]\gt ; [ 353 ] XII .
Thermal Properties of A at Low ( Second Paper .
) C. FREWEN NKIN , .
Inst. C. , Professor of Sciencc , and D. R. , ilI .
A. , ) of College , Oxford .
icated bSir EWING , , F.R.S. [ PLATE 6 .
] Fueccived , \mdash ; Rcad hIay 1 , 1915 .
CONTENTS .
Page Introduction . . . . .
354 The use of throttling experiments . . . . . . .
355 Description of the experiments\mdash ; 1 .
Measurement of the total hcat of . . .
356 2 .
Measurement of the total heat of liquid . . .
360 3 .
Throttling experiments on superheated gas . . .
362 Construction of the chart . . . . . .
364 Appendices\mdash ; I. A method of plotting essure cves in the superheated area of the chart . . . . .
369 II .
A method of plotting I lines in the superheated area of the chart . . . . . .
369 III .
Modifications produced in the chart by a correction in the values accepted for the total heat of the liquid ) . . . .
370 IV .
A method of plotting the results of throttling experiments . . . . .
371 .
A method of plotting gas-limit curve from the resulcs of a live of throttling experimertbs . . . . .
37 VI .
A method of correcting the results of a throttling experiment for a sm departure from dard conditions . . . . .
373 VII .
Allowance for the small drop of pressure in the iIlch calorimeter when used to measure the total of the gas . . . . . .
List of papers referred to . . . . . .
375 Tables\mdash ; A. Observations on the total heat of gas . . . . . . .
376 B. Mean spccific of ) gas at const prcssure for C. . . . . .
377 C. Spccific hcats of at . . . . .
D. Observations on the totnl of . . . . .
378 .
Throttling expel.iments . .
379 F. Por plotting throttling experiments . . . .
381 VOL. CCXV . .
C. JENKIN AND .
D. R. ON THE INTRODUCT ION .
IN a form ) the authors described a series of measurements on the properties of the construction of a chart embodying those results .
It was pointed out thattlle superheated area of the chart was incomplete , the constant pressure lines being only approximately accurate and the total heat ( I ) lines omitted .
It was also pointed out that throttling experiments on the supel.heated gas would form a valuable check on the accuracy of the chart , and the construction of chart was postponed till more accurate ements should have been made on the gas and the whole had been checked by rottling experiments .
present papel additional experiments required to complete and check the a re-measurement of the total heat of the liquid for which some had been used before , and finally the construction of the chart .
This chart been constructed graphically , as the chart was , directly from the ) data and its accuracy checked in various ways by thermodynamic equations .
These equations apply quite generally to ] charts and are independent of the particulal properties of ) onic acid .
The value of the chart in all calculations connected with refrigeration was originally pointed out by MOLLIrR , and has recently been emphasized in the Report of the Research Committee of the Institution of Mechanical Engineers ( 1914 ) .
The authors hope that the chart may be of practical use , as it extends and corrects the original o1te prepared by The nents w made , as before , in the Engineering Laboratory at Oxford .
The apparatus nployed was generally silnilar to that used in the former experiments , but all heat measurelne]lts were made in a new calorimeter , fig. 1 , which eliminates radiation and conduction losses .
The measurements of the total heat of the gas for each pressure were made at several temperatures , so that the variation of specific heat with temperature was determined .
The values of the specific heat found by these experiments were used to plot the new constant-pressure lines on the chart .
The method used for plotting these lines is explained in Appendix I. The re-measurement of the total heat of liquid showed , as was expected , that the values previously used required correction ; the liquid-limit of the chart was corrected accordingly .
The throttle experiments , which provide an independent check on both the gas-limit curve and constant-pressure curves , the accuracy of the pressure curves and of the limit curve above - C. , but showed that a small correction was required below that temperature .
This correction having been made , the chart was completed by drawing the I lines in the superheated area .
( See Appendix II .
, fig. 6 .
) The combined corrections of the two limit curves leave the values of the latent heat practically ] ; the ration a C. is only per cent. THERMAL PROPERTIES OF BONIC ACID AT LOW In a experiment gas flows through a throttle valve while pressures and temperatures are measured 011 each side of the .
No other measurements are required .
It is convenient to make a number of experiments starting from the same initial conditions , expanding the gas to a series of lower pressures .
Such a group of experiments may be called a of experiments , since it corresponds to expansion along one I line on the chart .
It is convenient to makcl a number of such lines of experiments all starting from the same initial pressul.e but from successively higher temperatures .
The series of final pressures to the gas expands must be the same in all the lines of experiments , if the method described below is to be used the results .
Throttling experiments cannot , without some other data , be used to plot any part of the chart , but with a single constant-pressure curve they can be used to plot the other pressure curves in the superheated area , and also the -limit curve .
resulting are not affected appreciably by errors in the total heat of the liquid ( see Appsndix III and are quite independent of errors in the latent heat , so that throttling experiments form an independent check on the accuracy of the gas-limit curve and all the constant-pressure curves .
A method of plotting the throttling so as to check the constant-pressure curves is explained in Appendix .
By this method a series of constant-pressure curves is drawn which are to be compared with the pressure curves already obtained from the specific heat measurements .
If the curves coincide , the throttle experiments confirlu tccuracy of the ; if the curves are ] but not coincident , the throttling experiments confirm the accuracy of the specific heat measurements , but indicate that there is an error in the position of the gas-limit curve , which may be shifted so as to make them cid .
If they are not parallel , there is a disagreement between the throttlin experiments and the specific heat measurements and one or other must be wrong .
Since shifting the gas-limit curve will make the curves , if parallel , coincide , a new gas-limit curve may be plotted derived directly from the throttle experiments\mdash ; a method of doing this is described in Appendix V. ; whether the new gas-limit curve is more reliable than the original depends o11 the reliability of the experiments .
In order to use the methods we have adopted the throttling experiments must be arranged exactly in , and the series of lower pressures must be exactly the same in all the lines , as already stated .
When actually making the experiment it is hardly possible to adjust the initial pressure and temperature and the final pressure exactly to the standard values , so that the observed results have to be reduced to standard conditions .
A method of making reduction is in Appendix VI .
PROF .
C. FREWEN JENKIN AND MR. D. R. ON OF THE IMENTS .
of the l'otal Heat of Gas .
The method employed in these experiments was the same as that used in the former series ( Series III .
, p. 73 ) , but the apparatus was modified*in several details .
The gas used was , as before , supplied by Messrs. Barrett and Elers , Limited ; the same instruments were used for weighing the gas , measuring its pressure and temperature , and the electric heat given to it .
The instruments were re-calibrated from time to time as a check , in the same manner as before .
The weighing flasks , condenser and drying flasks were the same as before , but calcium chloride was used in the drying flask instead of phosphorus pentoxide , which occasionally gave trouble .
The single-stage pump was replaced by a larger two-stage pump lent for the purpose by Messrs. J. and E. Hall , of Dartford .
With this pump the rate of working could be more easily controlled , and the attainment of low pressures was facilitated .
The large calorimeter ( No. I. ) was discarded , and the evaporation and varming of the gas was done in calorimeter No. II .
with the addition of a simple heating vessel made of a length of -inch steel pipe enclosing an electric resistance .
SCALE Fig. 1 .
All measurements of heat were made in a new 1-inch tubular calorimeter , fig. 1 .
This calorimeter consists of a central body with a thermojunction fitting at each end .
The central body is made up of three concentric pipes held between two vulcanite ends ; the outer pipe is made of steel , the intermsdiate pipe of vulcanite , and the inner pipe of fused silica .
Inside the inner pipe is a coil of nickel-chrome wire forming the heating resistance , Folded round the outside of the silica tube is a thin strip of copper which serves as the return lead for the current flowing through the heating coil .
The whole is held together by two external bolts .
The gas enters at one end , passes through the silica tube , and then returns outside the silica tube and inside the vulcanite one , and finally returns again outside the vulcanite and inside the steel tube .
At each end are the two thermojunction fittings ; they are the same fittings as were used with the throttle valve .
THERMAL PROPERTIES OF CARBONIC ACID AT LOW RATURFS .
It will be seen that the temperature of the entering is measured jnst before it enters the body of the calorimeter , and the nperabure of the issuing gas just after it leaves the body of the calorimeter ; neither case the mperatnre of the gas in the calorimeter be affected by conduction the pipes .
Most of the expel.iments were made with the issuing gas at atmospheric temperature , so that the outer steel shell was at atmospheric temperature and could neither rive out nor ] ueat .
make this doubly sure the calorimeter was wrapped in cotton wool and inside a 1nuff filled with water kept at approxilnately the temperature as the shell of the calorimeter .
The only remaining path by which heat can enter the neter is through the left-hand yulcanite end ; the amount flowing in by path must be extremely small .
There is a Hoskins alloy ocouple clamped to outside of the steel tube and connected to a very sensitive alvanolneter which gives a deflection of 20 mm. per degree change of temperature of the junction .
The electric power was regulated during an experiment so as to keep indicated by the anometer as steady as possible .
A few experiments were made with the issuing gas at C. The muff was then heated to about the same temperature , but it is possible that a little heat leaks out at the ends of the muff , so that these experiments may be slightly wrong .
The passages through this calorimeter are rather small , so that there was an appreciable drop of pressure in the gas as it passed ) rough it .
The diffel.eIlce of pressul.e between the inlet and outlet was measured by means of a differential pressure gauge , the construction of which is shown in fig. 2 .
An inner glass tube dips into mercury contained in the outer steel tube .
The difference of level of the mercury inside and outside the glass tube , which is proportional to the difference of pressure is measured by the change of resistance of a wire stretched the centre of glass tube .
Several sorts of wire were tried ; some made uncertain contact with the mercury , but a bright hard steel banjo wire answered very well .
The resistance of the SCALE INCHES wire was measured with a simple Callendar and Fig. 2 .
Griffith bridge .
The gauge was calibrated by direct comparison with a glass -gauge containing lnercury .
arrangement of the valves enabled the pressure to be applied and released quickly , so that the zero and readings could be repeated several times as a check on their accuracy .
PIiOF .
C. FREWEN NKIN AND MR. D. R. ON THE The general arrangement the apparatus is shown in fig. 3 .
The liquid from the weighing flask expands the throttle valve and is evaporated in the heater .
From there it passes in the condition of nearly dry vapour into No. II . .
where the evaporation is completed and the gas is warmed to the desired temperature .
It passes into the 1-inch calorimeter where its temperature is raised through an accurately measured range by the application of an accurately measured electric power .
From this calorimeter it passes on through regulating valves and a drying flask to the pump , which compresses it into the condenser and so back to weighing flasks .
The object of the -inch heater was to facilitate the regulation of the tempsrature of the gas .
By adding most of the heat necessary for evaporation in the -inch Fig. 3 .
heater , the temperature of calorimeter No. II .
could be kept only a few degrees above the required gas temperature , so it acted as a steadying reservoir of heat .
Without the -inch heater the arrangement was unstable , for then the calorimeter No. II .
had to be kept much warmer in order to transfer sufficient heat to the vapour .
The point where the vapour became dry was then near the top of the coil , and the least variation in the rate of flow made this point move up or down and so left a varying amount of surface for the gas , the temperature of which , therefore , varied widely .
For experiments at low pressures , , 150 , 200 , 300 and 400 lbs. per square inch , the speed of flow was controlled by the valve while and were open .
For the experiment at was also partially closed .
For the experiments at 600 and was open and was partially closed .
For the 700-1b .
experiment the weighing flask had to be slightly warmed by a gas ring .
During an experiment the valve was kept open and closed , so that the standard THERMAL PROPERTIES OF BONIC ACID AT pressure read the gas pressure , and the differential pressure gaugc was in equilibrium , which voided any nger of the mercury out of it .
Just before the end of each experiment the valvG was closed and opened and readings taken on the differential gauge .
The quantities measured nent w The rate of flow of the Its initial pressure the drop of pressure in calorimeter .
The temperatures of Iltering e the calorimetel The electlic power supplied to neler .
following quantities also noted:\mdash ; The temperature of the muff and of atmosphere .
The anometer r The results are given in able ( p. 376 ) , arranged in the sallle way as in of the former paper .
The correction for the small drops of pressure in the calorimeter is explained in .
The results are plotted in fig. 4 .
PROF .
C. JENKIN AND MR. D. R. ON THE of the specific heat of the gas for C. intervals have been measured from the curves and are given in Table B. The pressure curves in the chart were plotted with these values , by the method explained in Appendix I. , after the curve been corrected by the throttle experiments .
of the specific heat for each pressure at temperatures varying by C. interyals from .
to C. , as measured from the curves , are given in Table C. of the Total Hcat of Iiquid The inlents dibed in the former paper on the total heat of the liquid were not completely satisfactory for two reasons .
The lowest temperature ched w C. , so extrapolated values had to be used for drawing chart C. and C. made from C. ( the two lowest ) , both considerable corrections for\ldquo ; fall of bath temperature\ldquo ; and\ldquo ; radiation amounting ether to per cent. and per cent. respectively of the measured heat .
Subsequent experience has shown that the correction for the " " fall of bath temperature\ldquo ; is liable to error .
It was assumed to be equal to the -equivalent of the multiplied by the fall of erature during the experiment .
But the of the water-equivalent of a composite body such as the calorimeter , consisting of a good conductor surrounded by an insulator , depends on the quantity of heat by the , and this depends on the of temperature ; as this may be very different in the actual experiment from it was when the water-equivalent was measured , the value assumed for the water-equivalent may differ considerably from the yalue .
The effect of conduction along the pipes appears also to be somewhat uncertain ; the correction for this was included in the radiation correction .
These considerations made it desirable to repeat the measurements of the total heat more accurately .
The method employed for this purpose was the same as before , proved apparatus was used .
By using the new 1-inch calorimeter described above , radiation and conduction losses and the correction for the change of temperature of the bath were eliminated , and the introduction of a heat THERMAL PROPERTIES OF CARBONIC ACID AT LOW TEMPERATURES .
361 interchanger enabled the temperature of the liquid to be carried down to C. , eleven degrees lower than.before .
is made of two concentric copper feet long , coiled into a ring -inch diameter .
It is the same piece of apparatus which is described in the former paper on p. 92 .
By means of this Fig. 6 .
interchanger the liquid was cooled before passing the throttle valve by the cold gas leaving imeter II .
, an arrangement which considerably increases the available refrigeration ; it was only used for the last experiment at the lowest temperature .
The arrangement of the apparatus ( including the interchanger ) is shown in fig. 5 .
VOL. CCXV.\mdash ; A. 3 The liquid from the weighing flask passes through one coil in calorimeter then through the 1-inch calorimeter , then through the inner coil of the interchanger and expands through the valve V into the second coil in calorimeter II .
; from there it passes through the outer coil of the interchanger , through the drying flask to the pump .
The liquid is cooled to the desired starting temperature in calorimeter II .
It is then warmed through an accurately measured range of temperature in the 1-inch calorimeter by the accurately measured electric power .
The warm liquid is cooled in the interchanger and then , by expanding through the valve into the second coil , cook the calorimeter II .
, thus supplying the refrigeration required to cool the liquid in the first coil .
The cold gas passing out of calorimeter Ir .
serves finally to cool the liquid in the interchanger and then passes to the pump .
The quantities measured were : The rate of flow of the liquid The rise of temperature of the liquid in the 1-inch calorimeter .
The electric power supplied to the 1-inch calorimeter .
The results of the experiments are given in Table , arranged in the same way as in Table , of the former paper .
The new results have been used to correct the liquid-limit curve on the chart ( see fig. 6 ) , where the full line is the new and the dotted line the old curve .
The maximum correction , at C. , amounts to carnots and decreases to zero at C. A correction of the limit curve produces a very small correction on the I lines , as is explained in Appendix III .
The I lines have been shifted to the right at small a distance to be shown in the figure .
Throttling Experiments on Gas .
Throttling experiments are simple and quickly made .
The four quantities to be measured are the pressures and temperatures before and after the throttle valve .
The rate of flow is immaterial ; it is only necessary to adjust the apparatus till the conditions are steady and then to take the four readings .
The throttle valve used in the new experiments was the same one that was used for the experiments on throttling the liquid described in the former paper ( Series IV .
, p. 91 ) .
The arrangement of the apparatus is shown in fig. 7 .
The liquid coming from the weighing flasks expands through the valve into the -inch heater where it is evaporated and then passed into calorimeter II .
, where it is warmed to the required initial temperature .
It then passes through the special throttle valve where the pressure falls to the required lower pressure , and so through the drying flask to the pump .
The valve serves to regulate the lower pressure when necessary .
The results are given in Table where each line of experiment is denoted by a letter ( column 1 ) .
Columns 2 and 3 give the imtial and final pressures .
THERMAL PROPERTIES OF CARBONIC ACID AT LOW TEMPERATURES .
363 Columns 4 , .5 , 6 and 7 the observed gauge pressures and corresponding absolute pressures .
Column 8 gives the observed initial temperature .
Column 9 gives the same temperature ( see Appendix VI Column 10 gives the initial temperature .
Column 11 gives the observed final temperature .
Column 12 gives the same temperature ( Appendix VI Column 13 gives which is the observed final temperature ( Appendix VI The lines of throttle experiments do not all extend over the same range of pressures .
Lines and extend from 600 to .
( misses two lines .
) extends from 500-200 .
, , , , 400-200 .
, , , , 300-200 .
and were limited at their upper ends by the maximum temperature to which the experiments could be carried .
Lines and V start at 600 but Fig. 7 .
observations were only taken on the 200 and 150 curves .
They were made to link on the 150 curve , which had been omitted from the earlier experiments .
Line extended from 700 to 600 and 500 .
This was the only line taken up to 700-1b .
pressure .
These results were plotted on the chart in the way explained in Appendix IV .
The results for the 700 , 600 , 500 and 400 pressure curves were found to coincide with the curves already plotted , and the results for the 300 , 200 and 150 pressure curves were parallel to the curves already plotted , thus confirming the values of the specific heats used to plot these curves , but indicating that the limit curve required a small shift at pressures below 400 ; the limit curve was accordingly redrawn below the 400 curve in the way explained in Appendix V. ( see fig. 6 ) , in which the full line represents the new curve and the dotted line the old curve .
The maximum shift , at C. is carnots .
The three pressure curves , 300 , 200 and 150 were then redrawn to start on the new limit curve , and the th'ottle experiments for these curves replotted .
The calculations , made as explained in Appendix IV .
, for this second plottin are set out in Table , which gives the co-ordinates and , for the points on each line .
co-ordinates are arranged in columns ; each column contains the results for one cnrve , value of the pressure being stated at the top of the column .
The second line in each column the value of I on the limit curve for that pressure .
The 400-pressure curve was assumed to be correct in plotting the results of all lines which had readings on that curve , i.e. , and R. The line was plotted from the 300-pressure curve and , from the 200 curve .
was plotted from the 500 lrve .
The results , with the three bottom curves replotted in this way , should all fall on the pressure lines : agreement is remarkably good , the errors being too small to show in .
The alteration in the two limit curves which are shown to be necessary by the new experiments leave the width of the diagram between the limit curves practically unaltered , so that the values previous]y obtained for the latent heat are hardly changed .
The reasons for accepting the modification of the bottom of the gas-limit curve are as follows : the extreme simplicity of the throttling experiments make errors improb- able .
The close agreement between all the specific heat experiments and all the throttle experiments over wide ranges of temperature makes any systematic error in the throttle experiments improbable .
The close confirmation of the accuracy of the upper part of the limit curve also supports the accuracy of the throttle experiment .
The temperatures reached in the throttle experiments were lower than those given by the uncorrected chart .
Had there been errors in the throttle experiments due to conduction or radiation the difference would have been in the other direction .
Construction of the Chart .
The chart ( Plate 6 ) is plotted on skew co-ordinates , the angle between the axes being arc-tan ( degrees ) .
The vertical I scale is 1 inch Th. U. ; the horizontal scale is 1 inch carnots .
The whole chart is divided into a skew graticule , the distance between horizontal lines being 1 Th. U. inch ) , and the horizontal distance between the sloping lines being carnots ( 2 inch ) .
The part of the chart from\mdash ; 50o C. C. ( and the super-heated area up C. for pressures below 800 ) represents the results of the authors ' experiments .
The rest of the chart has completed by using MOLLIER'S ( 2 ) limit curve ( approximately ) and AMAGAT'S ( 3 ) results for the relation between pressure , volume and temperature .
These data alone are not sufficient\mdash ; some data on heat quantities are necessary ; we have therefore used experiments on the specific heat at constant volume , which cover a considerable ran ge ( between the lines and on the chart ) , and for THERMAL PROPERTIES OF CARBONIC ACID AT LOW ATURES .
365 the rest of the area we have assumed a value for the specific heat at constant volume , viz. , .
This assumption is probably close to the truth , since this value , found by JoLY for the line is exactly equal to the value deduced from our own experiments for the line A. MOLLILR assumed was constant and equal to for the whole of the area outside the limit curve .
It would appear at first that most of the chart be constructed by simply replotting the chart on the new co-ordinates , and our first chart was plotted in this way , but when it was checked it was found to be inaccurate .
Investigation showed that the errors were due to three causes ; the difficulty of reading the I values accurately enough from the chart ; the insufficient accuracy of table of AMAGAT'S results which we had nsed ; and lastly a small error in the limit curve which was greatly magnified when transferred to the new co-ordinates , where the scale is about five times as large as in the chart .
gives AMAGAT'S results in a very convenient form ; it is accurate for most purposes , but not for plotting the volume lines of the chart , particularly near the critical point .
We found it necessary to plot AMAGAT'S results and interpolate graphically to obtain the required data with sufficient accuracy .
The errors in the first chartled us to devise a number of ways of checking it , many of which turned out to be most useful in reconstructing it .
For this purpose the following equations connecting the total heat I with the other variables , and useful .
( 1 ) The slope of any constant-pressure curve in an chart is equal to the temperature , or . . . . . . . . . .
( i ) ( 2 ) The slope of any constant-temperature curve in an chart is equal to the absolute temperature minus , or .
( ii ) Cor. 1 .
At the critical.point the limit curve , the constant-temperature curve , and the constant-pressure curve are mutually tangent to one another , and their slope is equal to the critical temperature .
There is a point of inflexion in the constanttemperature curve , and the curvature of the constant-pressure curve is zero at this point .
Cor. 2 .
Constant-pressure curves do not change their direction on crossing either limit curve .
Cor. 3 .
There is no point of inflexion in any constant-pressure curve ; curves of the shape shown by MOLLIER ( 7 ) for the 80 and 90 curves in the chart are therefore impossible .
Cor. 4 .
The slopes of all the constant-pressure curves are the same at the points where they are cut by any one temperature curve .
Thus every constant-pressure curve on the chart has the same slope at the point where it is cut by the C. temperature curve , viz. , 308 .
Cor. 5 .
The slope of the constant-temperature curves is a convenient measure of the " " imperfectness\ldquo ; of the gas , for the slope is equal to the difference between the reciprocals of the dilatation of a perfect gas and of the gas represented , i.e. , slope ( 3 ) ( iii ) ( 4 ) The slope of any I curve in a chart may be expressed in any of the folowing forms : . . . . . .
' ( iv ) ( ivc ) Equations ( iv ) and ( ivc ) are particularly useful in the saturated area , where they reduce . . . . .
( ivd ) and . . . . .
of which the first is the more convenient .
Cor. 1 .
Equation ( ivd ) may be written:\mdash ; where is the dryness fraction .
Thus the slope of I lines varies with across the chart from one limit curve to the other .
Cor. 2 .
The forms , ( ivb ) and ( ivc ) become indeterminate at the critical point , but ( ivd ) or ( iv ) may be used .
For the slope of the I line at the critical point is about 3730 .
THERMAL PROPERTIES OF CARBONIC ACID AT LOW TEMPERATURES .
367 Cor. 3 .
The horizontal distance between two constant-temperature lines in the saturated area of the chart which differ in temperature by is The distance may be found for any value of by usin the corresponding value of V ( depending on the dryness ) .
This expression was used in plotting the chart to find the distance of the C. line from the C. line and the distance of the critical point from the C. line .
Using the above equations checks a second chart was drawn , plotting from the original data as far as possible instead of copying from the chart .
During construction it was constantly checked by the above formulae and small corrections made ; in this way we finally arrived at a result which represents all the experimental data and also complies with the above general theorems .
The only part of the diagram which does not fit in quite satisfactorily is the short piece of the liquid-limit curve between C. C. In order to make sure of the remarkable form of the temperature curves on the left of the diagram , these curves were extended some distance further than they are shown .
There was no difficulty in this , since AMAGAT'S results go to much higher pressures , but as the pressures increased we found it impossible to make the slopes of the pressure curves comply with equation ( i ) .
This discrepancy appears to point to some variation in the specific heat at constant volume in this region .
As there are no data available we have omitted the curves beyond , up to which pressure the discrepancy is very slight .
If the chart be examined it will be seen that experiments and ours overlap and may therefore be compared .
JoLY measured and we measured ; to compare the results it is necessary to obtain an equation connecting the two .
The most convenient equation is . . . . .
( i ) a form which may be derived directly from the fundamental thermodynamic relations , and is quite general .
The familiar relation for a perfect gas\mdash ; does not give even approximate results .
Comparing results with ours by equation ( i ) and using ANIAGAT'S figures to obtain the differential coefficients , we found that they agreed within the limits of accuracy of experiments .
The first lines drawn on the chart were the 700-1b .
line from zero down to C. and the line from zero up C. where it meets the limit curve .
The starting for these curves are the points for 700-1b .
line .
, , 900-1 .
, , The limit curve from C. C. was then set off from these curves ; it sses the zero temperature point where .
The sloping lines corresponding to pressures in the saturated area from .
to were then drawn from the limit curve at slopes equal to their temperatures .
The gas-limit curve was then determined by marking points on each of these pressure lines at values of taken from the chart .
The apex of the limit curve was then plotted by means of equation ( iv ) Cor. ( 3 ) .
The gas-pressure and temperature curves up to C. were then plotted from the throttle experiments ; and starting from the C. line four constant-volume lines and were plotted , ten degrees at a time , by means of the expressions and using values for .
Several more constantvolume lines were then drawn between and C. in the liquid area , assuming that values of were in all cases calculated from AMAGAT'S data .
points on all these volume lines were then marked where they are cut by the constant-pressure using AMAGAT'S data .
The constant-temperature curves for C. intervals were then drawn through the points already marked on the volume lines .
The constant-pressure curves were then drawn by drawing an envelope of tangents passing through the points already marked on the volume lines .
Any error in the chart was shown up to this stage by the envelope of tangents missing the marked points on some of the volume lines .
This check is a very rigorous one .
An error of thermal unit in the position of the liquid-limit curve will throw a pressure inch away from the point it has to pass through .
data do not go below zero , so a different procedure was necessary for the liquid area below zero .
Constant-temperature lines were drawn C. C. ; the first is horizontal because C. as is shown in fig. 9 of the former paper ; the secolld slopes upwards at an angle which can be calculated from the formula throttle drop*per using the data given in Table VIII .
of the former paper , extrapolated C. The throttle drop per 100 lb. is given in fig. 9 of the former paper , where it is called the JouleThomson effect .
To farc 1 ) .
368 .
THERMAL PROPERTIES OF CARBONIC ACID AT LOW MPERATURES .
369 The spacing of the pressure curyes on the C. temperature line is given by the expression the values of lvere calculated from fig. 11 of the former paper .
The spacing of the pressure lines on the C. temperature line was assumed to be uniform , which must be closo to the truth as may be seen from fig. 11 of the former paper .
The pressure curves were extended the C. voluIYle l through the points marked at C. ping them in form to the 700-1b .
line which was drawn from experimental ta .
The lower ends of the high-pressure curves have , however , been omitted from the final chart , because the data for plotting them were considered to be hardly sufFiciently accurate .
Additional constant-temperature curves were interpolated using the above formula to give the slope , those above C. being slightly curved to match those at higher temperatures whose form was known .
These complete the chart .
APPENDIX I. A Method of Plotting -pressn .
re Curves in Superheated of Each pressure curve starts from a point on the gas-limit curve whose temperature is given by the pressure-temperature curve for saturated vapour .
Above this point it is plotted in C. steps .
For each vertical step of C. the horizontal step , , is given by the equation:\mdash ; in which is the mean temperature of the step , C. , and is the mean specific heat of the gas for the step ( see Table B ) .
NDIX II .
A JIethod of Plotting I Lines in the Area of Chart .
The I lines have already been drawn in the saturated area for values differing by 5 , i.e. , for , 5 , 10 , 15 , \amp ; c. The last of these is .
It is required to plot the lines , 65 , \amp ; c. , which fall the superheated area .
The line is plotted by finding the points where it crosses the pressure curves .
Let ALP be any VOL. CCXV.\mdash ; A. 3 pressure curve , cutting the line in , and the limit curve in L. We have to find the point on this curve where With the notation in the figure Again and also Therefore Having found , the point on the pressure curve is marked off Similar points are found on the other pressure lines and joined up , thus the required I curve .
Points on all the curves after the first are plotted , each from the one before , by means of the equation APPENDIX III .
Produced in the Diagram by Correction in the Accepted for the Heat of the Liquid If the line AB represents the constant-pressure curve , .
per sq .
inch , then the area MABN under this curve down to absolute zero represents the heat required to raise the temperature of 1-lb .
of at 700-1b .
pressure from to , i. e. , the total heat from to .
If the total heat were assumed to be 5 per cent. less , then A would have to be shifted to , where AA ' is 5 per cent. of , i.e. , about carnots at .
The limit curve , which is set offfrom the constant-pressure THERMAL PROPERTIES OF CARBONIC ACID AT LOW PERATURES .
371 curve , would be shifted almost exactly the same distance to the right , pivoting round the point which has the same I as on the pressure curve .
The constant I lines would hardly be shifted .
They are set out from points on the limit curve in the way explained on p. 80 of the former paper , and calculation shows that they would only be shifted about one-tenth as far as the limit curve is shifted .
us on C. line would only be moved about carnots to the right .
APPENDIX IV .
A Method of Plotting the Results of hrottling Eiments .
In the following method the results of the ling experiments are used to draw all the constant-pressure curves but one ; the position of that one is assumed to be known .
The pressure curves drawn in this way should coincide , if the chart is correct , with the curves drawn by means of the observed specific heat of the gas .
Let be the pressure curve , assumed to be correctly drawn , on which represent the starting points of a number of experiments throttling down to a second pressure .
Let A be the starting point of the experiment at the lowest temperature , and let the observed final temperature , when throttling down to , be .
Draw a horizontal line at this temperature .
It is required to find the point on this line representing the final condition .
We have ( i ) and , where is the specific heat of the gas at , and , on the limit curve , is known .
Let be the point on the limit curve corresponding to and let be the horizontal distance between and B. Then an equation giving and thus .
B. Choosing the next experiment , starting at , we find the point by equating and so on .
Joining up the points \amp ; c. , we have the constant-pressure curve corresponding to the pressure derived from the throttling experiments .
A similar construction is used to plot the throttling results corresponding to all the other pressures .
If the chart is correct the curve will coincide with , the constant-pressure curve plotted through by means of observed specific heat of the gas .
APPENDIX V. A Method of Plotting the Gas-Limit Curve from thoe Results of Line Throttling Experiments .
The limit curve is plotted by finding the points where it cuts successive pressure curves .
Let and represent the initial and final conditions of the gas in one of the lines of throttle experiments .
COA and DPB are the pressure THERMAL PROPERTIES OF CARBONIC ACID AT LOW TEMPERATURES .
373 curves and through A and , OP is the limit curve , and CD the last I curve in the saturation area .
We assume that the points and A are known and we require to find where the limit curve cuts .
Draw a horizontal line DP at .
Then will be fixed when is found .
Let and be the difference of total heat between and B. Then ( i ) But therefore ( ii ) Similarly , selecting other pressure lines , we find a series of points which all lie on the limit curve .
The same method is applicable for finding points on pressure curves above , if is not on the highest pressure curve .
For checking our chart the point was chosen on the 400-1b .
pressure curve .
APPENDIX VI .
A Method of Correcting the Resutts of a Throttling Experiment for a Small Departure from Conditions .
Let be the standard starting point and the standard final pressure .
and are the actual starting point and finishing point .
and lie on an I line which cuts in and in S. The I line through A cuts the line in U. We have to find the temperature of , i.e. , the temperature which would have been observed if the conditions had been standard .
With the notation in the figure Similarly , and The required temperature of The values of were measured from the chart and plotted against temperatures , and it was found that for the range covered by the experiments the values were roughly constant for any one temperature independent pressure ) .
They may , therefore , be given with sufficient accuracy as follows:\mdash ; THERMAL PROPERTIES OF CARBONIC ACID AT LOW TEMPERATURES .
375 APPENDIX VII .
Allowance for the Small Drop of Pressure in the 1-inch when ?
to the Heat of the Gas .
Let be the point representing the condition of the gas entering the calorimeter , and the point representing the gas the lorimeter .
PR is the constantpressure curve meeting the I curve through .
Through draw a horizontal cutting the pressure line in S. Let the drop of pressure in the calorimeter be The amount of heat given to the gas If there had been no drop of pressure in the calorimeter , this amount of heat would have raised the gas to instead of .
The actual heat given is , but .
to raise the temperature to ( or ) , if there were no throttling , we should have needed only The heat may therefore be corrected by subtracting .
The " " throttle correction\ldquo ; ( Table A ) is calculated in this way ; it is in all cases very small .
PAPERS , \amp ; c. , REFERRED To .
Reference .
( 1 ) " " Thermal properties of carbonic acid at low temperatures\ldquo ; by JENKIN and PIE , ' Phil. Trans. , Roy .
Soc Series A , vol. 213 , p. 67 ( 1914 ) .
( 2 ) MOLLIER , ' Zeit .
fur die gesamte Kalte-Industrie , ' Heft 4 and 5 , pp. and 85 , 1895 .
( 3 ) AnfAGAT , ' Annals de Chemie , ' tome 29 , 6th Series , 1893 .
( 4 ) JoLy , ' Phil. Trans. ' A , 1894 , p. 943 .
( 5 ) , ' Zeit .
fur die gesamte dustrie , ' Heft 4 , p. 65 ( 1896 ) .
( 6 ) EWING , " " The Mech. Production of Cold 1908 .
The reference is to p. 195 .
( 7 ) MoLLIIiiR , 'Zeit .
des Vereins Deutsch .
Ingenieure , ' 1904 .
THERMAL PROPERTIES OF CARBONIC ACID AT LOW TEMPERATURES .
377 TABLE Specific Heats of for C. Intervals .
( Measured from the curves , fig. 4 .
) Notc.\mdash ; This table is used for plotting constant-pressure curves , see Appendix I. VOL. CCXV.\mdash ; A. TABL C.\mdash ; Specific Heats of Gas at Constant Pressure .
( Measured from the curves , fig. 4 .
) Temperature .
C. Temperature .
C. Pressure , lbs. per sq .
inch .
TABLE D.\mdash ; Observations on the Total Heat of Liquid IFEDAHWeight o THERMAL PROPERTIES OF CARBONIC ACID AT LOW TEMPERATUPRS .
379 .
THERIsIAL PROPERTIES OF CARBONIC ACID AT LOW RATURFS .
381

