]\gt ; XII .
On the of Con CHARLES H. LEES , M.Sc .
( Late Bishop Berkeley Fellow the ) nchester .
nunicated by Professor ARTHUB SCHUSTER , F.R. January 22 , \mdash ; Read , 1892 .
oduction .
OUR knowledge of the Thermal Conductivities of Crystals is derived mainly from the experiments of DE SENARMONT , V0N LANG , and JANNETAZ , who , using the wax melting or analogous methods , have determined the ratios of be called the " " principal conductivities\ldquo ; and the positions of the axes of conductivity within a number of crystals belonging to the simple systems .
to their experiments the isothermal surfaces about a heated point in a crystal , , in genel.al , ellipsoids , havin o their axes parallel to the optical axes .
the of a crystal , ellipsoid becomes a spheroid of revolution about the axis , and is , as a rule , oblate or prolate according as the wave-surface for the ordinary ray is oblate or .
Although this rule has a nnmber of exceptiollS , it is generaI to render it probable that there may be some xelation between the pass .
of light and of heat through a crystal .
The recent determinations of the refractive indices of metals by have shown that they stand in the same order as conductors of heat , as to the velocity of propagation of light through them , an this fact brings again into prominence the old determinations with respect to crystals .
That the comparison which KUNDT has made for the metals cannot be carl.ied to other bodies is at once seen from the fact that the index of refraction of iron differs little from those of glass and several commoner crystals , the condnctivities of which shown to be very small compared to that of iron .
A comparison may , however , be possible among transparent bodies themselves , and the following experiments were made with the object of furnishing data for this comparison , the results by previous observel .
S differing greatly from each other .
They have , owever , been extended to embrace non-transparent bodies commonly in use in a tory , and about the conductivity of which we have had a very meagre or absolutely no knowledge .
*SENABMONT , ' Ann. de Chim .
et de Phys ( 3 ) , vols .
21 , 22 , and 23 ( 1848 ) ; V0N LANG , ' Pogg .
Ann vol. 135 ( 1868 ) ; JANNETAZ , ' Ann. de Chim .
et de Phys ( 4 ) , vol. 29 ( 1873 ) .
Outline of Method .
The most important consideration in determining the method to be used is the fact that it is difTicult to get large pieces of the crystals to be experimented on .
This excludes methods requiring large plates , such as that of WEBER and TUSCHMI or large spheres or cubes , such as that of KIRCHHOFF or THOMSON .
A method which to several advantages was the one first suggested by LODGE , and which may be called the " " divided bar\ldquo ; method , and after some preliminary experiments had been made to its suitability , it was finally adopted .
It consists in observing the temperature a bar heated at one end and cooled at the other , and diyided halfway between the two ends by a plane perpendicular to its axis , when ( 1 ) the divided ends are together , ( 2 ) a disc of the crystal or other body is between .
If temperature observations are taken at several points in each half of the bar , the corrections to be applied to the second set of observations for the distances of the points of observation from the contacts can be determined from the first set , and thus the temperature at each side of the crystal disc can be found .
If , in addition , we know the thermal conductivity of the bar used , the amount of heat flowing through the disc can be found from the temperature slope ; and thence we have the thermal conductivity of the crystal .
It may be noticed that there would be considerable uncertainty as to the nature of the conta , cts between bar and crystal unless some special precautions were taken .
LODGE proposed to use pads of tinfoil to obviate this difficulty , but this method is not satisfactory .
The difficulty has , however , been completely overcome by using bars of a material which would amalgamate , and making the contacts by means of mercury .
Description of .
( Fig. 1 .
) The bar used in the experiments was one brass , which presents several advantages .
( 1 ) Tt readily amalgamates , and therefore enables good contacts to be made .
( 2 ) Its conductivity is not so high as to make comparison between it and that of crystals , \amp ; c. , impossible .
( 3 ) Its conductivity , according to LORENZ , increases with the temperature , a fact which , as will be seen later , partly neutralises the deviation of the cooling from NEWTON'S law .
The diameter of the bar was cm .
, and its length 67 cm .
To each end of the bar was soldered a can , through one of these cans steam was sent and through the other water .
The whole was arranged in a wooden frame , so that the bar was horizontal and exposed to the air over a covered tank through which a current of water could be sent to keep the temperature constant .
The apparatus was surrounded by screens to protect the bar as far as possible from air currents in the room .
* ' Phil. Mag ) , vol. 5 , p. 110 ( 1878 ) .
OF CRYSTALS AND OTHER BAD CONDUCTORS .
The cans at extremities of the bar were supported on slides attached to the framework in such a way that the bar could be moved parallel to itself by screws passing through the framework and bearing in conical indentations in the backs of the cans in the prolongation of the axis of the bar .
These were sufficient while the bar was entire , and the experiments to determine its conductivity were being made .
After the bar was cut for the insertion of the discs , it ecame necessary to support the cut ends in such a way as ( 1 ) to enable the motion of the bar parallel to its axis still to take to enable the cut faces to be adjusted parallel to each other ; ( 3 ) to cause as small as possible a disturbamce of the isothermal surfaces .
These were secured by supporting each free end in a loop of thin string , Fig. 1 .
hanging the upper part of the framework .
ends of this loop were ttached to screws for the purpose of raising or lowering it as required .
The sides of the loop made an angle of or with each other at the bar .
These loops were sufficient to support bars properly if the arrangement was kept horizontal .
, was , however , found that the mercury contacts could be made with much greater certainty if the bars were vertical and the reury surfaces horizontal .
On this account two other loops were attached to the lower part of the framework and passed over the bars .
These were elastic , and served merely to keep the upperloops taut , and the in the same positions with respect to each other , whether the framework was horizontal , as in the experiments , or vertica ] , as in making the mercury contacts .
Measuremen of The most direct method of measuring the temperature along the bar is to sink thermometers into holes in the bar , but this method , although it has been used by FORBES , TAIT , and MITCHELL for their large bars , is very objectionable , as it diminishes the available of flow , and therefore makes the change of temperature along the bar more rapid than it would be for a contin uous bar .
The method first MR. C. H. LEES ON THE THERMAL CONDUCTIVITIES lsed by and FRANZ of measuring the temperature by means of a thermojunction ) into with the bar is free this ob.jection , and will produce no effect on the distribution of temperature in the bar if the mass of the junction is not used by WIrDEMANN and had a silvered surface , and it was sufficient to bring the thermo-couple into contact with the surface in order to get the temperature .
In the later of the following experiments the bars were painted , so that contact could not be made in this way .
Small conical holes about5 mm. deep were , however , made at opposite extremities of horizontal diameters of the bars , and at nces along .
Eight such diameters were taken , the two nearest the centre of the uncut being 1 cm .
apart , and the rest .
apart .
Each of these small was amalgamated , and sufficient mercury left in them to make good contact with the ends of the two wires which were used as a thermo-couple .
The cuit woulrl thus consist at the bars of\mdash ; first wire of couple ; mercury of first hole ; brass of bar ; mercury of at opposite extremity of diameter ; second wire of couple .
The mercury in the holes was constantly cleaned by touching the surface with dilute nitric acid , washing and drying ; and the ends of the thermocouple wires were kept bright .
Under these conditions the arrangement worked satisfactorily .
The wires used originally for the thermo-couple were of brass and iron , but these gave unsatisfactory readings , mainly on account of the poor contact between iron and mercury .
As there is an in using copper as oue of the elements , a few combinations of copper with other metals and alloys were tried , and eventually the platinum-silver alloy used for resistance coils was selected as the other element .
This gives an E.M.F. of aoout volt for difference of temperature , and its constant only increases about 7 per cent. for C. It is , therefore , well suited for thermo-electric measurement of temperature .
The wires used for the thermo-couple were about 20 .
long and 2 mm. diameter .
The ends which seryed to make contact with the galvanometer circuit dipped into mercury cups in blocks of wood which could be moved along guides parallel to the bars in order that the whole of the temperature observations could be made with one couple .
The cup into which the platinum-silver wire dipped was provided with a thermometer graduated in C. The galva ometer was a low resistance one of the WIEDEMANN typ , with SIEMENS ' and copper damping sphel.e .
The resistance of the whole circuit was only ohm , and as the resistances of the contacts at the holes in the bar might it was necessary to take measurements of the resistance of the circuit when a galvanometer deflection was taken .
For this purpose modification of the bridge method was used .
The diagram ( fig. 2 ) shows the arrangement of the circuit .
is the alyanometer in series with the tbermo-e ] ement .
are three mercury cups , arranged so that when and are connected , the galvanometer and thermo-element are in circuit alone , and when and are connected the galvanoOF CRYSTALS AND BAD CONDUCTORS .
48.5 meter element form one side of the bridge .
is a , and a key .
and are each ] 00 ohms from post-office box ; is a multiple arc box capable of ad.justment to ohm .
The resistance of the galvanometer circuit is found by so that there is no eflect on the nometer on pressing down , hrst with in one direction then the other .
The coil is introduced in the battery circuit to counteract the effect of the current through the galvanometer while making these observations .
will be noticed that the current in it is reversed by the operation which reverses the current in the galvanometer coils , and that the galvanometer needle is therefore kept close to the zero of the scale during the whole of the observations .
Let the values of the resistance determined as above be and , and let be the resistance of each of the arms , AB , that of the coil the .
of the battery , that of the thermo-element .
Then writing down the condition that and have the same potential in each case , we have Hence which gives The last term of the right member of this equation never exceeds , and may therefore be neglected , and we The constants of the thermo-couple were determined by comparison with thermometer reading to degree Centigrade .
An auxiliary iron bar of 80 and cms .
square crross-section , was placed horizontal heated at one end and cooled at the other by the waste and water respectively from the principal .
The upper surface of this auxiliary bar was tinned and amalgamated .
On this surface two copper mercury cups of 1 cm .
diameter and cm .
height were placed .
The under .
C. H. ON THE BMAL C of these so as bo make good contact with the iron bar .
The insides of were malgamated , and each cup contained sufficient mercury to fill it heIl t thermometer was placed in it .
At opposite extremities of a diameter , about half way down each cup , two small holes like those in the divided were punched , and into these holes , which were amalgamated , the wires of the could be .
The constants of the couple are in this way deter .
lnined under ctly the same conditions as when they are in use .
The iron bar was in wdust , and the copper mercury cups placed in such positions on the upper the rmolneter read in them C. and C. respectively .
the insertion of the wires of the thermo-couple into the holes of the mercurv cups , the cups surrounded with sawdust , and minutes allowed for the tem- perature of the cups to get as nearly as possible uniform observations were taken .
In eriment made in the early part of the work the following observations were taken :\mdash ; peratnre of hot junction , C. Telnpcrature of coldjunction , Deflection of galvanometer in scale divisions , 296 .
Resistance of circuit , mean of observations taken , as previously described ohm .
E.M.F. of element in units used , ( b ) Temperature of hot junction , Temperature of coldjunction , Deflection of galvanometer , Resistance of circuit , ohms . .
of element in units used , These observations are sufficient to determine the constants in the formula where snd are the temperatures of the hot and cold junctions respectively .
It was found more convenient for the reduction of obseryed E.M.F. to emperature , to draw up a small table of values of for diffel.ent values of .
Thus , the above observations , we have : OF CRYSTALS AND OTHER BAD CONDUCT O By inspection , we have the divisor for .
E. , F. to teln perature difference , since in eyery case is known ) tely to one two degrees .
The value of is seen from bove table to only about 7 per cent. for C. of ctJange of element is therefore well suited for thermo-electric observations of temperature .
The determination of he constants of the eIement was ) every few days during the progress of the work .
The of was found to slowly ; the total change during the three months during which the couple used about per cent. The temperature of the enclosure in which the divided bar was suspended was determined by means of a thermometer graduated in degrees , aced about 7 above the top of the water-tank and about 15 .
below the bar .
The bulb of thermometer was protected from radiation from the hot by a small sheet of just large enough to screen the bulb .
Determination of the of the The thermal conductivity of the bars was determined , the chod of FORBES , by two series of experiments , the firsb to determine the law connecting the rate of loss of heat from a heated surface , with the of that , and the second to determine the law of steady distribution of temperature bar heated at one end , and exposed to the air ] its entire length .
These two observations furnish data for determining the thermal conductivity required .
Cooling It is customary in making these experiments to use a short length of the .
used in the steady experiments .
I have , however , shown that the law of cooling is the same for bars of different material , but of the same cross-section and ) as a knowledge of the specific heat of the bar is required , it seemed better to make use of a bar of a material the specific heat of which is well known , rather than a brass bar in these experiments .
As the values found by different obseryers for the specific heat of copper are practically identical , this metal was chosen .
As it is better conductor than brass , it also enables the assumption temperature throughout the bar is the same to be made without much chance of error .
The of the same diameter as the brass bar , cm .
cms . .
It had polished nickel-plated surface .
To make the cooling experiment correspond more closely to of an infinitely long bar , the ends of the bar were covered with a layer of sulphur about 5 mms .
thick .
This was done by placing the bar vertical , ) pping a sheet of paper ound the upper end so a to project about 1 .
above the end , and then melted sulphur into the paper cylinder .
Since ] phur is an extremely poor conductor of heat , it may be that the loss of heat from the ends of the .
is very small .
The bar suspended in the position occupied afterwards by the divided bar by meitns of .
The temperature was determined by means of the platinum-silvercopper inserted into small amalgamated and filled with mercury , on opposite sides of a Lneter at the middle of the .
Observation of , resistance , tenlperature of the cool junction , and of the enclosure , were made every two minutes during the experiments .
From these observations the temperature of the bar at any is determined , and curves drawn for each experiment showing the connection between time and temperature of bar and of enclosure .
Of these curves that which most nearly with the mean of all was selected , and is referred to in what follows .
If is the mass of the bar , its snrface , the specific heat of the material of the bar , the excess of the temperature of the bar over that of the enclosure , which will supposed constant , we have , as the connection between and time , ( 1 ) , where is a which is to be constant ( NEWTON'S law ) , but which I have shown increases for a nickel-plated bar of the size used 70 or 80 per cent. for C. of In the } ) aper referred to this rise is expressed by writing where n is small tion about .
In that paper was always positive .
As in the following work is sometitnes negative , in which case would require special interpretation , it is preferable to write where is a small quantity .
The foregoing equation becomes , then , ( 2 ) , the solution of which is .
( 8 ) , , here A is an constant .
The table shows that this equation represents the cooling with a amount of accuracy for .
The temperature of he enclosure bas been assumed constallt C. , but in reality it varied from at the commencement to at utes , and then down to at 40 minutes , at 60 minutes , ab 70 minutes .
On this account no attempt has been made to determine with greater accnracy ; the value given above is determined by trial .
* Phil. Mag , p. 429 The heat lost from a nickel-plated surface exposed to air is , therefore , glam degrees per .
cm .
per second , where is the excess of temperature of the surface over that of the air .
Experiments .
These experiments were made with the brass : before it was cut .
One end wag by steam , and the other cooled by water .
Eight observations of temperature were taken at different points of the bar in the way previously described .
Let be the excess of temperature at point of bar over that of surrounding air , perimeter of cross-section of bar , of cross-section of bar , conductivity at temperature excess Then , at point of the surface of the bar , we have where is the rate of change of along the normal to the surface , , and having the meanings assigned to them in last section .
At C. this gives Now will be seen later to have value 27 approximately .
Hellae , at a point of surface , at the temperature of C. excess , At the same point on the ) .
Hence , the inclinaticn of the normal to the isothermal surface at this point to the axis of the bar is arctan . .
If the isothermal surface be assumed to be part of a sphere , the radius of curvature is about 42 .
At C. excess , a similar calculation gives the radius of curvature of the isothermal surface to be 60 Hence we may assume the isothermal surfaces to be planes perpendicular to the axis of the bar .
The equation for the motion of heat in the bar is under these conditions .
( 4 The conductivity has generally been considered constant in treating equation , but this is scarcely justifiable , as most experimenters find changes OF CRYSTALS AND OTHER BAD CONDUCTORS .
per cent. in conductivity for 10 C. change of temperature .
Taking therefore as equal to the above equation becomes .
( 5 ) 1 Multiplying through by 2 , this becomes or , From which by integration we obtain constant .
If the value of when be , the above equation becomes .
( 6 ) If , the integral of this equation can be expressed in terms of logarithmic and circular functions .
If , the integration introduces circular functions and elliptic integrals of the 1st and 3rd kind .
If the problem under discussion were \mdash ; given the conductivities\mdash ; to determine the distribution of temperature throughout the bar , this integration would be necessary , but as we are given the distribution and have to find the conductivity , the problem can be solved without further integration .
Although this method is probably not as accurate as that depending on the integrated equation , its accuracy is sufficient for the present purpose , where the value of the conductivity is only required over a small range of temperature .
To determine at the points of observation is represented by an empirical function of and the differential coefficient with respect to taken .
The conditions which satisfies , fix , to some extent , the function to be used ; we see , e.g. , that it vanishes for some value of , and that at that point the first differential coefficient is finite ; that it increases in one direction with in an approximately exponential manner .
These considerations lead at once to the function A , and this function I has been used .
It eviden cannot express accurately throughout the bar , since it is the solution of a linear differential equation , but it ( jan be made by proper choice of the constants to represent the main feature of the curve , the nces between its values and those being afterwards represented by an additional : which has generally only small values .
For the curve which correspollded best with the mean of the three sets of observations taken , value of is represented the cooler part of the bar by sinh .
The di[ferences , with the exception of the last two , are within the limits of error of observation , and 1nay be neglected .
The last two may be taken into account by adding to the above expression for , between and , an expression of the where The values of the coefficient of this expression at are , 128 , 294 respectively .
Differentiating the hyperbolic expression for we have sinh .cosh .
, which gives , on substituting for the coordinates of the points of observation , the required values of , to which , at and , we must add the quantities and 294 to get the true temperature slope .
The values thus obtained are given the above table .
Having thus got the values of and at eight points on the bar , we can by trial determine so that the equation ( 6 ) holds .
A few trials show that is positive and equal approximately to 002 .
It is not necessary to determine with any great accuracy it Hects very little the subsequent work .
Assuming the of results:\mdash ; OF CRYSTALS ANI ) OTHER BAD CONDUCTORS .
Column ( 5 ) gives the values of , and column gives values of ' which by equation ( 6 ) two numbers in column ( 6 ) , which are uncertain on account of the nallness of the quantity ( 5 ) at the cool end of the bar , we have , as the mean value of , 00141 .
Hence The temperature of the air in this experiment was C. hence for the thermal conductivity of the br ss bar used* .
ser. Experiments on Before cutting the bar the insertion of the discs the conductiyities of were to be measured , obseryations were of the tion of ture along the bar after it had been inted with Aspinall 's enamel .
It is necessary to have the diating s of and discs the same , and this is most easily secured by painting both .
loss of heat from a point of the at temperature is about 40 per cent. greater than the loss at the same temperature for the nickel-plated , but does not rapidly with rise of ' ture .
On this account it is possible to express the distribution of tempel a along the bar in the form , , and be determine ed so that this equation holds with closer ) tion t in the of the unpainted bar , throughout the whole of the bar .
It is assumed to hold in what follows for each half of the divided aving the , but different values for each half .
The bar was divided in the middle , between the contact hoIes 1 .
the ends ground down to be as nearly as possible planes perpendicular to the of the bar .
To secure this a vertical hole , a little larger than bar , drilled in a pristn *The following values of brass have been found by different imentcrs : , ' de Chim .
et de Phys. ' ( III .
) , vol. 66 ( 1862 ) , 302 ; , ' .
Berlin Akad for1880 , p. 457 , ; tIlNZ , 'Wied .
Anu vol. ( 1881 ) ; ) of wood , .
base , and 15 .
height .
This was done in a drilling nlachine the table of which was found to be perpendicular to the spindle .
One of the then fixed by into this hole in such a way that a point , which when the bar was in position would be the highest point of the cut surface , came in contact with a murk the block of wood .
The block and bar were then moved about on a stone afterwards on a slate slab , till the surface was plane and smooth .
The other bar then fixed so that the vest point of the cut surface came into contact with the same mark , and the surface rubbed down .
This process secured that the cut should be parallel to each other when the bars were coaxal .
To secure and test that the surfaces were parallel when the bars were in position in the framework , two test gauges were made , one for testing the upper surface and the other the side of the bars .
The bars were taken out of the frame and placed vertical , the upper bar resting with its cut surface accurately on the cut surface of the lower bar , and not being supported by any other means .
Four screws on each gauge were then adjusted so that they would touch four points , two on each bar near its extremities .
By applying these gauges to the bars in position , either in contact with each other or with a disc between , it is therefore easy to make the required adjustment of the cut surfaces .
The surfaces were now amalgamated , the amalgamated surfaces brought together , and experiments made on the distribution of temperature throughout the bar , to test the efficiency of the mercury contact .
When the bar was uncut , the difference of temperature observed between the points in the middle , cm .
apart , was , as a mean of three experiments , C. With the cut bar and the mercury contact , the distribution of temperature in the other parts of the bars being the same , the difference of temperature observed C. The cutting and grinding of the bar had reduced the distance between the points of observation to cm .
, hence if the contact has no ppreciable resistance the fall should be \mdash ; , which is , within the limits of error , what was observed .
Hence , as far as the distribution of temperature along the bar is concerned , the mercury contact secures practical continuity .
Occasional tests were made during the course of the work to see if this continued to hold .
The of the contact was found to increase slightly .
Three tests made at the end of the work as the difference of temperature .
This is equivalent to an increase of distance between the points of observatiou from cm .
to cm .
, and is probably caused by the gradual soaking of the into the brass .
Account is taken of this in the following calculations .
The conditions of the above experiments are not quite the same as in experiments with crystals , \amp ; c. , for while in the above there is only one mercury film , in the crystal experiments there are two .
To investigate the effect of this , a thin film of mica , about 0003 cm .
thick , was inserted .
Although there are then two films of mercury a film of mica between the bars , no change could be detected in the OF CRYSTALS AND BAD DUCTORS .
encs of temperature of the .
points .
The same hcld on insertion of a piece of platinum foil0025 cm .
thick .
The corrections for the length of bar and mercury contact between the points at which the temperatures are observed the surfaces of the discs , can , therefore , be determined from the observation with the gamated ends of the bars in contact .
In preparing the apparatus for an experiment , the bars and framework placed so that the bars were vertical , with the cool .
at the top .
The were taken out of the frame and the holes and contacts cleaned with dilute nitric proyided with mercury , washed , and then dried by filter paper .
The hot bar was then inserted into lower part of the frame , and if a disc was to be experimented on , it was placed on the ercury surface and the excess of mercury forced from under it .
In the case of a transparent disc it was possible to see that the mercury formed a perfect mirror .
The cool bar was then put into the upper part of the , and held up , in contact with the centre screw at the top of the frame , by means of a spring .
The amalgamated end would be thus brought about cm .
above the upper surface of the disc .
It was then supplied with mercury , which hung down as a pendent drop .
The upper screw was then turned till the end of the upper bar was about mm. from the upper surface of the disc .
In this position the gauges were applied to the bars , and the loops so adjusted by means of the screws at their ends , tbnt the four points of each gauge touched the bars .
The upper screw was then rotated till the upper bar came into contact with the , and the excess of mercury was reed out .
The frame was then placed over the water tank with the bars horizontal and connections made to the steam and water supplies .
The apparatus allowed to stand about one and a half hours , in order that the distribution of temperature might become steady before observations were taken .
Almost invariably on taking the bars apart at the conclusion of an experiment , the discs were found to adhere to the cool bar , and in the case of transparent discs this enabled it to be seen if the contact between the cool bar and the disc had been good .
It was generally found as a mirror as the contact between the hot bar and the disc .
Some experiments were vever made to try the effect of making contacts which were poor as far as could be judged by the appearance of the mirrors .
No difFerence could be detected in the observations between an optically good and a bad mirror , so that a few experiments , hicI ) the contacts were optical]y defective , have been included in the results given .
Reduction of with Discs .
We have seen that the temperature throughout any short length of the bar be represented by the empirical equation .
We assume that this equation holds for the first three observations on each side of the disc , and thus have , if are the of the points of observation , measured from the surface of the disc , and the observed temperatures , , R. H. RMAL These three equations are sufficient to fix the values of the constants , but their solution is difficult unless .
This relation is very nearly satisfied by the points of obseryation , in the hot bar , cms .
; in the cool bar , The error introduced by assuming each of these intervals .
is small enough to be neglected in the cool bar , but a small correction is necessary in the of the hot bar .
If be the value of at the point we have for the temperature at the point Now is found to differ little from 1 for any of the experiments , we may , therefore , with sufficient accu.acy , take Making use of we have from the three atures , where \amp ; c. This equation determines from any three observations of temperature .
The mean of the vahJes thus determined for different points along the bars and for different experiments is used in the subsequent work .
We have then the equations *The value of may be mined as .
By Theorem we have\mdash ; Therefore Pntting we h , since for any short length of the ) . .
Thereforc .
valuo of for , point of is in all experiments approximately 04 .
CRYSTALS AND BAD CONDUCTORS .
where are known quantities , to determine the values of A and B. Differentiating the equation nd putting , we have at the surface in contact with the disc , The isothermal surfaces in the iscs themselves may be assumed to be planes , for a small calculation , like that made ( p. 490 ) for the bar itself , shows that the radius of curvature of these surfaces is about 40 cms .
iting for the coordinates of the surfaces of the the temperatures deternlined as above , the values of in the bars at the surfaces in COlJtact with the disc , being the conductivity of bars , and that of the disc , we have , for the temperature in the disc itself : Differentiating and writing down the expressions for the flow of heat into and out of the disc we have the equations\mdash ; and where refer to the disc and have the usual meanings , being in most cases , differing only slightly in others .
Now , the thickness of the is small enough to make small .
Hence , writing - .
, and expanding the hyperbolic functions , we have , as a close approximation-- Therefore similarly ; either of which equations deterlnines k , the conductivity of the disc , if its approximate value is substituted in the terms involving .
The value of used is that for a painbed surface , at the mean temperature of the disc , and the value of is given with sufficient accuracy for the above substitution by the equatiop Observations .
* 1 TRANSPARENT ISOTROPIC BODIES .
\mdash ; The temperatures given are the means of three experiments .
CROWN Glass Disc , .
diam. , 169 cm .
thick .
C. Glass Disc , cm .
diam. , 177 cm .
thick .
Air , C. mean ROCK Salt Disc , cm .
diam. , 83 cm .
thick .
Air , C. *The results } colleoted in a table on page 606 .
OF CRYSTALS AND OTHER BAD CONDUCTORS .
II .
CRYSTALS.\mdash ; Temperatures , mean of three experiments .
QUARTZ Disc , cm .
diam. , cm .
thick .
Thermal stream lines parallel to optical axis .
Air , C. QUARTZ Disc , cm .
diam. , 811 cm .
thick .
Thermal stream lines perpendicular to optical axis .
Air , C. mean ICELAND Spar Disc , cm .
diam. , 70 cm .
thick .
Thermal stream lines parallel to optical axis .
Air , C. ICELAND Spar Disc , cm .
diam. , 602 cm .
thick .
Thermal stream lines perpendicular to optical axis .
Air , C. MlCA , large plate , .
thick .
Thermal stream lines perpendicular to cleavage planes .
Air , ) C. llCtu As the lllica was in the form of a large plate , the above method of finding is not strictly correct , this accounts for the diffe , rence between the calculated values of and .
It found impossible to use a disc , as in the other experiments , on account of the ) itting .
The value of given must be taken as approximate only .
III .
ROCK SPECIMENS .
MARBLE Disc , cm .
diam. , 26 cm .
thick .
The marble was the white variety used for statues .
Air , C. mean SLATE Disc , cm .
diam. , 31 cm .
, cut from an ordinary school slate . .
mean OF CRYSTALS AND OTHER BAD CONDUCTORS .
Where no diameter is given the plate is of ular shape , and rather larger than the section of the bars .
SHELLAC Plate , 068 cm .
thick .
Air , C. This plate was made between two microscope slides having plane surfaces .
One slide was placed horizontal , and the upper surface covered with a smooth sheet of tinfoil .
On the foil was laid a sufficient number of the thin films of shellac suppliod by varnishmen to make a plate of the requisite size and thickness .
The slide was then warmed gently from underneath till the shellac was soft , and the cold slide brought down on to the soft mass .
In order to get the requisite thickness , and the surfaces of the plate parallel to each other , the two slides were kept apart by two short lengths of wire cm .
diameter laid across the lower slide near its ends , the upper slide being pressed against these .
The tinfoil which attaches itself to the plate is readily dissolved off by mercury .
This method is very convenient and gives excellent plates .
mean PARAFFIN Plate , 038 cm .
thick .
Air , C. This plate was made by dropping a little melted paraffin on to the cold upper surface of the lower slide , and placing the upper slide on it described above .
The cold surfaces are sufficient to ensure the paraffin being deta , chable from the glass .
Cool bar .
1378 1378 1378 1378 1378 1378 1378 1378 1378 1378 1378 Hob bar .
1407 mean PARA RUBBER , pure .
Sheet0264 cm .
thick .
Air , C. SULPHUR Plate , 0584 cm .
thick .
Air , C. This plate was made in the same way as the plate of shellac , except that it was found better to a sheet of tinfoil on the surface of each slide .
mean EBONITE Disc , cm .
diam. , 0414 cm .
thick .
Air , C. mean * Kindly supplied to m oy Mr. C. H. GRAY , of the Silvertown Company OF CRYSTALS AND BAD CONDUCTORS .
GUTTA PERCHA Sheet , 0617 cm .
thick .
Air , C. This sheet was made of the material used for insulating wire .
A small was cut from a wire , melted , and pressed into a sheet between the two cold microscope slides .
Cool bar .
mean PAPER Disc , cm .
diam. , 019 cm .
thick .
Air , C. This was cut from an ordinary visiting card .
Cool bar .
\mdash ; 1515 Hot bar .
mean " " ASBESTOS Paper\ldquo ; Disc , cm .
diam. , 047 cm .
thick .
Air , C. This was the ordinary asbestos millboard of commerce .
It consists of paper to the pulp of which sufficient asbestos has been added to render it incombuQtible . .
C. H. ON RMAL c Disc , 1.93 cm .
dialn .
, 056 cm .
thick , cub from a sheet of " " veneer\ldquo ; used by kers .
Air , C. mean WALNUT Disc , ; cm .
diam. , 069 cm .
thick , cut from a sheet of veneer .
Air , ] C. Cool bar .
Hot mean CORK Disc , cm .
diam. , 05 cm .
thick , cut by a section-cutting machine from a sound cork .
Air , C. Sheet , 0084 cm .
thick .
This was cut from a dress piece of plauin brown silk .
Air , 1 .
C. CRYSTALS AND OTHER BAD CONDUCTORS .
SILK Sheet , 017 cm .
thick , cut from a piece of ribbed silk ribbon .
The thickness is the mean of that across ribs and across grooves .
Air , 1 C. mean COTTON Sheet , 085 cm .
thick .
A piece of thick tape .
Air , C. mean Sheet , 1 cm .
thick .
Air , C. mcan As the silk , cotton , and flannel yield ements of are not very accurate in these three cases , The error will be tcst in the cases of flannel and cotton , and probably small jn the case of silk .
; MR. C. H. LEES ON THE THERMAL CONDUCTIVITIES TABLE of Results .
' .
of cr . . . . . . . . . .
Rock salt . . . . .
lartz t to axis . . .
Iccland slltr axis .
perpendicular to axis Mica pcndicular to cleavage marble . . . . . . . . . .
er oil . . . . .
raffin .
cnnised .
Ilph bonitc Gut ) . . . . . .
across .
cross fibre Cork Silk . . . . . . . . .
Thermal co C. G. S. units .
0016 . . . . . . . . . . .
0158 . . . . .
0084 \ldquo ; ivity between 2C .
Refl a 00046 PECLET OF CRYSTALS AND OTHER BAD CONDUCTORS .
References : \mdash ; LORENZ , ' WIEDEMANN'S Anmalen , ' vol. 13 ( 1881 ) .
F. WIlBEiR , ' Monatsber .
der .
Akad 1880 , p. F. NLUMANN , ' Annalcs de Chim .
et do Phys III scr .
, } ol .
6 .
, ' ' vol. 123 , p. ( 1861 ) .
H. MEyER , ' nIlalcn , ' vol. , p. 600 ( 188S ) .
PECLET , ' nales de Chim .
et de ' III ser. , vol. 2 .
G. FORBES , ' Roy .
Soc. .
Edin vol. 9 , p. 64(1873 ) .
TUSCHMIDT , 'Beibliitter , ' 1884 , p. 49t ) .
YAlfAGAwA , ' Journal , College of Tokio , ' vol. 2 .
p. 263 ( 1888 ) .
WINKELMANN , ' POGGENDORFF'S AnnaleIl , ' vol. 153 , p. 481 ( 1874 ) .
In the foregoing the conductiyities of a few metals and liquids are given in order to show the position of the bodies experimented on amongst other conductors .
Where experiments have been preyiously by other experimenters their results are given for comparison .
Such comparison shows a fair degree of concordance between the present results and those obtained by totally methods , with the exception of the results of G. FORBES , which were obtained by the following method .
A can about 9 cIns .
diameter was filled with a freezing mixture , placed on a sheet of the material the conductivity of which was to be deternlined , the lower surface of the sheet brought down on to a surface of water at C. Ice is formed on the surface of the sheet , and FORBES works out an expression for the thickl]ess of ice in terms of the temperature of the freezing mixture ( which was obseryed by a tl)ermometer placed in it ) , the thickness of the material , the time , and the conductivities of ice and th material .
In working out this expression , however , he assumes the conductivity of the materials between the thermometer in the freezing mixture and the upper surface of the sheet to be infinitely great , which , when one considers the layers of liquid , metal , and air present , is scarcely justifiable .
On this account almost all FORBES'S results are low .
PECLET'S results for metals were all found be low , and this led to the assumption that all his results were the same .
It is , however , evident from the present experiments that his results for bad conductors are correct .
This is probably to the fact that the layers of still water which , in his experimen{s , adhered to the faces of the sheet experimented on , are of much less importance when the sheet is a bad conductor .
TUSCHMIDT'S results are obtained by a method which WEBER used for liquids .
The sheet of liquid is simply replaced by one of the crystal , the contacts being made by means of glycerine .
As glycerine has a conductivity is about of those of the crystal plates , the glycerine layers have a great effect on the flow of heat through the plates .
This is a defect of the method , and , in addition , the plates themselves must be large , and are therefore expensive .
TUSCHMIDT'S results and those of T the present paper agree fairly , the only great difference being in the case of Iceland spar the axis .
MEYER'S determinations depend on the rate of rise of the temperature of a calorieter into which a heated cube of bhe material is plunged .
It is questionable whether which he Inakes , that the temperature of the surface of the cube at any instant is that given by the thermometer in the calorimeter , is justifiable on account of the difficulty of preventing a layer of water adhering to the surface .
The objection applies to the method which he used to confirm the results obtained by his first method .
Remarks .
From the Table of Results it is at once evident that , for transparent bodies , no such comparison between thermal conductivity and velocity of propagation of light can be made , as has been made by KUNDT for the metals .
The thermal conductivities vary enormously for very small change of refractive index , and the variations are sometimes in the same direction as the variation of the index of refraction , sometimes in opposite .
It is remarkable that the bodies quartz and rock salt , which are diathermanous bodies , should also be good thermal conductors , quartz being a better conductor than bismuth .
To prove that the high value of the conductivity is not due to heat passing through these bodies by radiation from the hot bar to the cold , several experiments were made with the apparatus arranged as it was during the experiment in quartz , but the qnartz disc was removed , thus allowing the hot bar to radiate heat through the intervening air space to the cool bar .
Under these conditions , no change of temperature at the point of observation near the end of the cool bar , could be detected when the hot bar was suddenly cooled .
As air is a more diathermanous body than quartz , it is thus evident that the amount of heat radiated through the quartz from one bar to the other , is too small to affect the above results .
The high conductivity ofquartz would render the use offused quartz ( the conductivity of which does Jlot probably differ much from that of quartz crystal ) advantageous for vessels subject to sudden change of temperature , and for delicate thermometers .
The results for Iceland spar and marble seem to indicate that the irregular arrangement of the crystals in marble interferes with the of heat .
Water and salt solutions have conductivities about equal to that of glass .
The solid insulations used in electrical work rank with glycerine and the oils , so far as thermal conductivity is concerned .
Silk has about one-third the conductivity of shellac .
Hence , if a silk covered wire the covering saturated with shellac , it will have its heat conducted away to surrounding bodies much faster than previously .
The practice ofsoaking galvanometer coils in shellac is , tnerefore , good from a thermal point of view , as it enables the heat generated in the wire to pass more rapidly to the exterior of the coils , and be radiated OF CRYSTALS AND OTHER BAD CONDUCTORS .
On account of the low conductivity of paraffin , standard resistance coils with temperature coefficients ought not to be embedded in paraffin , as the B.A. standard is .
Cork is the worst conductor experimented on .
Almost all the substances experimented on are insulators electrically .
Although tables of their electrical conductivities are available , the values given seem to depend more on the condition of the surface of the body experimented on tlJan on its material , and , on this account , it is ctle use to make any comparisons between the electrical and the thermal conductivities of these substances .
The above experiments were carried on in the Phvsical Laboratory of the Owens College .

