was so placed that the straight line formed by the intersection of two of the sutfaces of the cube was horizontal .
Previous to heating the rocker , I laid it , according to practice , upon the mass , merely to ascertain whether the arrangement was likely to answer .
To my astonishment a deep musical sound commenced immediately .
The temperature of the rocker was at this time far below that of boiling water , and when it had ended its song it was scarcely above a blood heat .
The heated rocker was laid upon a large boulder-shaped mass of the salt ; it commenced to sing immediately .
I scarcely know a substance , metallic or non-metallic , with which vibrations can be obtained with greater ease and certainty than with this mineral .
To the remarkable properties which the researches of MELLONI have shown to belong to rock-salt a new one may now be added .
Avanturine.-I was tempted to try this mineral from having met a piece of it possessing a clean sharp edge .
The large brass rocker placed hot upon this edge gave a decided tone .
Sulphate of Potash.-Care is required with this artificial crystal , as it readily flies to pieces on the sudden communication of heat .
With proper precautions , feeble , but well established vibrations , were obtained .
Onyx.-A distinct tone .
Tourmaline.-After many trials I obtained a continuous vibration and low tone .
Fossil Wood.-Two different specimens were exarnined and distinct tones obtained with both of them .
Banded Agate.-Strong and continuous vibrations , when the rocker was very hot .
Chalcedony.-Loud and long-continued knockings on a knob of this mineral .
7 Glass.-Decided vibrations on the smooth rounded edge of the foot of a drinkingglass .
Mr. TREVELYAN believed that he once obtained vibrations upon glass , but the fact is doubted by Professor FORBES .
This is the only experiment on non-metallic bodies , as far as I arn aware of , hitherto on record .
Earthenware.-A feeble tone , which soon ceased , was obtained on the edge of a dinner plate .
Flint.-A decided tone , though not so strong as that obtained from rock-crystal .
Lydian Stone.-Permanent vibrations .
Heliotrope.-A durable tone .
Iceland-spar.-A lighter rocker than any of those hitherto described was found necessary in experimenting with this crystal .
The mass is soft , and is readily bruised by the rocker , when the latter is heavy .
With a suitable instrument a continuous feeble tone was obtained .
Red Hematite.-Distinct tones were produced by several specimens of this mineral .
Arseniacal Cobalt.-A strong tone .
Meteoric Ironfrom Mexico.-A low musical tone .
This list might be readily extended .
The substances mentioned in it were chosen on account of their accidentally presenting the conditions favourable to experiment .
The principal condition is a clean even edge .
Several of the minerals possessed such edges cut artificially ; others possessed them naturally .
In the case of chalcedony , the rocker was placed upon a rounded knob ; in the case of tourmaline , one of the ridges , which usually run along the surface of the prism , served as a support ; with glass and earthenware the surfaces were smooth and rounded .
As a general rule however I have found an even edge best .
With such an edge , and rockers similar to those described , no difficulty will be experienced in repeating and extending these experiments .
It is usual to permit the knob at the end of the handle of the rocker to rest upon a flat surface , while the instrument itself leans slantingly against the bearer .
In delicate experiments I think a knife-edge is a better support for the handle , the rocker being placed horizontal , or nearly so .
Omitting the last three substances , which might , perhaps , with some justice be regarded as metallic , we find a number of exceptions to the law under consideration which far exceeds the number of bodies mentioned in the paper of Professor FORBES .
These exceptions demonstrate that the second law also is untenable .
The third general law runs as follows : " The vibrations take place with an intensity proportional ( within certain limits ) to the difference of the conducting powers of the metals for heat , the metal having least conducting power being necessarily the coldest .
" The evidence adduced against the validity of the first law appears to destroy this one also ; for if the vibrations are to be ascribed to a difference in the conducting powers of the rocker and bearer , then when there is no such difference there ought to be no vibrations .
But we have shown , in half-a-dozen cases , that vibrations 8 'FTllH CONTACT OF BODIES HAVING DIFFERENT TEMPERATURES .
9 occur when rocker and bearer are of the same metal .
The same facts deprive the latter part of the third law of its significance .
I will however cite one or two experiments , in which the conditions regarded necessary by Professor FORBES were reversed , and the effect was produced notwithstanding .
1 .
Silver stands at the head of the conductors of heat .
A copper rocker was laid upon the edge of a thin plate of this metal ; strong musical notes were obtained from the arrangement .
2 .
Forcible vibrations were produced by placing a blass rocker upon the same silver plate .
3 .
A feeble , but distinct tone , was produced by the iron rocker .
4 .
Gold is a better conductor than brlass ; nevertheless starong vibrations were obtained by placing a hot brass rocker upon the edge of a half-sovereign .
These experiments are , I think , sufficient to prove the non-existence of the third law .
In the prosecution of his inquiry Professor FORBEs discovered " that at least two metals were perfectly inert in either situation , namely , antimony and bismuth .
" Considering the explanation given , that the effects are due to the mechanical repulsion exerted by the heat in its passage from a good conductor to a bad one , the inertness of the two bodies mentioned presents a grave difficulty .
Reflecting on the subject , the thought occurred to me , that if a mass of bismuth or antimony were cut so that the plane of most eminent cleavage might be vertical , the superior conductivity which the mass probably possesses in the direction of the said cleavage might aid in the production of the vibrations .
I cut such a piece from a mass of antimony and fixed it in a vice , so that the horizontal edge on which the large brass rocker rested was perpendicular to the surfaces of principal cleavage .
Loud and sustained vibrations were the consequence .
I repeated the experiment in the case of bismuth with equal success ; and after a little practice found that the precaution of cutting the substances in the manner just described was wholly unnecessary , and that tones could be obtained with facility , no matter what might be the direction in which the mass was cut .
We have thus proved antimrony and bismuth to be active in one position at least ; but antimony is active both as rocker and bearer .
Two irregular masses , the one weighing about a pound and the other five pounds , were so filed down as to plesent suitable surfaces for rocking .
Heated , and placed upon a flat mass of lead , both masses vibrated permanently .
These experiments add their evidence to that already adduced against the third law ; for antimony is a worse conductor than lead , and antimony is here the hottest metal .
These results appear to leave the theory of Professor FORBES without any foundation .
One point only remains to be considered .
Professor FARADAY attributes the superiority of lead as a bearer to its great expansibility by heat , combined with its deficient conducting power .
Against this view Professor FORBES argues in the ingenious manner already described .
It cannot be denied that when the supporting metal is a good conductor a greater quantity of heat will pass into it during contact than when it is a bad one .
It cannot be denied that the greater the quantity of heat transmitted , the greater will be the expansion ; and hence the conclusion seems unavoidable , that , if the vibration be due to expansion , both rocker and bearer , other things being equal , ought to possess the power of conduction in the highest possible degree .
Assuming then that the effects are produced by ordinlary expansion , the argutment of Professor FORBES stated in its severe logical form would be as follows : The greater the expansion the greater will be the effect ; but , The greater the conducting power the greater is the expansion : therefore , The greater the conducting power the greater will be the effect .
This , to all appearance , is conclusive .
A slight inadvertence , however , in the use of the term 'expansion ' appears to deprive the argument of much of its force .
In the first proposition the term means expansion in a vertical direction ; for if this be not meant the proposition would be untrue .
In the second proposition , however , it is the total expansion that is referred to* .
Now supposing the conductivity of the bearer to be infinite ; that is to say , that the quantity of heat which it receives from the rocker during contact is instantaneously distributed equally throughout its entire mass , then , although the total expansion might be very great , there would be no local expansion at all , and therefore none of the effects in question .
The expansion we require is a sudden elevation of the point where the rocker comes into contact with the bearer , and it is manifest that " a rapid communication with the interior " may , by suddenly withdrawing the heat from the point where it is communicated , almost extinguish the requisite elevation , and thus prevent the vibrations .
This appears to be the precise reason why Professor FORBES has failed to obtain the numerous results described in the foregoing pages .
His bearers were of such a form that the mass of matter inmmediately surrounding the point of contact quickly abstracted the heat communicated to that point , and thus destroyed the condition upon which the vibrations depend .
The success of the experiments described in this memoir depends on the precaution , that the abstraction of heat was prevented , to some extent , by reducing the bearers to lamine and mere spikes ; and the fact that a thin edge gave a better tone than a thick one thus receives a full explanation .
These considerations , 1 think , render it clear that the cause of the superiority of lead assigned by Professor FARADAY is by no means an " oversight .
" On the other hand it would not be safe to affirm generally , nor has it been affirmed by the philosopher last mentioned , that the less the conducting power the greater will be the effect .
In the case of glass and earthenware the vibrations soon come to an end , for the requisite difference of temperature between rocker and bearer , as anticipated by Professor FORBES , soon ceases .
Perfect non-conductibility would be just as inefficacious as perfect conductibility , and the region of practical results lies between these two extremes .

