On some Thermo-dynamic Properties of Solids .
By J. P. JOULE , LL. D. , FRI .
S. , F.C.S. , Hon. lem .
Phil. Soc. Cambridge , Vice-President of the Lit. and Phil. Soc. Manchester , Corresp .
Men .
B.A. Sc. Turin , &amp;c. Received April 22 , -Read June 10 , 1858 .
1 .
AFTER finding the numerical relation between heat and work in 1843 , it immediately occurred to me to investigate various phenomena in which heat is evolved by mechanical means , and of these one of the most interesting and important appeared to be the evolution of heat by the compression of elastic fluids .
If the heat given out in this case proved to be the equivalent of the work spent , then the natural inference was that the elastic force of a gas and its temperature are owing to the motion of its constituent particles , both being proportional to the square of the velocity of the particles ; and that the force of a falling body employed in compressing a gas is exhibited anew in the form of temperature .
On the other hand it was possible , secondly , to conceive of an elastic fluid which would not give out any heat by compression .
This would be the case if it was made up of mutually repelling particles , the temperature of which was that of the mass .
The work required to compress the fluid would be the same on either hypothesis , but in one supposition the effect would be developed in actual energy , in the other in the potential form .
Thirdly , we may suppose a fluid exhibiting as heat a portion of the force employed in its compression , and retaining the rest in the potential form .
Or we may have a fluid giving out more heat than the equivalent of the work spent upon it when it is compressed and maintained at a constant Fig. 1 .
temperature .
Experiment proved that the heat actually l evolved was very approximately that due to the compressing force .
Nevertheless , it seemed desirable to demonstrate the possibility of a gas so constituted that heat shall not be evolved by its compression .
It occurred to me || that a bag full of elastic metallic springs would illustrate A such a gas .
If the springs were properly formed , the elasticity of the bag would follow the laws of gaseous __ ___ pressure .
To the question , Would such a bag evolve heat on compression !
I could readily answer , no ; for in the bending of a spring one part is extended while the other is compressed , and thus it might be expected that WY the thermal effect on the whole would be neutral .
Still it seemed desirable to decide the point by experiment .
The apparatus I employed is represented by the adjoining sketch , where a spiral spring of tempered steel is seen immersed in the can A. By applying weights to a lever connected with the link B the spring could be compressed , and the heat evolved , if any , measured by the increase of temperature observed to take place in the water or mercury filling the can .
The plan I pursued was to note the temperature successively ; first , two minutes after the weight had been laid on ; second , two minutes after the weight had been removed ; and third , after two minutes more had elapsed .
The mean of a number of these observations taken in succession gave me , -first , the thermal effect resulting from the laying on of the weight and the atmospheric influence ; second , that of the removal of the weight and the atmospheric influence ; and third , the atmospheric influence alone .
3 .
The pressure applied was 318 lbs. , which pushed down the spiral spring 1'136 inch .
In the following summary of results , each number is the mean of eight or ten observations , given in terms of the graduation of a thermometer , of which each degree was equal to *0558 of a degree Centigrade , First Series .
Spring immersed in 8 ozs .
of water .
Weight laid on .
Weight taken off .
Atmospheric influence .
First experiment ... . .
-548 -*568 -'560 Second experiment -'009 -'023 -*012 Third experiment ... ..+ '282 + '271 + '252 Mean ... . .
-092 -.107 -*107 Second Series .
Spring immersed in 7 lbs. of mercury .
Weight laid on .
Weight taken off .
Atmospheric influence .
First experiment ... .
+ 180 +'175 +'169 Second experiment v. + 066 + 092 + 059 Third experiment.+ 041 + '039 +'020 Fourth experiment ... -060 -0059 -053 Mean .
+057 +-062 +'049 Mean of both series. .
-'032 -'038 *041 4 .
The capacity for heat of half a pound of water being about twice as great as that of 7 lbs. of mercury , I have divided the mean result of the second series by two , before combining it with the first series in a general mean .
The mean of both series therefore represents the thermal effect in the capacity of half a pound of water .
It indicates , on subtracting the effects of the atmosphere , a heating effect of '009 after laying on the weight , and a heating effect of '003 on taking off the weight .
The highest of these numbers represents a temperature less than one thousandth of a degree FAHR. 5 .
Now the actual force expended in the compression of the spring was 14'268 footpounds , which is equivalent to 00.037 FAHR. in half a pound of water , a thermal effect which would have been made manifest with the greatest facility .
Hence it was obvious that a gas might be conceived as so constituted that the heat evolved by its compression would be in no respect the equivalent of the mechanical force employed , and that therefore we had no right to assume such equivalency except as a hypothesis to be tested by experiment .
Accordingly after this hypothesis had been proved by me approximately* , Professor THOMSON devised the experimentst by which we have succeeded in defining the limits of its accuracy .
The same philosopher has also applied his powerful analysis to the investigation of the thermo-elastic properties of matter4+ The results of the experiments I have just given an account of can only be considered as negative , but it has been decided by Professor THOMSON that the compression of a spring gives a certain , though excessively small thermal effect , owing to the almost exact counterpoise of heating and cooling effects on the compressed and extended sides .
The method of obtaining appreciable results was obviously to examine these opposite effects separately .
I have , therefore , on the suggestion of Professor THOMSON , undertaken some experiments with a view to ascertain the heat developed by longitudinal compression , and that absorbed on the application of tensile force .
6 .
At the outset it was obvious that a very delicate test of temperature would be required , and no means appeared to offer so many advantages as that of thermo-electricity .
Professor J. D. FORBES had constructed a thermo-multiplier capable of detecting temperatures not exceeding one thousandth of a degree FAHRENHEIT .
Adopting some of the refinements introduced by MELLONI and FORBES , I have simplified the instrument so as to render its construction and management very easy , and also increased its sensibility by immersing it into the vacuum of an air-pump .
My thermo-multiplier is represented by the adjoining sketch ( fig. 2 ) , where a is the air-pump firmly clamped to a strong stool , the legs of which pass through holes in the Fig. 2 .
laboratory floor and are driven into the ground beneath ; b is a glass chimney receiver ; ca block of wood supported on feet which rest on the pump plate ; da piece of glass rod fixed to the block , over which is thrown the filament'.which supports the astatic needles .
Two thick copper wires ( e ) dip into mercury cups formed in the block , and , being carried out of the receiver through holes drilled ill the ground glass plate f , are bent into two mercury cups ______ placed on the top of the instrument .
Pitch was employed a to close all orifices air-tight .
_l 7 .
The details of the astatic needles , which are poised &amp; 0 according to the plan first suggested by Professor THOMSON , will be better understood by inspecting fig. 3 .
The needles are parts of one sewing-needle magnetized to saturation , one part being a little longer than the other , so as to exceed it in magnetic moment and give direction to the system .
A piece of glass tube drawn very fine , and bent as represented in the sketch , is attached at right angles to the upper magnetic needle and serves as Fig. 3 .
the pointer .
The lower needle is hooked to the pointer by means of the fine glass tube to which it also is attached .
The coil consists of twenty turns of silked copper wire , -416th of an inch in diameter , the ends of which dip into the , mercury cups gy formed in the block of wood .
8 .
In order still further to increase the sensibility of the instrument , a steel magnet one yard long , the permanency of which had been tested , was placed so as to counteract and almost entirely overcome the action of the earth 's magnetism in the locality of the needle .
A small telescope placed at the distance of a few yards , and looking obliquely downwards through the chimney glass at the graduated circle , completed the apparatus .
9 .
With air in the receiver at the atmospheric pressure , the mere standing at the distance of two yards on one side of the instrument , would in a short space of time cause the needle to travel through 100 in consequence of the currents of air produced by the unequal heating of the walls of the glass receiver .
But when the air was reduced to a pressure of only half an inch in the mercury gauge , this did not take place , though still , when the hand was put in contact with the receiver , a very considerable deflection of the needle was -speedily produced .
10 .
On working with my instrument , I was agreeably surprised to find that when the bar-magnet was placed so as to make the needle take up one minute in being deflected to a new position , no perceptible return swing of the needle took place , even when the rarefaction of the air was carried to half an inch pressure .
If a small magnet was suddenly placed where it could deflect the needle 300 , the pointer would steadily travel towards that degree of deflection , and on arriving there would remain settled without any previous oscillation that could be discerned .
When the time of a swing was reduced to 30 " , a return swing was observed amounting to 15O , 25 , and --L of the first swing , according as the gauge was reduced to 1 , 1 , and 2 inch respectively .
11 .
As a test of the delicacy of the instrument arranged so as to give the swing in 45 " , I may mention that , after increasing the resistance of the coil and its appendages a hundredfold by the addition of 100 yards of fine copper wire of Birth of an inch diameter , a single degree Centigrade communicated to the junction bismuth and antimony produced a deflection of 20 57 ' .
Therefore , as it was quite possible to estimate a deflection amounting to 2 ' , it followed that a change of temperature in the junction bismuth and antimony directly connected with the multiplier would be estimable , if it were only -00 of a degree Centigrade .
12 .
An objection which might be raised on account of an air-pump being permanently occupied , might be got rid of by the following arrangement:-A rim a , fig. 4 , contains pitch or other cement to secure the joint between the Fig. 4 .
chimney and top plate glass a rim at b contains pitch to secure the orifices where the wires pass through the top plate .
The bottom of the chimney rests upon &amp; around metallic plate , into which a metal pipe c is screwed , which is attached by india-rubber tube to the glass tube d. This tube is hermetically sealed after the air of the receiver has been exhausted through it .
All the lower joints are then rendered permanently tight , by putting the instrument in the shallow dish e filled with * melted pitch , Thermo-electricity of Iron in different states .
13 .
In my earliest experiments on the thermal effect of stretching a steel bar , 1 placed a copper wire in contact with the steel , and completed the circuit by an iron wire in contact with another part of the steel bar not under tension .
I found anomalous results which I was ultimately able to refer to the strong thermoelectric relation between hard steel and iron .
It was at once obvious that the existence of such marked differences between ferruginous metals in various states of aggregation or purity might render the thermo-multiplier a valuable test in the hands of practical men : I therefore allowed myself to be diverted awhile from the main object of inquiry in the endeavour to throw some light on so interesting a question .
14 .
Professor THOMSON has described* the changes in thermoelectric position which are produced by the various conditions under which metals are placed .
He has shown that the process of hardening by plunging into cold water brings copper and most other metals nearer antimony in the thermoelectric series , but that iron is brought nearer bismuth : I find that in steel the change is in the same direction as in iron , but of enormously greater magnitude .
In a softened state , the position of steel is about -midway between copper and iron , but after hardening it by plunging it at a bright red heat into water , I have in some instances found it to be on the bismuth side of copper , the alteration of the thermoelectric position of the same specimen amounting to as much as -pth of the entire range between bismuth and antimony .
In , all the specimens of wrought and cast iron I have examined , there is a notable effect in the same direction produced by plunging at bright red heat into water ; but although a great change in the molecular state was thus occasioned , evinced by the iron bending with twice the difficulty it did when annealed , and the cast iron resisting the action of the file , either metal was only brought nearer bismuth by 200 of the interval between bismuth and antimony .
A fresh illustration of the extraordinary physical change produced in iron by its conversion into steel is thus afforded ; and I believe that the excellence of the latter metal might be tested by ascertaining the amount of change in thermoelectric condition which can be produced by the process of hardening .
15 .
The different varieties of cast iron I have tried present a surprising range of thermoelectric intensities , extending almost from that of wrought iron on the one hand , to that of German silver on the other .
By the kindness of Professor F. C. CALVERT , I have been enabled to examine several interesting specimens , of which he has furnished me with the analysis .
The general conclusion arrived at is , that the metal is brought nearer bismuth as the quantity of carbon in combination is increased , but much more so than would be the case if cast iron exhibited merely the combination of thermoelectric intensities which carbon and iron separately possess , at least if the intensity generally assigned to carbon is correct .
The intensity of the junction wrought iron and highly carbonized cast iron is as much as one-fifth of the intensity of antimony and bismuth .
Thermo-electric Intensities of Metals , Alloys , &amp;c. 16 .
In constructing the following Table , I employed the thermo-multiplier already described , furnished with a variable extra resistance .
I examined first those metals whose thermoelectric qualities were most widely separated , and then those which lay intermediate and differed less from each other .
Small arcs of deflection were observed , so that the deviation of the needle was a sufficiently correct measure of the intensity of the current , and constantly repeated comparisons were made with a standard thermoelectric junction of copper and iron included in the same circuit with the junctions experimented on .
17 .
Scale of Thermo-Electric Intensities at 120 Centigrade .
Antimony , specimen of commercial .
100 Antimony , pure , prepared by Mr. CALVERT..98'06 Antimony , pure , prepared by Mr. CALVERT , not well annealed..95*95 Alloy consisting of 5 equivalents of antimony +1 equivalent of bismuth 83'75 Alloy consisting of 4 equivalents of antimony +1 equivalent of bismuth 80'50 Alloy consisting of 3 equivalents of antimony +1 equivalent of bismuth 78'94 Alloy consisting of 2 equivalents of antimony +1 equivalent of bismuth 68'79 Alloy consisting of I equivalent of antimony +1 equivalent of bismuth 45'51 Iron , thick wire , Professor CALVERT'S..80 ' Iron , thin wire..79-24 Iron , thick wire , very well annealed. .
78-24 Iron , thick wire , hardened by plunging at bright red into water ... 77'62 *Iron ( Professor CALVERT'S No. 1 ) drawn into wire ... 77'23 tlron ( Professor CALVERT'S No. 1 ) after puddling , but previous to being drawn into wire..76'00 ?
Iron ( Professor CALVERT'S No. 1 ) , cast , previous to puddling ... 65 2l Carbon .
Silicium .
Sulphur .
Phosphorus .
Iron .
Total .
Cast iron ... 2'275 2 720 0 301 0-645 94 059 100 Puddled bar. .
0-296 0-120 0-134 0-139 99-311 100 Iron wire ... 0111 0 088 0 094 0.117 99 590 100 From the above digression I now return to the main subject of the present paper , by describing my Experiments on the Thermal Effects of Tension on Solids .
18 .
All the metals employed , with the exception of lead , were in the form of cylindrical bars about a foot long and a quarter of an inch in diameter .
The upper end of the bar was screwed into a piece of metal supported by a wooden framework , the lower end was attached to a lever , to the extremity of which weights could be hung without approaching the apparatus .
In the first method the thermoelectric junction was made , as represented in the adjoining sketch , by binding to opposite sides of the bar copper and iron wires 4---th of an inch in diameter , hammered -flat at the ends .
In the second method , the junction of fine wires was inserted in a hole -1-th of an inch in diameter , bored through the centre of the bar , a small portion of wire being bound to the bar by means of cotton thread , but metallic contact prevented by an intervening slip of paper .
The terminals of these wires were immersed in deep mercury cups formed in a solid block of wood , whence thick copper wires proceeded to the thermo-multiplier .
19 .
Immediately after each experiment on the effect Fig. 5 .
of tension , the thermometric value of the deflections was Ft h. od ... . .
Be~~~~~~~~~Frst method .
Second method .
ascertained by immersing the bar , to within one-third of an inch of the junction , in water of different temperatures .
The deflections thus produced were about twothirds of those occasioned by the same changes of tem_ J perature when the junction was completely immersed .
The diminished effect in the former case is owing for the most part to the conduction of heat from the air by the thermoelectric wires .
The experiments on tension LVV\ were liable to be affected in the same way , but they were not subject to the loss arising from the conduction of heat from the surface of the bath to the junction .
The error intervening from this latter circumstance could not be great , and was moreover in all probability almost exactly neutralized by a small error in the tension experiments , arising from the escape of - &amp; th of the thermal effect from the quarter-inch : bars during the 40 " occupied by the swing of the needle .
20 .
Iron.-The weight of the bar per foot was *1568 of a lb. The lever alone gave it a constant tension of 70 lbs. The additional tensions successively given were 194 lbs. , 388 lbs. , 583 lbs. , and 1166 lbs. , or as 1 , 2 , 3 , and 6 .
The mean of five trials gave 20'46 as the deflection indicating cold on applying the tension of 194 lbs. , and 20 ' 2 as the deflection in the opposite direction , indicating heat , when the weight was removed .
With 388 lbs. , I had 35 ' 8 and 42 ' 2 : with 583 lbs. , 56'A4 and 57 ' 8 : and with 1166 lbs. , 10 55 ' 8 and 10 58 ' 6 .
Hence it appeared that the quantity of cold produced by the application of tension was sensibly equal to the heat evolved by its removal ; and further , that the thermal effects were proportional to the weights employed .
21 .
The above trials having been made without an accurate determination of the thermometric value of the deflections immediately afterwards ( a practice I found it important to follow in order to obviate the effects of any alteration in the magnetic intensities of the earth , astatic needles , or controlling magnet , which might be taking place ) , I repeated the experiment , using the first method above described , and found a mean deflection of 10 22 ' 8 with a tension of 775 lbs. The thermometric value of this deflection , determined in the manner already explained , was 00.115 Centigrade .
22 .
Professor THOMSON , in his researches on the thermo-elastic properties of metals , has demonstrated the following formula as applicable to the phenomenon in question , t peII where H is the thermal increase in degree Centigrade , t the temperature Centigrade from absolute zero , J the mechanical equivalent of the thermal unit , p the pressure applied in pounds , which in the case of tension is of course a negative quantity , e the longitudinal expansion per degree Centigrade , s the specific heat , and w the weight in pounds of a foot length of the bar* .
Applied to the above experiment the formula gives 1390 1 X81200X yX.1568 1 23 .
In another experiment by the first method , but with another thermoelectric junction , I obtained a thermal effect of 00.124 ; in this case the formula gave H= 7X 75X1 X X- .
, ,XX1568001099 .
24 .
A third experiment , in which the second method was used , and in which I had the advantage of Professor THOMSON'S assistance , gave a thermal effect of00-1007 .
In this case the formula gave 28 -725 111 H=j39X _X 2X- .
, , X. 568=-0 ? ?
1069 .
25 .
Hard Steel.-With stretching weights of 194 , 388 , and 775 lbs. , I observed deflections of the needle amounting to 27 ' , 48 ' , and 10 37 ' .
Another experiment by the * The formula signifies in fact that the heat evolved by compressing a solid is equivalent to the work required to compress the volume of it due to temperature , just as in the case of a perfect elastic fluid , applied to which the expression becomes simplified to -P first method to obtain the absolute thermal effect gave 53 ' , which was found to indicate a temperature of 00.162 .
The theoretical result in this case was 275-4 775 1* it 1 H=390X 758163 X -04 X 101 26 .
Cast Iron.-The deflections produced by tensile forces of 194 , 388 , and 775 lbs. were 17 ' , 31 ' , and 59'-9 respectively .
The thermal effect indicated by the last was 0-16 05 .
The formula gives H 277-4 -775 1 1 ?
X 1=_ 1390 1 X90139 X1 98x.128127 .
In another experiment I obtained a thermal effect of -1481 , theory giving 283 -784 111 H--X 78X lXlX1o H139 01 90139 1281 28 .
Copper.-With tensile forces of 191P6 , 383-3 , and 766-6 lbs. I obtained by the first method deflections of 21 ' , 41 ' , and 10 12'16 respectively , the absolute thermal effect indicated by the last being 00.174 .
The formula gives 274-9 -766-6 1 1 ?
f 1 H= lsoX --XX I8-2 X.5X 007154 1 x58200x 'x1781=-0014 29 .
With copper , and also with iron wires , I have observed the cooling effect of tension close up to the breaking-point .
But whenever the pull occasioned permanent change of figure , the heat frictionally evolved overpowered the cooling effect .
30 .
Lead.-The bar employed was half an inch in diameter , and weighed 6308 grains to the foot .
The stretching weights were , first , 193 lbs. with 70 lbs. , the weight of the lever , constant ; and second , 263 lbs. , the weight of the lever itself being used along with the 193 lbs. laid on it .
Using the first method , I obtained with the above weights deflections of 21'-5 and 30 ' 7 , indicating thermal effects of -00.0531 and -00 ?
0758 .
The formula gives in the two cases , H2785X-193 1** Il H= ~x3510x.03 x --00-0403 , 1390 l 35100 03 901 and _278 5 x-263 111 H~ 3X3510X.00X -.00.0550 .
-1390 1 35100 0 .
*0 31 .
Gutta Perchl.-The piece used was half an inch in diameter , and weighed 1246 grains per foot .
The thermoelectric junction , formed of very thin copper and iron wires , was inserted in a slit made two-thirds through the substance of the gutta percha .
The weights employed were , first , that of the lever alone amounting to 70 lbs. ; and second , the lever with the further addition of 80 lbs. , making in all 150 lbs. As with the metals , cold was produced when the gutta percha was stretched , and the heat restored when the stretching weight was removed .
With the first-named tensile force a.mean deflection of 16'-4 was observed , and with the second a mean deflection of 30 ' 2 .
S SME .
TON .
t POTTER .
+ RoY . ?
Myself .
1i Livoisim and LAPLACE . ?
T DULONG and PETIT .
*@ LAVOTSIER and LAPLACE .
In order to turn the above into thermal measure , the gutta percha was plunged into water of different temperatures , deep enough to immerse the part in which the thermoelectric junction was imbedded ; the former plan of operation being inapplicable in this instance , on account of the low conducting power of the substance used .
It was found in this way that the quantities of thermal effect due to the above deflections were -0-0 ?
284 and -0 0524 , quantities , however , which are likely to err in defect by a very small quantity , owing to the nature of the method of finding the value of the deflections .
The theoretical results , assuming as I did for the metals that the expansion of gutta percha by heat is the same whether suffering tension or not , are _276_ -70 11l 1390 X1 X6354 X 402 XO and 2764 -150 111 OoO65 0X -1 1X 6-3-54X X 7j--0656 .
32 .
The values just given for the expansion and specific heat of gutta percha are those which I arrived at by experiments on a part of the same sheet out of which the cylinder above used was made .
The method pursued in obtaining the expansion was one I have frequently found very convenient .
It consists in weighing the body in water at two temperatures , one as much below as the other is above the point of maximum density .
In this way I found the linear expansion between 20.4 and 50.8 to be 6A54 per 10 , and the specific gravity to be 1P00462 at 40 .
The specific heat at 40.5 was found by the method of mixtures to be *402 .
33 .
India-rubber.-The extraordinary physical properties of this substance appear to have been first remarked by Mr. GoUGH X. By placing a slip of it in slight contact with the edges of the lips , and then suddenly extending it , he experienced a sensation of warmth arising from an augmentation of the temperature of the rubber ; and then by allowing the slip to contract again , he found that this increase of temperature could be destroyed in an instant .
In his next experiment , he found that " if one end of a slip of caoutchouc be fastened to a rod of metal or wood , and a weight be fixed to the other extremity , in order to keep it in a vertical position , the thong will be found to become shorter with heat and longer with cold j. " The third experiment described by Mr. GOUGH has indicated the means of using this substance in the production of textile fabrics ; ^ he says , " If a thong of caoutchouc be stretched , in water warmer than itself , it retains its elasticity unimpaired ; on the contrary , if the experiment be made in water colder than itself , it loses part of its retractile power , being unable to recover its former figure ; but let the thong be placed in hot water , while it remains extended for want of spring , and the heat will immediately make it contract briskly .
" In addition to the above , my own experience has led me to the curious fact that a piece of india-rubber , softened by warmth , may be exposed to the zero of FAHRENHEIT for an hour or more without losing its pliability ; but that a few days rest at a temperature considerably above the freezing-point will cause it to become rigid .
35 .
The ends of a piece of elastic india-rubber , about --ths of an inch square , were attached to an apparatus by means of which it could be stretched by various weights .
A thermo-electric junction of thin copper and iron wires was placed in an orifice pierced through the centre of the slip ; and at either side , at -6-ths of an inch distance from one another , pins were stuck which afforded the means of knowing the exact length to which the rubber was stretched .
With small tensile forces no thermal effects were observed , but when the rubber was stretched by a weight of 6 lbs. , a sensible deflection of the needle took place , indicating heat ; the needle returning to zero , and thus indicating absorption of heat when the weight was removed .
Experiments with different weights gave the following results : Length of rubber in inches Stretching weight Deflection of in pounds .
When stretched .
After weight thermo-multiplier .
indicated by When trethed .
was removed , deflection .
0 0.59 0'59 002 0o67 Not observed .
004 0.77 Not observed .
-0'002 60 94 Not observed .
11 0020 8 113 Not observed .
32 0 059 10 1-45 0.71 1 15 0-139 12 1.75 0.9 2 47 0-370 14 Not observed .
Not observed , 4 50 0 537 16 2-14 Not observed .
5 55 0o657 18 2-36 1.02 6 45 0 750 36 .
The low temperature ( about 6 ?
Centigrade ) at which the above Fig ( ; experiments were tried considerably interfered with the perfection of the elasticity of the india-rubber , which in consequence became gradually elongated on the removal of successively increasing weights .
This circumstance no doubt interferes slightly with the absolute quantity of thermal effect recorded in the last column of the Table .
Still the essential characteristics of the phenomena are plainly indicated , viz. that little heat , or even a reverse effect , is produced by moderate stretching weights , but that after a certain weight is reached , a rapid increase of thermal effect takes place .
37 .
The same piece of rubber was afterwards immersed in water in the manner represented by the adjoining sketch .
A wire attached to the rubber at the upper extremity was made fast , and another wire affixed to the lower end passed through a cork in the glass tube , and being bent into a hook , supported a scale for weights .
The temperature of the water in the tube was kept uniform by constant A _ additions and removals of water , which was kept well mixed by a current of bubbles of air blown from time to time through a narrow tube into the lower end .
The positions of the pins , showing the length of the rubber , were ascertained by means of a graduated scale placed behind the tube , viewed through a telescope which was raised or depressed according to the level of the pins .
38 .
In my first experiment the stretching weight , inclusive of the weight of the tube , was 3 lbs. , and the water in the tube being adjusted to 70 Cent. , the length of the thong between the pins was 1-026 inch .
The temperature of the water was then raised to 170 , and kept there for ten minutes , when the length was found to be reduced to 0X964 inch .
A third observation at 28 ' gave the length 0 9 .
It was evident that this contraction was partly owing to the recovery by heat of the rubber from the constrained state in which it was left by the experiments of the preceding Table ; for the lengths observed on lowering the temperature successively to 160.6 and 704 , namely , 0X91 and 0-922 , indicated a lengthening effect from the application of cold , only one quarter of the amount of the previous shortening by heat .
I now used a stretching weight of 6 lbs. , which immediately increased the length to 1-22 , and in the course of thirty-six hours further increased it to 2'548 inches at the temperature 504 .
Raising the temperature successively to 120 and 170 , the length was diminished to 2-527 and 2 459 .
Then on depressing it to 90.5 and 40.3 , I observed 2A476 and 2 490 .
After leaving the same weight on for five hours longer , I observed lengths of 2-559 , 2-514 , and 2X47 with the temperatures 6 ?
2 , 120 4 , and 20'-2 , and 2-483 and 2-498 at temperatures successively depressed to 120 and 50 , After the lapse of twenty-two hours more , I found the lengths 2-624 , 2-568 , and 2'51 at the temperatures 6 ?
2 , 13 ?
*6 , and 220 ; and 2-523 and 2-548 on depressing the temperature to 13 ' and 6-4 .
All these observations show the gradual elongation effected by tension , the tendency of heat to restore the original length , and the increase or decrease of elastic force with the elevation or depression of the tempera ture of the rubber .
39 .
I now reduced the stretching weight to 1 lb. , being that of the tube and the water contained by it .
The observed length was 1-63 , and in forty-five minutes further decreased to i-621 .
Afterwards it began to increase gradually ; in five hours attaining to 1i635 , in seventeen more to 1-724 , and after a further lapse of forty hours to 1-751 .
These observations were at 70.1 .
Then raising the temperature to 21'-6 , the length was reduced to 1191 ; and on depressing it again to 70.4 , the length was further reduced to 1181 .
Keeping the rubber at the same temperature it still continued to contract , and in two hours its length was only 1I1 ; .
Then on raising the temperature to 230 , and afterwards depressing it to 7 ?
^2 , the length was successively reduced to 1'151 and 14141 .
These and similar observations were continued for several days with like results , confirnl ing my previous remarks , and also developing the curious fact , that when a piece of indian rubber , which has been previously gradually lengthened by stretching , is shortened by the temporary application of heat , the shortening effect continues to go on for some time , even if the rubber has been reduced to its previous temperature .
Although my experiments , already detailed , on the thermal effects of stretching india-rubber were not , on account of the imperfect elasticity of that substance at low temperatures , well calculated to give exact numerical results , the general conclusion derived from them is perfectly in accordance with Professor THOMSON'S formula , where , in this instance e , the expansion , being negative , the result indicated is a rise of temperature on the application of the tensile force .
41 .
When , by keeping india-rubber at rest at a low temperature for some time , it has become rigid , it ceases to be heated when stretched by a weight , and , on the contrary , a cooling effect takes place as in the metals and gutta percha .
The experiment by which I ascertained this fact was made with a piece of india-rubber 4 of an inch square , which had been exposed for several days to a temperature not lower than 50 Cent. In this state it was found that the application of a tensile force of 14 lbs. produced a deflection of 20 ' , indicating cold , and that a contrary deflection , indicating heat , was produced by removing the weight .
After raising the temperature of the rubber to 15 ?
, the elasticity of the rubber was restored , and the reverse phenomenon of heat on stretching produced , indicated by a deflection of 15 ' .
42 .
" ; V ulcanized " India-rubber.--I observed the principal physical characters of this substance before I was acquainted with Mr. GOUGH'S discovery of the properties of simple india-rubber , above noticed .
The superior permanency of the elasticity of rubber in combination with sulphur , and its unimpaired existence at low temperatures , rendered it better qualified for experiments in which accurate numerical results were desired .
43 .
I determined the specific heat of vulcanized india-rubber from portions of the specimens used in the experiments about to be detailed .
The method of mixtures was employed , a quantity in small bits being raised to a given temperature in a bath of air , and then plunged into a thin copper can partly filled with cold water .
The result , reliable to at least one-tenth of its amount , was 0'415 .
44 .
The expansion by heat was found by weighing in water 20.25 above and 20.25 below the maximum density .
The result showed a cubical dilatation of 0 000526 per degree Centigrade , an expansion greater than that of any other solid hitherto examined .
45 .
Mr. GOUGH has stated that the specific gravity of india-rubber is increased by stretching it ; and to this he attributes the heat thereby evolved .
The experiment , however , is a very delicate one , and it does not appear that this philosopher possessed the means of arriving at a reliable result .
My experience with vulcanized india-rubber leads me to doubt the accuracy of Mr. GoUGH'S conclusion .
I weighed elastic bands alternately stretched and unstretched in water .
The result of the first series of experiments was , that the bands unstretched had a specific gravity *996057 , but when pulled twice their natural length , '994641 .
The second series gave me for the specific gravity of the unstretched bands 990918 , but when pulled to two and a half times the natural length , *988483 .
The same vulcanized india-rubber bands were used in both series , and the diminution of the specific gravity in the second series is owing to the constant loss of sulphur taking place .
On account , however , of the alternation of the experiments in each series , this circumstance does not affect the conclusion that the action of tension on a band of vulcanized india-rubber sensibly diminishes its specific gravity .
46 .
In my experiments on the thermal effects of stretching vulcanized rubber , a piece about 8-ths of an inch square , and weighing 1452 grs. to the foot , was employed .
The upper end was made fast , and to the lower , weights could be applied and removed at pleasure .
A thermo-electric junction of thin copper and iron wires was inserted in a slit made in the direction of the length of the thong and near its middle .
Also pins were stuck into the rubber at 4 inches distance from each other , in order to measure its length under various tensile forces .
The following is a Table of results:-Stretching Length of measured Deflection of Thermal effect weight in part , under tensile thermo.ultiplier .
the deflection .
pounds .
force , in inches .
tem-utpir h elcin 04002 4-06 -1*4 -0-003 4 4'12 -2 -0O004 7 4.3 -24 -0-004 14 4-8 -0 5 -0'001 21 5-21 7.9 0'014 28 5-87 29.1 0-053 35 6*6 51.9 0095 42 7-25 1 15*7 0137 49 7-75 1 3913 0-180 47 .
The data for turning the deflections in the third column into the thermal effects in the fourth , were obtained by immersing the rubber , to a point above the place where the thermoelectric junction was inserted , in water of different temperatures .
48 .
These experiments developed the following facts:-1st .
That the effect of laving on the weights was not sensibly different in quantity from the reverse effect of removing them .
2nd .
That with light weights and a low temperature there was a slight cooling effect on the application of tensile force , which , first increasing with the weights laid on , ultimately changed into a heating effect , increasing much more rapidly than the stretching weight .
49 .
The temperature of the thong during the experiments was 70.8 .
I found that at temperatures a few degrees higher , the reverse action with weak tensile forces did not take place , but that there was , on the contrary , a very slight heating effect .
50 .
Professor ThOMSON suggesting to me that in those circumstances in which the rubber was heated by being stretched , it would contract its length under tension when its temperature was raised , I arranged the same piece with which the foregoing experiments were made in the manner already described in the case of common india-rubber .
The method pursued was to observe the lengths at several temperatures successively raised , and then at similar temperatures descending .
The necessity for this arose from the circumstance that a gradual elongation took place , particularly at the higher temperatures , and by taking the mean of the ascending and descending series , I hoped to eliminate the error thus arising .
I give the following experiments in the order in which they were made : Tensile Temperature Length Temperature Mean tempeMean Shortening force in successively in inches successively Length .
rature of the leagth .
per degree pounds .
raised .
lowered .
two series .
Cent. from 00 .
F 13'4 4'899 9'2 5 ' ; 008 113 4'953 14 ) 22'5 4'894 21-5 4e979 22 4'936 1 32 4-888 30'8 4e952 31'4 4'920 j 85 4 48-6 4-888 ... ... ... . .
486 4-888 J 7'95 4'569 8 4'476 7G97 4.52 2 22 4'526 19'1 4'480 20'55 4'503 7 33 2 4-481 31 e5 4 481 32 35 4 481 j &gt; 5 5001 4A468 ... ... ... .
50.1 4'468 J 2'8 5'497 65 661 4.4 5.579 11 18 5468 17'4 5'612 17'7 5'540 1 21 c ) 33 5*458 32 4 5'540 32'7 5-499 212 &amp; t ~~~50 5-459 ... ... ... ... .
50 5-459 Fr 1'4 6'398 4.4 6'513 2'9 6'455 ' is 6,298 18 6'435 18 6.366 4336 6'224 34 6'302 33 8 6-263 If ' iT 4926 6'153 ... ... 41986 6153 f 1-8 7'467 3'4 7e589 2'6 7-528 J 19 7-264 18 7-432 18-5 7-348 _ 35 &gt; 34 7-121 33 7-250 33.5 7'185 687 49'8 7'009 ... ... ... ... 49'8 7'009 J -0'4 8'604 2'6 8792 S.1 8 69S 42 17 8'415 16-6 8'575 16'8 8497 42 n 34 8195 32 8'327 33 8261 62L S0 8020 ... ... . .
50 8'020 J 51 .
The above results completely justify the anticipation of Professor THoMSON , and , as we shall presently see , afford a remarkable confirmation of his theory .
Before , however , instituting a comparison between theory and experiment , it will be proper to observe that the length of the rubber at the low temperatures beginning and ending each experiment show a gradual elongation .
To this circumstance chiefly is it owing that so considerable a shortening effect took place with 7 lbs. tension ; for a tensile force of 14 lbs. having been just previously tried , a certain amount of increased length of the nature of set was produced , which being taken out by heating the rubber under less tension , caused a greater apparent contraction by heat .
On lowering the temperature again , a small contraction took place , which , being corrected for the gradual elongation of the rubber , gave me yjl09 as the expansion by heat under the tension of 7 lbs. 52 .
The following Table gives the theoretical estimate , compared with the experipiental result , for each additional tensile force of 7 lbs. ~~~~ , or mea ... pri eexornmean 1 Theoretical Experiexpansion - ' rslo etlr-mental reWeight laid on ~~~~~~by heat he1 feut krmna e suIt corWeight laid on t tween 1 or reciproproduct of sult taken ruet for in pounds .
P. e er ec o manent and s cal of mean the prefrom the elongation temporary weight per ceding five foregoing of rubber Permanent .
Temporary .
tension .
foot .
columns .
Table .
by uise .
07 251 -7 +L11 -0-002 -0-004 -0OII4 .
07 1390 -7 9009 '415 '1917 7 14 81 -7 11 +0-004 +0-003 + 0-003 1390~ 5003 '45 1759 14 21 281 2435 j'I +0-009 +0-015 +0-016 21 28 281 7 1360 II+ 0,01 8 +0-039 + 0-043 281 I~ 11 05+04 07 28 35 1390 -7 814 '415 '1194 +0-035+ 0042 +0 047 35 ~~~~~~281 111t 35 [ 42\ 1-390o t -7 -6j6 415 1 +0-050 + 0-042 ++0.050 0 42 +0-114 +0-137 + 0-155 53 .
In the last two columns of the above Table are recorded the actual quantities of heat observed in the experiments given in ( 46 ) , and the same corrected on account of the thong not having then been so much stretched by use .
I am of opinion , however , that this circumstance would not cause any material difference in the thermal effect of a given strain , and that therefore the last column but one is the proper one to compare with theory , which it will be found to do quite as well as could have been expected .
54 .
After the experiment on the contraction by heat of the thong when it was stretched by 42 lbs. , I removed all the weights except 2 lbs. , and then a day or two afterwards observed as follows : Temperature Temperature Contraction per Contraction per successively Length .
successively Length .
degree of temdegree of temraised .
lowered .
perature raised .
perature lowered .
-0-4 4-445 3.2 4-236 25 4-339 26 4-251 50 4-26 50 4-6 12 12 8 26 0 55 .
It will be observed that the effect of a rise of temperature in the above experiment was to take out the set which had been given to the thong by previous experiments at higher tension .
Afterwards , the effect of cooling , as shown in column 4 , is a contraction so great , as to prove that the set continued to be taken out after the heat which commenced its removal was withdrawn .
That this was the case was proved by repeating the experiment several times , using the same weak tension of 2 lbs. ; when the lengths at 8 ? ?
8 and 52 ? ?
8 were found to be 4-241 and 4-254 respectively , indicating an expansion of only T4W76 per degree .
56 .
On removing all tensile force , and applying a moderate warmth , the thong returned to the same length it had at the commencement of my experiments with it , viz. exactly 4 inches .
57 .
In order to try the thermal effects of stretching vulcanized india-rubber for a greater range of tensions , I inserted a very fine thermo-electric junction into the breadth of an elastic ring 2 inches in diameter and weighing 30 grains .
The ring , when pulled sufficiently to make the two sides parallel , became a double band 33 inches long , which elongated further to 68 inches with a tension of 2lbs .
The results I arrived at are as follow Permanent weight Length with Weight laid on in Length with the Thermal effect in pounds .
permanent weight addition to the additional weight in in inches .
permanent weight .
_ degree Centigrade .
0 3-3 25 658 0 110 2@5 6@8 2 10.9 -242 4.5 11 2 14'6 0'330 6-5 14'8 2 17-2 0'132 8.5 17-8 2 1815 0'088 10.5 18'9 2 19'4 0.068 12'5 1917 2 20.1 0'004 14'5 20-5 4 20-9 0'001 18-5 21'1 2 21'3 0'009 20*5 21'7 4 22 0'044 24'S 22 4 band broke .
Sum. .
1j028 58 .
The Table shows that the rise of temperature occasioned by stretching vulcanized india-rubber bands up to their breaking-point is about 1 ?
Centigrade , a quantity which I regard as erring somewhat in deficiency , owing to the attenuation of the band under high tension preventing the full communication of heat to the junction .
The thermal effect is evidently greatest when the rubber receives the greatest elongation by an additional weight .
When the band was stretched to six times its original length , little further elongation took place by increasing the tensile force , very little thermal effect being at the same time produced .
I thought it possible that in this case a small cooling effect might exist , but after many trials did not succeed in finding it , owing probably to frictional generation of heat near the breaking-point .
59 .
Wood.-The physical properties of this substance have great interest , from its various applications to the wants and luxuries of mankind .
On making experiments on the thermal effects of its extension and compression , I soon found anomalies which induced me to investigate at some length its expansion by heat , its elasticity , and the variations of these arising from such modifying influences as temperature , moisture , and tension .
60 .
The apparatus I employed will be understood from the following sketch , in which a wooden rod is represented as passing through a glass tube , which again passes through a wide tube of gutta percha .
This latter fits tightly at its lower end on the glass tube , which again is made tight to the wood by means of a piece of indiarubber tube .
Mercury or water may be poured into the interstice between the glass tube and the wooden rod , and so serve as the means of convection of heat from the water with which the gutta-percha tube is filled .
A stopcock Fig. 7 .
enables one to change this water when desired .
The ends of the rod are held by suitable clamps , which are attached , one to Li a firm upper support , the other to a lever of which one extremity has a knife-edge resting on a steel plate , and the other is furnished with a micrometer consisting of a glass plate , divided into 2 ths of an inch , examined by a microscope .
The weight of the lever alone produced a tension of 35 lbs. , which could be further increased by the addition of weights .
When a very weak tension was desired , a lighter lever was employed .
61 .
Preliminary trials informed me of the time ( never exceeding five minutes ) required to bring the rod sensibly to the temperature of the water in the gutta-percha tube ; I also found that the most __ ready means of obtaining uniform temperature in this water was to fill the tube twice , allowing a little intervening time .
The observations of the micrometer were generally made at intervals of five _ minutes ; the first observation being at a low temperature , the second at a high temperature , and the third at a low temperature again .
The mean of the first and third observations gave me the length at the low temperature , the difference between which and the length at the high temperature , gave me the expansion of the rod .
When a high tension was employed , gradual elongation frequently took place for so long a time , that I did not wait till it had entirely ceased .
The system of observation employed was , however , such as prevented error arising from this circumstance ; and since the gradual elongation was more rapid as the temperature was raised the precaution was taken to expose the rod to the high temperature for the same space of time immediately before and after each micrometer observation at the high temperature .
62 .
Bay Wood.-The piece selected was well seasoned and very straight in the grain .
Its length was 46'63 inches ; the part passing through the glass tube and exposed to change of temperature measuring 33 inches .
The diameter of the rod was -3fths of an inch , and its weight , in the first instance , about 196 grains to the foot .
Mercury poured into the inner glass tube formed the means of conducting heat from the surrounding water to the rod .
The following Table contains the observed expansions after various intervals of time : Interval of time between the Tension on the rod , Expansion successive determinations .
in pounds .
per degree Centigrade .
35 -00000.461 A few hours.35 . ? ?
4000S665 Two or three days 435 000004372 A few hours ... ... ... ... ... ... 435 00000685 Two or three days..35 ... .
A few hours.4 ... ... ... 35 -000004395 Two or three days ... . .
35 '000004003 A few hours.~~435 -000004850 Mean for 35 -000004235 Mean for 435 *000005225 63 .
The above experiments manifest most strikingly the effect of tension in increasing the expansibility of wood , that with 435 lbs. being nearly one quarter greater than that with only 35 lbs. The gradual diminution of the coefficient perplexed me a good deal at first , particularly as the mercury in contact with the wood prevented the access of air to its sides , but I was ultimately able to refer it to the gradual absorption of moisture through the pores in the direction of the length of the rod .
64 .
The rod was now exposed to a tension of 435 lbs. for twenty-two days , at the end of which time its expansion by heat was found to be 000004784 .
The weights were then removed so as to reduce the tension to 35 lbs. After remaining two days in this state , the first effect was found not to be expansion , but , on the contrary , a contraction amounting to *000002408 per degree .
Then the subsequent application of cold produced a contraction of 000005705 .
On raising the temperature again after an interval of five minutes I found an expansion of 000002170 , and then on cooling , a contraction of *000005017 .
After twenty minutes I found that the expansibility by heat , indicated by the mean effect of raising and depressing the temperature alternately , was *000003575 .
65 .
From the above it appears that wood has to a certain extent the property possessed by india-rubber , of returning from a state of strain on the application of heat .
Mr. HoDGiINSON has observed this property in iron ; I have noticed it in whalebone , hair , leather , and to a slight extent in copper .
66 .
The increase of expansibility with tension suggested that the force of elasticity of wood decreases with a rise of temperature .
To try this I affixed a graduated glass plate to the top of the glass tube ( see fig. 7 ) , while parallel and close to it another divided glass plate was affixed to the wooden rod .
The extension of the rod under various tensile forces was thus read off by means of a microscope in - , jMths of an inch and fractions of the same appreciable to -1lth of the actual divisions .
The mean of several series of trials gave me , at a low and high temperature , the following indications of the micrometer , with stretching weights increased by 100 lbs. at a time until 435 lbs. , and then decreased by the successive removal of the weights again .
Micrometer observations , Micrometer observations , Tension .
the temperature of the the temperature of the wood being 150.99 .
wood being 410,78 .
35 001 35 4-199 4'381 235 8'539 891]4 335 12'890 1 3-41 5 435 17*135 17-743 335 12 830 13-392 235 8-579 8.968 135 4'373 4-602 35 0-9231 0-268 -231 '268 67 .
17-135-==17e02 , and 17e743--2 =17e609 , which are therefore the elastic strains in Soths of an inch , produced by a tension of 400 lbs. at the respective temperatures 150.99 and 410.78 , the length of the rod being 33 inches .
These results , reduced to unity of length and temperature , give the difference between -000099953 and *000103410 , or -000003459 , as the quantity by which the expansion by heat of the wood ought to be increased by adding a tension of 400 lbs. Referring above , it will be seen that this increment is actually only about -000001 .
I believe that two causes are concerned in producing this discrepancy .
In the first place , I have reason to think that when tension is applied to wood and various other substances , a strain takes place of the nature of set besides the true elastic strain , which set is almost instantaneously taken out again on the removal of the tension ; also that this temporary set increases with the temperature .
Secondly , I think that when the expansion by heat of a body under tension is observed , the tendency of heat to take out set operating in a contrary direction to the expansive action , causes a slight apparent diminution of the latter .
It will thus happen that the apparent decrease of elasticity by heat is greater than the real , while the apparent increase of expansion by heat , in consequence of the application of tension , is less than the real .
These observations will be found to receive further confirmation from the phenomena of the elasticity of moistened wood , and also in the case of whalebone , a substance in which the characteristics of wood as regards set are very strongly developed .
68 .
YOUNG'S modulus of elasticity , deduced from the above data in terms of the length of a rod weighing one lb. to the foot , which would be extended by unity by the tensile force of one lb. , is 5539710 and 5354349 for the temperatures 150.99 and 410.78 respectively .
69 .
The same rod of bay wood was'now employed to determine the effects of hygrometric condition on its expansion and elasticity .
Mercury or water was employed as the medium for conducting heat between the glass tube and the wood .
When mercury was employed , some inconvenience resulted from its entry into the pores of the wood .
Ultimately I found that the heat was conveyed with sufficient rapidity when air only occupied the narrow space intervening between the glass and the wood .
Medium of YOUNG'S modulus of Remarks on the Weight Proximate communicaLmits o erLinear contracelasticity or length of Remarkson the Weigh sPecxifictioe fha eso perature between condition of of rod specific tion of heat Tension which the expe .
tion or expansion rod , 1 lb. per foot , the wood .
per foot gravity .
between the in pounds .
riments were per degree which would be in grains .
glass tube made .
Centigrade .
extended unity by and the rod .
1 lb. tension .
lbs. oz. 0 Dried before the fire for several days un , til it could be eleec 45 36P9 and 7-35 -00004i666 E. trified by rubbing 2042 428 Mrcury ... 16 5 38 5 and 8-9 -000045928 E. 203732 with the hand , and.'retain the charge a long time ... ... .
) .
Boiled in water a few minutes , then left 1-122 Water .
45 33q1 and 4-38 000004497 Cimmersed in water 593-2 t45 52 and 6-1 *000006455 C. for one day ... ... .
5a 72 .
The length of the wood saturated with water was 1-062 of its length in the dry state .
In the saturated condition it broke after a tension of 17lbs .
had been applied for half a minute .
The specific gravity of the wood , considered apart from the air or water absorbed , was found to be 1-506 .
73 .
The following Table contains the results of experiments with a rod of deal , similar to the last , also cut across the grain : Medium of l YOUNG'S modulus Weight Extreme communicaLimits of temperaLinear contracof elasticity or Remarks o the conition of of rod Tesinlength of rod , Remarks on the condition of perfoot rlength tion of heat Tension ture between which tion or expansion we ngt hofr the wood .
per oot in between the i the experiments per degree Centiweighing 1 lb. per in inches .
glass tube pounds .
were made .
grade .
foot , which would grains , and the rod .
be extended unity by 1 lb. tension .
Dried before a fire until it could be electrified byfric193-5 15-625 Mercury 43 37-5 and 10-9 *000042470 E. 188120 tion ... ... ... ... ... ... .
On immersion in water , the wood absorbed in a few minutes sufficient moisture to make it weigh 243 grs. per foot .
Its length was then 15-8 ; but after allowlng it to rest two hours its 228*2 16-3 Mercury 4-3 35-1 and 10-7 -000065153 E. 104600 length became 16-24 , without receiving any additional waterin the interim .
Twelve hours afterwards its length was 16-3 , and its weight per foot 228-2 grs ... ... ... .
) Further immersion in water. .
270-5 16-594 Mercury 4-3 41-3 and 10-91 -000000380 E. 60033 Immersed several hours in water ; specific gravity then 446 16-625 Mercury 4-3 43-2 and 10-95 -000005404 C. 38257 was found to be -891 ... ... Boiled in the exhausted re- ) l ceiver of an air-pump , and 4-3 34-5 and 13 25 000002201 C. afterwards left immersed in 5734 1i6-7 Water. .
3 345 and 1325 -000002201 C. ] water several hours ; sp. gr. 17.3 27-2 and 12-12 -000001105 C. 29618 1-143 ... ... ... ... ... ... ... ... , Dried a little by placing it a 434-4 16-63 Air 43 39-03 and 128 |000003159 E. short timebefore afire ... f 4440 16^63 Water. .
4.3 34-9 and 12-7 -000001561 C. Driedstill further f 3581 16-62 Air 4-3 37-38 and 13-3 |000010390 E. 370 1 6-62 Water. .
4-3 43-8 and 15 ?
000001356 C. ( continued ) .
Medium of YOUNG'S inoduIus Weight Extreme communicaLimits of temperaLinear contracof elasticity or Remarks on the condition of of rod length tion of heat Tension ture between which tioni or expansion length of rod , the wood .
per foot in between the the experiments per degree Centi .
weighing 1 lb. per in inches .
glass tube pounds .
were made .
grade .
foot , which would grains , and the rod .
he extended unity by 1 lb. tension .
Dried still further ... ... ... . .
231-5 16-48 Air. .
; 4.3 37-7 and 10-24 -000084212 E. 96170 Water poured betaeen the.rod ... . .
W ... .
Water ... 4.3 372 and 10,3 000058836 E. 82430 Then after a short time ... | After ten minutes more had ... ... ... ... .
Water ... 43 40 ?
9 and 12-15 '000019457 E. elapsed ... ... ... ... .
After another ten minutes ... ... ... ... ... ... Water ... , 4*3 37-6 and 12-3 *000004815 E. After another tell minutes ... ... ... ... ... ... Water ... I 4-3 42 and 12-3 *000001017 C. After another ten minutes ; W23 16-6s6 Water..i 4-3 39-34 and 12-53 000002951 C. sp. gr. -5646 ... ... J Air .
4.3 37*5 and 917 *000000817 C. Dri ed om un.a.mode .
I 206 15455 Air 4-3 41-5 and 9.77 -000038620 E. Immersed in oil under the l exhaststed receiver of an 329 15-54 Air .
4*3 38-8 and 10-6 -000043670 E. air-pump ... ... ... ... ... ... 74 .
The specific gravity of the wood , considered apart from the air or water in its pores , was in the above specimen found to be 1-595 .
75 .
The following are deductions from the above experiments:-lst .
That tension increases the expansibility of wood by heat .
2nd .
The expansibility of dry wood cut across the grain is about ten times as great as its expansibility when cut in the direction of the grain .
3rd .
The length of wood cut longitudinally is a little increased by the absorption of water ; but when cut crossways , a very great augmentation , as is well known , takes place .
4th .
When water is absorbed into the pores of wood cut longitudinally its expansibility decreases , until ultimately it passes into contraction by heat .
5th .
When water is absorbed into the pores of wood cut across the grain , its expansibility appears to increase in the first instance , then decreases until it changes into contraction , which , after increasing with the absorption of water to a certain point , seems to diminish as the wood becomes completely saturated .
6th .
Cut lengthways , the elasticity of wood appears to be lessened by the presence of water ; when , however , it is cut across the grain , the effect of moisture is very greatly to impair the elastic force .
7th .
A short period of immersion enables wood cut crossways to take up sufficient water to make it contract by heat ; but , on the contrary , wet wood partially dried on the surface is expanded by heat .
So that we may have contraction or expansion of wood charged to the same extent with water , according probably to peculiarities in the distribution of the water among the pores .
76 .
When wet wood , cut across the grain , is dried , it gets shorter very gradually , in consequence of a tendency to retain its former dimensions .
The strain thus arising may be removed by raising the temperature , which speedily reduces the wood to the length due to its state of dryness .
The same kind of effect takes place , but in the reverse order , in wood which is taking up water .
It does not immediately assume the length due to the state of hygrometry , but if heated the set is taken out , and the wood expands to its proper length .
77 .
Inasmuch as at a certain degree of humidity wood contracts in every direction with a rise of temperature , it might be inferred that on weighing it in water its specific gravity would be found to increase on the elevation of temperature .
Such , however , is not the fact .
A piece of saturated wood , being part of the specimen employed in the last series of experiments , was found to weigh 392-6 grs. in air , 42 26 grs. in distilled water at 00 , and 42'01 grs. in distilled water at 8 ? .
Hence the cubical expansion was *0000892 per degree , but calculated on the wood considered apart from the water in its pores , 000396 .
Therefore on raising the temperature a decrease of specific gravity occurs simultaneously with a diminution in the external dimensions of the wood .
This can only take place in consequence of a contraction of the surfaces of the walls of the cellular structure , while the actual bulk of the material of which they are composed is increased , a certain minute quantity of water exuding at the same time .
78 .
The phenomenon in question is therefore , I believe , owing to capillary attractions , which , diminishing with elevation of temperature , has the effect of removing a part of the swelled condition due to that action .
Dr. YOUNG has given -y-15 as the descent of water in a narrow tube due to a rise of temperature of 10 Cent. The fraction arrived at by M. C. WOLF is 52 To apply the latter estimate to the foregoing results , we find from the last Table that between extreme dryness and perfect humidity the wood increased in length from unity to 1P068 .
Hence , on the hypothesis that the contraction by heat is owing to diminution of capillary attraction , it ought to be equal to -0001293 minus the expansion in the dry condition .
An examination of the foregoing Tables will be found to afford ample support to this view .
79 .
Ratan Cane.-My observations of the expansion and elasticity of this substance are comprised in the following Table .
The length exposed to thermal influence was 33 inches .
Medium of YOUNG 's modulus of Weight communicaLimits of temperaLinear contracelasticity , or length of Remarks on the condition of rod per tion of heat Tension in ture between which tion or expan-rod , weighing 1 lb. per of the wood .
foot in between the pounds .
the experiments sion per degree foot , which would be I grains .
glass tube and were made .
Centigrade .
extended unity by 1 lb. the rod .
tension .
1094242 at 10 6 jDriedbyplacingitbeforea 150*2 rur J 35 25 3 and 11'67 '000024750 L. 9 918179 at 34 fire for a short time f.ercury 91 28-05 and 12'09 '000027000 E. 956720 At 29 1048546 At 12 Wettedbyimmersion under 208 Water ... 35 26%6 and 11-04 -000001436 C. water for a few hours ... f Dried by exposure before a 15 V ?
O A. Otf17A ' 247 t1 Driedbyeposurebforea } 14 e MrCu r y ... 3 35 127-8 and 11*1 I-000016740 E. I12 14179 at -12 fire for several hours. .
ecy. .
5 2oa l uuuz id ; __ ~_ ______ __ .
___ _ .
___ 80 .
The following are the results of my experiments on rose , vine , and poplar .
These were sprouts of recent growth , selected for their straightness .
They were cut from the tree in November , directly before the experiments .
I have also inserted in the Table the expansion and elasticity of wheat-straw .
Medium of YouNG 's modulus of Weight communicaLimits of temperaLinear expansion elasticity , or length of Kind of wood , &amp;c. of rod per tion of heat Tension in ture between which or contraction rod , weighing 1 lb. per foot in between the pounds .
the experiments per degree foot , which would be grains .
glass tube and were made .
Centigrade .
extended unity by 1 lb. the rod .
tension .
Rose wood in the green 35 46 and 14'53 000000240 E. state , the bark not re171'6 Water 91 42 and 16.000000J40 E. moved ; sp. gr. *638 ... 91 42 and 16 000001101 E. Vine in the green state , } | 116 Water ... ... 35 28'6 and 1 13 '000002121 E. Poplar in the green state , 1586 Water 35 232 and 10 '000000773 C. bark not removed ... ... f 40 39 and 9 *000000037 E. Wheat-straw dried before 14'8 Air .
52 33'4 and 11 It will be observed that an increase in the expansibility by heat of paper and leather was produced by moisture .
It is probable , however , that had they been more thoroughly saturated , the expansion would again diminish , as in the case of wood cut across the grain , in which , although the expansibility in the first instance increased with the application of moisture , it ultimately changed into contraction when the , wood became more completely saturated .
I think that a cause tending to increase the expansibility in consequence of the application of moisture exists in every case ; but that it is modified , and in wood sometimes completely overcome , by the effect of decreased capillary attraction with rise of temperature , to which I have already adverted .
84 .
Upon the facility with which hot moistened whalebone can be moulded into form , and the permanency with which it keeps the shape in which it is cooled , much of its use in the arts depends .
When heat is applied to whalebone , thus cooled in a constrained state , it at once returns to its original shape .
Even when cold moistened whalebone is strained , it returns almost immediately to its original dimensions on the application of heat .
The slip of moistened whalebone employed in the experiments recorded in the last Table , 46 inches long and weighing 71-8 grains to the foot , became rapidly stretched 7 inches longer by a tension of 64 lbs. Left to itself after the tensile force was removed , it began to return gradually towards its original length , which however it at once assumed on immersion in hot water .
85 .
The apparent elasticity of whalebone is always somewhat less than the real , owing to the observed elongation , when tension is applied , consisting of a considerable amount of set as well as of elastic strain .
Also when the tensile force is removed , the contraction which takes place consists of the return from set as well as of the elastic recoil The interesting phenomena connected with imperfect elasticity might , I think , be very advantageously studied in this substance .
The limits I had set myself did not , however , allow me to do more than obtain the following determinations of its elasticity in the dry state at two different temperatures , and with various intervals between the application or removal of tension and the corresponding observations .
Time allowed to YOUNG'S modulus of apparent elasticity , elapse between laying ' or length of rod , weighing 1 lb. per foot , which on or taking off a would be extended unity by 1 lb. tension .
tensile force of 35 lbs. , !
and making the observation of length .
At 100 .
1 .
At 3001 .
| 15 " 1176554 962450 30 " 1147383 927100 1 ' 1119624 894545 2 ' 1080415 4 ' 1043860 15 ' 1028395 jh .
0 922481 86 .
Thermal effect of tension on Wood.-My first experiment was tried on a square rod of straight-grained pine , dried , and weighing 132-4 grains to the foot .
A thermo-electric junction of thin copper and iron wires having been inserted into it , its upper extremity iI Dc CcLIx .
T was clamped to a firm support , and the lower attached to a lever giving a tension of 35 lbs. On applying an additional tension of 200 lbs. , the motion of the needle of the thermo-multiplier indicated cold ; on removing the tension , a reverse motion of equal extent indicated the evolution of heat .
The mean of several trials gave me 9'87 , the thermal value of which , determined by immersing the junction in water of various temperatures , was -01364 .
Owing to the large surface of the wood as compared with its capacity for heat , it was affected by the temperature of the surrounding atmosphere nearly three times as much as the quarter-inch metal bars .
I found , in fact , that during the time occupied by a swing of the needle , viz. 45 " , one-fifth of the thermal effect was lost .
Adding , therefore , one quarter to -01364 , we find -01705 for the thermal effect , which , it may be observed , errs a trifle in defect , owing to the conduction of a small portion of heat along the wires of the thermoelectric junction .
87 .
In order to arrive at the theoretical result , it was necessary to determine the specific heat of the wood in the condition it was in when used .
This was done by raising the temperature of a faggot of it in a bath of mercury , and then plunging it in a thin copper can filled with mercury at the atmospheric temperature .
The fall of temperature of the wood and the rise of temperature of the can and mercury , corrected for the atmospheric influence , indicated the specific heat ; which , in the case of dried St. John 's pine , was found to be -3962 , and in that of dried bay wood , -3582 .
88 .
The theoretical thermal effect of tension in the foregoing experiment , derived from the above data and the expansion by heat at the mean tension , is , 279'4 .
200 111 11=1390 XX -1 x231053 X 3960x01890 89 .
Bay Wood.-The rod I employed was dried until it could be electrified by rubbing .
A tensile force of 400 lbs. produced a deflection of 25 ' 8 , which the test experiment on the junction proved to indicate 00.0475 .
This , increased one quarter , gives -0594 as the actual thermal effect at the moment of the application of the tension .
90 .
The theoretical result , using my own determination of the specific heat of the wood , and its expansion by heat at the mean tension 235 lbs. , is 278 -400 111 H-1390x 1X 196858 X*358201886 91 .
St. John 's Pine , cut across the grain.-This specimen was dried slowly until it became electrical by rubbing .
A tensile force of 14 lbs. produced a deflection of 3 ' 4 , indicating a thermal effect of -00494 .
This , increased by one quarter , gives -00617 as the cold produced by the tension .
The theoretical result is 276 -14 111 H19X-1 X jgX.39~X _0139( ) 1 23256 *3962 03214 92 .
Bay Wood saturated with water.-The specific gravity of this specimen was 933 ; it weighed 333 grains per foot , 157 grains being imbibed water .
The mean of twenty experiments gave me -0073 of heat on applying a tension of 200 lbs. , and -0013 , likewise of heat , on removing the tensile force .
There could be no doubt that in this instance the imperfect elasticity of the moist wood caused a considerable quantity of heat to be generated frictionally .
It may therefore be safely concluded that the thermal effect , considered apart from the result of friction , was , as in the case of indiarubber , one of heat on the application of tension , and cold on its removal .
Its actual '07-0013 value would be over-estimated by 073 2 =00 ?
003 , on account of the set communicated by tension being always greater than that taken out when the tension is removed .
The approximate theoretical result is 277 -200 111 1390 X 2000000 *68 04760 93 .
For the sake of ready comparison I have collected in the following Table the results of the foregoing experiments on the thermal effects of tension , placing byS their side the results of Professor THOMSON'S theory .
Theoretical thermal effect of Material .
}3Experiment .
Theory .
1 lb. tension on a prism Material , ~~~~~~~~~~~weighing 1 lb. to the foot , at the temperature O0 Cent. Iron ... ... ... ... ... ... ... ... ... ... ... ... . .
-115 -*110 Iron ... ... ... ... ... ... ... ... ... ... ... ... -'124 -'110 -'0000220 Iron ... ... ... ... ... ... ... ... ... ... ... ... . .
-101 -*107 J Hard steel.-'162 --'125 -*0000235 Cast iron ... ... ... ... ... ... -'160 -'1 12 0016 Cast iron..-'148 -'_11.l.5 .
} -f 06 | Copper .
-'174 -'154 -0000355 Lead.- ' ... ... ... ... ... . .
053 _-040 Lead.-|'076 -'055 1 } -'0001847 Lead ... ... ... ... ... ... ... ... ... 0 ... ... ... _ 07 05 Gutta percha.028 -066 } -0000769 Vulcanized india-rubber .
+*114 + 137 Pine wood ... ... ... , ... . .
017 -'023 -'0000021 Bay wood ... ..- ' ... ... ... . .
--059 -'060 -'0000028 Pine , cross-grained.006 -'009 -'0000213 Wet bay wood ... ... ... .
+003 +001 +00000015 On the Thermal Effects of Longqitudinal Compression on Solids , 94 .
Wrought Iron.-A pillar 2 inches long and one-quarter of an inch in diameter had a fine hole bored through it , in which a thermo-electric junction of fine copper and iron wires was inserted as described in ? ?
18 .
It was placed under the lever which had served for the tension experiments , the weight of which alone gave a pressure of 50 lbs. When , in addition to this , a pressure of 1060 pounds was applied , a deflection of 21'"3 , indicating heat , was produced ; and a like deflection in the opposite direction , indicating cold , was produced by removing the pressing weight .
The value of the deflection , ascertained by immersing the pillar , in water of various temperatures , to within oneeighth of an inch from the thermo-electric junction , was 0'-1517 .
The theoretical result is ____7 X X1060 1111 Xsl00 X.714910'l1645 .
My next experiment was with a pillar of the same length , but half an inch in diameter , using a pressure of 1060 lbs. In pillars of this and greater diameters the experiment to test the value of the deflections was made by plunging the pillar in water , which rose a little above the orifice in which the thermo-electric junction was inserted .
Thus in the present.instance I found the deflection of 8'"94 to indicate a rise of temper rature equal to 0 ?
'0319 .
The theoretical result is 289 1060 1110 H= j-sso X X8200X 96 .
Wishing to try higher pressures , I employed .
a very excellent hydraulic press , the pressure exerted by which I had ascertained by experiment as far as 2000 lbs. The friction amounted to only one-seventh of the exerted force .
By supplying the correction thus indicated , I had a very convenient and tolerably accurate mode of applying pressure by simply placing weights on the handle of the pump .
A pressure of 6458 lbs. applied in this way , gave a deflection of 42 ' at the thermo-multiplier , placed at a distance of 40 yards , which deflection was found to indicate a thermal effect of 00.2344 .
The theoretical result is H 85-5x6458 111 H= 130X ~1 X81200X.1X 632=0 235 .
97 .
Next I tried a pillar 2 inches long and 1 inch in diameter .
When a pressure of 1780 lbs. was applied to this by the lever , the needle was deflected 10"6 , indicating a rise of temperature of 00.01803 .
The theoretical result is 290 _ 170 11H= 130X 7 X8120 XyX2.533 0 ?
01642 .
98 .
With the hydraulic press , pressures of 4154 , 8762 , and 20282 lbs. , produced deflections of 9 ' , 21,6 , and 10 15 ' 3 respectively , indicating temperatures of 0'03157 , 00'07578 , and 0 ?
'2642 .
The theoretical results are289__ 4154 111H =2189 X 4l X 81200 X T1 X2 533=00 03822 , 11=28958 8762 111 =0003 f139X0 X X1200 X.11 X2 533 08063 , and H= 289x820282 111 0-86 X X81200 Xb 2.533O0'l866 .
99 .
On applying a pressure of 47930 lbs. , the pillar was squeezed to the length of 1-88 inch , giving out about 6 ?
, or nearly the thermal equivalent of the work thus done .
Then using the lighter pressure of 37744 pounds I obtained a deflection of 20 6 ' , indicating a thermal effect of 00.442 , The theoretical result is 290 37744 111 H=~x x 510X.JJX~~40-3-266 .
1 pl390l 2e 20lo adon2694-e rahmi 100 .
Cast Iron A pillar 2 inches-long and one-quarter of an inch in diameter being Ill. Whien the same block was pressed , in a direction perpendicular to the grain , with* a weight of 1792 lbs. communicated by the lever , a deflection of 371'3 was produced ; but when the pressure was removed , the deflection in the reverse direction amounted only to 31'-2 .
I believe that the excess in the result of pressure is owing to frictional evolution of heat through the imperfect elasticity of wood cut crossways to the grain .
The mean , 34 ' 25 , represents a thermal effect of 0 ?
04461 .
Theory gives 288-6 1792 1110 H-1390x X 20x16 X7 0210 112 .
A second experiment gave a thermal effect of 00.016 , the theoretical result being 00 ?
0264 , In this latter experiment the position of the block was reversed .
I attribute the discordance between the two results to the difficulty I experienced in obtaining with my lever apparatus a perfectly even distribution of pressure over so large a surface as 3 inches square .
113 .
On exposing the block , still crossways to the grain , to a greater pressure by means of the hydraulic press , I obtained a deflection of 40'5 , 'indicating a rise of 00 .
whenever a weight of 4154 lbs. was laid on , but observed no perceptible effect whatever when the pressure was removed .
In this case the bulging of the wood under pressure extended to about half an inch .
It was evident that frictional heat increased to a great extent the thermal effect of pressure , and that a similar though smaller frictional effect diminished the cooling effect on removal of pressure .
The theoretical result , if bulging had not taken place , and supposing perfect elasticity , is H= 290 4154 111 -0 0609 .
114 .
Vulcanized India-ru6bber.-A pillar , 1'92 inch long , 1'22 inch in diameter , and weighing 692 grains , had a thermoelectric junction of thin copper and iron wires inserted into its centre .
On applying pressure , the multiplier showed a rise of temperature , and when the pressure was removed , a depression took place , sensibly equal to the previous rise .
The following is a Table of the results : Weight laid on the pillar , Deflection .
Heat by laying on the weight , in pounds .
or cold by removing it .
01 -28 3 040058 33 3.7 00072 47 6#7 00131 62 16 00312 93 32 00625 124 1 4-3 01254 With the pressure of 124 lbs. the pillar was compressed to the length 1'05 inch , and returned to 1-78 when the weight was removed .
115 .
The expansion of the pillar by heat , in the direction in which the pressure was applied , was found to be about 1 per degree Cent. Hence the theoretical result for a , pressure of 124 lbs. would appear to be iDR .
JOULE ON SOME THERMO-DYNAMIC PROPERTIES OF SOLIDS .
125 287 124 11 H1390X 1 X16i60X413XO77=0 ' .
The excess of the actual result may be attributed to the central part , in which the junction was placed , bearing an undue share of pressure in consequence of the bulging of the waist of the pillar .
116 .
On examining the above Table , it will be remarked that the thermal effect increases more rapidly than the pressure .
This is owing partly to the influence of a rise of temperature in increasing the elasticity of rubber under strain , to which we have already alluded , and partly to the unequal distribution of pressure to which I have just adverted .
The longitudinal expansion of a pillar of vulcanized india rubber , 2-21 inches long and 1'44 inch in diameter , under various pressures , was found to be Pressure in poun~ds .
Expansion per degree Cent. on the length under pressure .
0 1..5 00te^* ' ~~~~~~~~5 70U 55 * ' * ' 3730 55 ~~~~~~~~~~~~~~I 135 1 The above results , as in the case of the tension experiments indicate a slight excess of experiment over theory .
I at first thought that this might be owing to the diminution of elastic force by heat in metals , of which I had not taken account as in wood , for in applying or removing tension the thermal effect would be increased in consequence of the increased expansion by heat in such a case .
This cause would , however , diminish the thermal effect of the application or removal of pressure .
The discrepancy must therefore be referred to experimental error , or to the incorrectness of the various coefficients which make up the theoretical results .
Having , however , been led to believe that with a rise of temperature a certain change of elasticity takes place in metals , although too minute to be appreciated in the foregoing results , I made some experiments in which spirals weighted at one end were measured when exposed to cold air , and then again after they had been heated in the atmosphere of an oven .
In the latter case there was considerable elongation , indicating , in the case of steel , a diminution of elasticity amountingto *00041 per degree Centigrade , and in the case of copper , to 00047 .
After I had made these experiments , I became acquainted with M. KUPFFER 's valuable researches on this subject , by which , using the method of vibrations , he finds the decrement of elasticity per degree Centigrade in steel and copper to be -00047 and *00048 respectively .
121 .
Another source of error exists , which , although not of sufficient amount to make a sensible alteration in the thermal effects of tension and pressure on metals , yet ought not to be neglected in a complete view of the subject .
Mr. HODGKINSON has long ago shown that any force , however small , is able to produce a certain permanent deflection in a bar , and that this deflection increases rapidly with the force which has produced it .
Professor THoMSON has added the observation , that even after a metal has been exposed to great tensile force , its elasticity is not thereby rendered perfect for smaller degrees of stress .
Thus he finds that when weights are successively hung to a wire so as gradually to increase its tension , and then successively removed , the wire never assumes immediately its just length , but is always shorter during the putting on of the weights than during their removal .
Hence work is done on the wire which must necessarily evolve a certain quantity of heat ; and if , as is probable , a greater quantity of work is thus done whilst the tensile force is being removed than whilst it is being applied , the result will be that the cold of tension will not be diminished to the same extent as the heat , in consequence of the removal of tension , will be increased , and so the mean thermal effect will be increased .
On the other hand , it is probable that in the act of compression less work is done on the wire than during the removal of the compressing force , the result being that the mean thermal effect of applying and removing the pressure is " lessened .
The foregoing experiments do not afford sufficiently delicate tests to detect the excessively minute quantities of heat developed frictionally in the above manner .
122 .
Professor THOMSON has pointed out that the dynamical theory of heat , with the modification of CARNOT'S principle introduced by himself and .
CLAUSIUS , show that " if a spring be such that a slight elevation of temperature weakens it , and the full strength is recovered again with the primitive temperature , work done against that spring by bending or working in whatever way must cause a cooling effect .
" The quantity of cold expected was excessively small , yet I hoped to measure it by taking the mean of a large number of observations with the thermo-multiplier .
I took a spiral of tempered steel wire 1th of an inch thick , of which each convolution was 1-inch in diameter and one quarter of an inch distant from its neighbour .
A thermoelectric junction was attached to one of the convolutions , and means were provided to compress or extend the spiral at pleasure without approaching it .
In the case of a spiral stretched by a weight hung to it , the application of heat causes , as we have seen , a considerable elongation , in consequence of diminution of elasticity ; but in the case of a spiral compressed by a weight laid on the top of it , the effect of the same cause is to diminish its length .
Hence either the pulling out or the compression of the spiral must cause the absorption of heat , and the return of the spiral to its normal state must be accompanied by the evolution of heat .
123 .
The above thermal effects of bending a spring are evidently proportional to the square of the pressing or tensile force ; for if these be increased , the elastic spring and the alteration of length by rise of temperature will be also proportionally increased .
124 .
Having arranged the thermo-multiplier so as to give one swing in 30 " , I pursued the experiments as follows:-A weight of 7 lbs. was laid on the top of the spiral to compress it , and after 30 " ' had elapsed , the change in the position of the needle was noted .
Then the weight was removed , and the needle observed after 30 " as before .
One hundred such experiments were tried alternately .
Afterwards I made another series of one hundred experiments , on the effect of stretching and removing the stretching force .
The results are placed in the following Table , in which the signs + and distinguish deflections indicating evolution and absorption of heat .
coefficient for the diminution of the elastic force of steel by rise of temperature , I find for the theoretical result , in the case of compression , H= X390X1X-5 X -X0.00403 , and in the case of extension , -1390X -iX7500X.4280 043 the results being necessarily the same in both cases .
The deficiency of the actual result is not great , and is on the side of the probable error , in consequence of the unavoidable loss of a portion of the thermal effect by conduction from the junction .
126 .
Thus even in the above delicate case is the formula of Professor THOMSON completely verified .
The mathematical investigation of the thermo-elastic qualities of metals has enabled my illustrious friend to predict with certainty a whole class of highly interesting phenomena .
To him especially do we owe the important advance which has hbeen recently made to a new era in the history of science , when the famous philosophical system of BACON will be to a great extent superseded , and when , instead of arriving at discovery by induction from experiment , we shall obtain our largest access sions of:new facts by reasoning deductively from fundamental principles .

