189 The Boiling-point of Sulphur on the Thermodynamic Scale .
By 1ST .
Eumorfopoulos , B.Se .
, Fellow of University College , London .
( Communicated by Alfred W. Porter , F.R.S. Received December 20 , 1913 , \#151 ; Read January 29 , 1914 .
) The experiments described in this paper were carried out with the constant-pressure gas thermometer .
The form of the apparatus has been already given in some detail in a previous paper , * the results of which were not regarded as final owing to the alteration in volume of the glass bulb .
, The latter was , therefore , changed for one of fused silica , and , of course , the compensation tubes were likewise of silica .
It is unnecessary to repeat all the details given in the previous communication , any change will be mentioned here ; it can be assumed that the mode of experiment has remained the same in all other particulars .
To avoid continual references back to this paper , figs. 1 and 2 are here repeated .
Fig. 1 represents the bulb and expansion vessel , while fig. 2 shows the two gauges ( the mercury and oil gauges ) ; it 9*468 lit !
Scale .
Fig. 1 .
* N. Eumorfopoulos , 4 Roy .
Soc. Proc. , ' A , vol. 81 , p. 339 ( 1908 ) .
VOL. XC.\#151 ; A. Mr. N. Eumorfopoulos .
The Boiling-point of should be added that the oil-gauge tap has since been changed for the tap shown in fig. 5 of the present paper .
The silica bulb and connecting tubes are drawn roughly to scale in fig. 3 .
Two silica bulbs have been used of practically the same volume ; the one has the form shown in the figure ; in the other , the bulb is connected immediately to the tube B , the capillary tube D being continued down to C. M ' M Fig. 2 .
2 4 6 8 iiit Fig. 3 .
The compensation tubes are similar to the above , except that the bulb is missing , the capillary tube passing directly across from B to D. The silica apparatus was made by the Silica Syndicate ; the capillary tubes were selected from a large number , so as to obtain proper compensation .
How far this has been successful will be seen from the following table:\#151 ; Bulb 1 .
Bulb 2 .
Compensation tubes .
Tube B ... . .
OT103 grm./ cm .
0*1048 grm./ cm .
0*1246 grm./ cm .
" D ... ... .
0*0266 " 0*0298 " 0*0176 " 0*1369 " 0*1346 " 0*1422 " ( All volumes are given in grammes of mercury at 0 ' C. ) Sulphur on the Thermodynamic Scale .
To join the end N of the silica tube to the rest of the apparatus , a little Prout 's glue was smeared over the silica , the latter gently warmed and pressed into position against the glass tube to the left of F ( fig. 1 ) , over this junction was slipped a glass sleeve a few centimetres long , leaving a clearance of about a millimetre , into which was poured cement .
This cement was made according to Faraday 's recipe , but a larger quantity of beeswax was added than is given in the directions .
Finally , the ends of the sleeve were covered with true Faraday cement .
It was , however , found necessary to introduce mercury into the capillary tube while carrying out the above operation , otherwise the w^ax dropped into the tube and completely closed it .
The tubes were then evacuated , and the cement at the two open ends partially melted , so that the wax could be sucked in , if the Prout 's glue did not hold properly .
Pressure Coefficient of the Silica Bulb .
This was measured with the following results .
With the bulb full of mercury , an increase of pressure of 1 cm .
of mercury produced an apparent change of volume :\#151 ; At 0 ' of 0*00124 grm. " 100 ' " 0*00126 " " 184 ' " 0*00124 " " 245 ' " 0*00125 " " 294 ' " 0*00128 " These numbers indicate that the coefficient does not vary appreciably with temperature .
Volume of the Bulbs .
Each bulb was heated for some hours in an air-bath to a temperature of approximately 550 ' C. before being used .
Volume of Bulb 1.\#151 ; 1401*735 .
Mean of 10 determinations , varying from 1401*728 to 1401*743 .
Volume of Bulb 2.\#151 ; Two rough determinations were made before use These gave 1385*980 and 1385*959 .
Immediately after the experiments on the S.B.P. two very careful measurements were made ( in July ) , viz.:\#151 ; 1385*9870 and 1385*9886 .
Three months later , two more , carried out with equal care , gave 1385*9835 and 1385*9833 .
The difference between the two last sets is small , but represents , I think , p 2 192 Mr. N. Eumorfopoulos .
The Boiling-point of a real change in volume .
In working out the results 1385*987 has been taken as the volume .
If the true value be 4 mgrm .
smaller , then the coefficient of expansion of the gas would have to be increased by I5 unit ( with six significant figures ) , and the value of S.B.P. by *0015 ' .
It may be mentioned that this bulb was heated on one occasion to 550 ' C. for seven hours in free communication with the air , and on the following day for another six ; it was then filled with mercury , and the freezing-point taken within an hour of being removed from the air-bath .
It was kept in ice continually for 24 hours ; no change of volume to a milligramme could be detected .
Coefficient of Expansion of the Bulb .
It was originally intended to determine the expansion of the bulb by the weight thermometer method .
In view , however , of the results obtained , * this idea had to be abandoned .
I have , therefore , made use of the following values derived from other observers:\#151 ; Between 0 ' and 100 ' C. , mean coefficient of expansion , 145 x 10"8 , " 0 ' " 445 ' C. " " 168 xlO-8 .
An increase of 10 x 10-8 in the first coefficient will increase the coefficient of expansion of the gas by 14 units with six significant figures , and diminish the S.B.P. by 0*017 ' , while a similar increase in the second coefficient will increase the S.B.P. by 0*032 ' .
The Barometers .
For description of these reference can be made to the earlier paper .
As the German glass used for the standard barometer was somewhat faulty , thus rendering the setting more difficult , it was replaced by English glass\gt ; which was found much more satisfactory .
The length of the barometer was determined before and after the experiments on the boiling point of sulphur by comparison with an invar scale recently redetermined at the National Physical Laboratory .
The values for the length of the barometer were as follows:\#151 ; Length at 16 ' .
April , 1911 ... ... ... . .
75*9776 October , 1913 ... ... ... 75*9770 * Eumorfopoulos , 'Phil .
Mag.,5 [ 6 ] , vol. 23 , p. 653 ( 1912 ) .
Sulphur on the Thermodynamic Seal .
Course of the Experiments .
As before , only three temperatures were observed , viz. , W.F.P. , W.B.P. and S.B.P. The ice was tested regularly every day with silver nitrate and with barium chloride solutions .
The former very seldom gave even an opalescence with the melted ice , the latter did so as a rule , but not as much as was given by a solution of sulphuric acid with freezing point \#151 ; 0*001 ' C. Only on one or two occasions was it necessary to reject the observations due to the precipitate exceeding this amount .
The S.B.P. apparatus is similar in construction to the one given in my former paper , except that it was of larger dimensions .
The part of the boiler exposed to the flame is 7 cm .
high , the portion covered by asbestos 32 cm .
, and the uncovered portion above this 5 cm .
The diameter is 11*5 cm .
, while the walls are just a little over 1 mm. thick .
A thin tin-plate screen surrounded , but did not actually touch , the bulb so as to prevent radiation .
The shape of the screen is indicated in fig. 3 by the dotted line .
To allow circulation of the vapour four holes about 1 cm .
diameter were punched in the screen at X. There were also openings at Y. In addition a horizontal plate was placed 2-3 cm .
above this screen ; this plate almost closed the opening of the bath .
A plate of uralite with the necessary openings for the four tubes was cut into two and placed over the mouth of the bath .
Into a fifth opening , nearly 2 cm .
in diameter , was luted a glass tube , in which the line of condensation of the vapour could be observed .
This occurred a couple of centimetres above the uralite plate .
To test how far the temperature depends on the size of the flame heating the bath , the gas was turned very low , so that the sulphur was just boiling .
The temperature became apparently steady .
The gas supply was now increased , and the temperature rose about 0*06 ' .
A further increase indicated no rise of temperature ( to 0*002 ' ) ; by this time the vapour was alight , and large flames about 10 cm .
long were shooting out horizontally between the top of the bath and the uralite .
It would thus appear that with this bath it is almost impossible to superheat the vapour . .
The Experiments .
The bulb was cleaned with nitric acid and with mixed chromic and sulphuric acids ; it was then thoroughly washed with distilled water , dried , and connected up with the rest of the apparatus by means of the wax joints already mentioned .
It was then heated for a few hours to 550 ' ( approximately ) , evacuated with a Fleuss pump , and finally with charcoal in liquid air , 194 Mr. N. Eumorfopoulos .
The Boiling-point oj Connection was made to the pump through the mercury gauge ( tube 0 , fig. 2 ) .
The method of connection is shown in fig. 4 .
Air was passed into the thermometer after purifying it by means of potash , sulphuric acid and phosphoric anhydride .
The bulb was several times evacuated , and finally filled with air in the cold at a pressure of 76 cm .
On determining several times the coefficient of expansion it was found to be 0*0036756 .
This value is evidently too high , so the bulb was evacuated again in the cold and refilled , no other change being made .
The coefficient was again found to be 0*0036756 .
As the error could be accounted for , if the compensation tubes were not acting , the junction of the silica and the glass at the wax joint was examined , but was found to be in order .
The volume of the bulb was roughly determined , but no appreciable change was found .
As it was thought that some impurity might come from the sulphuric acid or the phosphoric anhydride , the former was done away with , while the latter was changed , Kahlbaum 's being used this time .
The coefficient found was 0*0036755 .
The cause of these high values has so far remained a mystery .
If due to an impurity in the air , it seems extraordinary that the quantity should be so constant ( no burners were alight in the room at the time of filling , no other experiments were going on , and the window was open ) ; while , if the source of the trouble is something condensed on the silica , it seems equally remarkable that this impurity does not get permanently condensed on the glass bulbs kept at 0 ' C. ; on the contrary , the coefficient found on heating the silica from 0 ' to 100 ' agreed perfectly with that found on cooling down again .
It was now decided to try some experiments with nitrogen .
This gas was prepared in the usual way by heating together in solution potassium nitrite and ammonium chloride .
The resulting gas was passed through a tube containing solid potash sticks , another containing phosphoric anhydride , then through a red hot tube containing copper and copper oxide , then again over potash and phosphoric anhydride , and finally through a glass spiral ( 2SS I* Pump Sulphur on the Thermodynamic Scale .
195 immersed in liquid air .
This last tube was connected to the one .
first mentioned , and by a suitable arrangement the gas could be circulated .
This was done continuously for some three or four hours .
It should be mentioned that the whole system of tubes was evacuated before letting in the gas ; then some gas was let in , and the tubes again evacuated before proceeding with the final preparation .
After exhausting the bulb at a temperature of over 500 ' , with a charcoal vacuum , the nitrogen was introduced after passing it through a spiral immersed in liquid air .
The pressure being 76*0 , the coefficient was found to be 0*0036714 .
The bulb was again evacuated at 550 ' , left at this temperature for about two hours in connection with the charcoal tube , filled with nitrogen at the high temperature , and the gas drawn backwards and forwards for about half-an-hour through the spiral and over sodium-potassium alloy , the former tube being the nearer to the thermometer .
At a pressure of 76*3 cm .
, the coefficient was now 0*0036702 .
As the diminution in the coefficient is quite large , the bulb was again heated up , but not evacuated , the gas being merely passed to and fro through the spiral and over the alloy .
At a pressure of 78*4 cm .
the coefficient was numerically the same ; this means , in reality , a diminution of three in the last figure .
As this was deemed satisfactory , the following measurements of the coefficient were made :\#151 ; Up* ... 367074 xl0"8~l Down ... 367004 xl0"8 Down ... 367012 J Down ... 367034 Up ... 367031 1 !
Down ... 367016 Down ... 367022 J Down ... 367015 The mean of these , omitting the first value , is 367019 x 10"8 .
* The word " up " means a value obtained on heating up the bulb from 0 ' to the boiling-point ; the word " down " on cooling down .
Numbers bracketed together are obtained for the same boiling-point .
196 Mr. N. Eumorfopoulos .
The Boiling-point of Th.e following values for S.B.P. were obtained , reduced to 76 cm .
by Harker 's formula [ viz. , 0*904 ( j ?
\#151 ; 76)\#151 ; 0*00519 ( p\#151 ; 76)2]:\#151 ; .
Up 444'078~1 Mean ... 444117 Down ... 444-156 J Up ... 444-1701 444-170 Down ... 444-169J " Up ... 444-1491 444-146 Down ... 444-143 / " Up ... 444-1441 f \gt ; 5 . . .
444-138 Down ... 444-131J Up ... 444-131 " ... 444-131 Mean 444-140 Unfortunately at this stage the oil-gauge tap got jammed , and in freeing it a little air went up the tubes .
As it was July , and I was leaving town , I merely pumped out the nitrogen and let in some dry air at 76*3 cm .
pressure and determined its coefficient ; it was found to be 0*0036710 .
The air remained in the bulb till October , when the latter was heated up to the usual temperature for some hours , pumped out with a Fleuss pump and with charcoal in liquid air , some nitrogen let in over sodium-potassium alloy and through the spiral in liquid air , pumped out again , and again filled with nitrogen , the gas being drawn to and fro over the alloy .
Finally the bulb was cooled to 0 ' , the pressure adjusted , and the coefficient measured .
It was found to be 0*0036728 .
This coefficient is again too high , in spite of the fact that the heating up , etc. , occupied two successive days , say 12 hours in all .
The above process was , therefore , repeated , i.e. , bulb was heated up , and the same nitrogen drawn to and fro through the usual tubes .
The coefficient at a pressure of 75*6 cm .
was now found to be 0*0036704 .
The same process was again repeated for two days , and at a pressure of 75*7 cm .
the coefficient was found to be 0*0036701 .
As it was thought the gas was now sufficiently pure for measurements to be made , the following values for the coefficient were obtained:\#151 ; Sulphur on the Thermodynamic Scale .
uP ... 366999x 10-8^ Up 366985 x 10_81 Down ... , ... 366991 J Down 366978 J Up ... 366987 Up 366994 Up ... 366981 Up 366990 Up ... 3669791 Up 3669941 Down ... , ... 366990 J Down 366983J Up ... 367008 Up 367004 Up ... 366997 Mean 366991 x lO"8 .
And the following values for the S.B.P. :\#151 ; Up .
444-0841 444.106 Up 444-1181 444.120 Down .
444-129.1 Down 444-122 J Up .
4441011 444.092 Up 444-154 444-154 Down .
444-083 J Down 444-161 444-161 Up .
444135\444-134 Up 444-1511 444.138 Down .
444-133 J Down 444-124 J Up .
444134 U44-128 Up 444-1271 444.128 Down .
444-123 J Down 444-129 J Up .
444'13n444-120 Up 444-089l 444.097 Down .
444-107J Down 444-105 J Mean . .
444-125 ' .
The occasional somewhat large differences between the " up " and " down " values made me suspect that the trouble of the high coefficient had not been entirely eliminated ( it is obvious that this difference does not in any way depend on the observations at the S.B.P. ) , and I decided to repeat the whole set , substituting , however , a Topler pump for the Fleuss .
When the nitrogen was prepared , it was noticed that a considerable amount of ammonia was given off , and .
as it was feared that the purifying tubes might not be able to cope with this impurity , a fresh quantity of the gas was prepared , and passed first through a sulphuric acid bulb , and then through the same tubes as in the previous preparation .
Before the measurements with the new gas were carried out , I made , at Mr. J. Bose-Innes ' suggestion , a few observations at a lower pressure , so as to obtain the thermodynamic correction .
These experiments cannot claim the same accuracy as the ones carried out at a higher pressure .
In the first place , the sensitiveness of the oil-gauge is proportional to the pressure , and then serious trouble arose over the solubility of air in the oil ( Fleuss pump oil ) used in the oil-gauge .
Occasional minute air-bubbles made their appearance at the points marked 198 Mr. N. Eumorfopoulos .
The Boiling-point of Fig. 5.\#151 ; Oil-gauge Tap .
A and B ( especially A ) in the accompanying figure .
At the low pressure , oil slowly leaked into the oil-gauge through the tap .
As air could not do so directly , the only explanation seemed to he that it got carried in by the oil in solution , the change of solubility in the pressure accounting for the air-bubbles .
In addition , it was very likely that diffusion might take place through the oil .
In fact , it was noticed that , if two W.F.P/ s were taken in succession ( say on Friday and on Monday ) , the oil-gauge did not give the same reading on the two days .
The C side had apparently increased in pressure , presumably because the C tube is nearer the bulb C of the tap , from which the leakage apparently came .
This change of pressure was measured on three separate occasions , and was found to be equivalent to a change of the W.F.P. of 0*60 mgrm .
per hour\#151 ; this is equivalent to a change in the coefficient of 080 per hour with six significant figures , and in the S.B.P. of *00080 ' per hour .
This correction has been applied in all the low pressure measurements ; to show its importance , its magnitude is indicated in the tables of results .
Owing to the pressure in the thermometer being smaller than the atmospheric , mercury could not be removed from the expansion bulb ( bulb E ) with the apparatus as it stood , nor could the oil-gauge be conveniently manipulated .
To overcome these difficulties , a small piece of apparatus , shown in fig. 6 , was adapted to the thermometer .
By adjusting the reservoir B , the pressure in A could be varied .
Three tubes , with stopcocks , , were sealed to A. One communicated with the atmosphere , the second was connected to the open tube of the oil-gauge ( see fig. 2 ) , and the third to the bulb C , in which was collected the mercury from the expansion bulb .
The experiments with the old gas at a pressure of 39*5 cm .
gave the following results:\#151 ; Coefficient XlO8 .
Corrected for diffusion .
Up . .
3664791 ( 24 hours ) Down . .
366499 J r ( 4 ) Up . .
3664851 , ( 68 } ) ) Down . .
366580 J ( 4 ) Up . .
3663021 L ( 188 ) Down . .
366640 J r ( 4 J\gt ; ) Up . .
366555 ] L ( 20 )\gt ; ) Down ... . . .
366652 J r ( 4 )\gt ; ) Up . .
3666011 1 ( 20 \#187 ; ) Down ... . . .
366590 r ( 4 ) Sulphur on the Thermodynamic Scale .
Coefficient x 108 .
Corrected for diffusion .
Up . .
3665311 1 ( 68 hours ) 366585 Down . .
366564 j r ( 4 \#187 ; ) 366561 Up . .
3665041 ( 20 " ) 366520 Down . .
366556 J i ( 4 \#187 ; ) 366553 Down . .
366575 ( 4 \#187 ; ) 366572 } } Mean ... ... .
366561 x 10"8 .
S.B.P. Corrected for diffusion .
o o Up . .
444-4071 ( 20 hours ) 444-4231 Down . .
444-439 J r ( 4 33 ) 444-436 J Up . .
444-3831 i ( 20 33 ) 444-3991 Down . .
444-414j 1 ( 4 33 ) 444-411J Up . .
444-3341 l ( 20 33 ) 444-3501 Down ... . . .
444-339 J 1 ( 4 33 ) 444-336 J Mean ... ... ... 444*392 ' .
-rubber - glass - ground joint Fig. 6 .
Mr. N. Eumorfopoulos .
The Boiling-point of Once , when the nitrogen was giving an abnormal coefficient , an attempt was made to see if any gas was given off by the bulb .
For this purpose , an ordinary vacuum tube with electrodes was connected to the tube coming from the thermometer .
The bulb , kept in ice , was evacuated with the Topler until practically no discharge passed .
On placing the bulb rapidly in boiling water , in general , no change of luminosity was observed , although the nitrogen was on one occasion left for a month in the bulb to enable the silica to absorb the gas , if it felt inclined to do so .
Once only was there a sudden increase of luminosity .
For the following experiments , the fresh sample of nitrogen was used , and the second bulb :\#151 ; Pressure = 79-2 cm .
Coefficient\#151 ; Up 366987 x10~81 Down 367011x10~7 Down 366997 J Down 367003 Down 367013 Down 367004 Down 367005 Down 367009 Mean 367005 x 10-8 .
In taking the mean , I have counted as one observation the first two values , .so as not to give undue weight to a single boiling point observation .
S.B.P.\#151 ; Up ... ... . .
4441441 Down ... ... 444113 J Up ... ... 444120 " !
Down ... .
444116J Up ... ... .
444132 " !
Down ... .
444120 J Mean .
Up ... ... 444126 " !
Down ... .
444112 J Up ... ... .
4441181 Down ... .
444121J Up ... ... 4441431 Down ... . .
444140 J 444125 .
The pressure was now reduced to 41*5 cm .
In all these observations , a S.B.P. was taken in the morning , a W.B.P. in the afternoon ; and finally , about an hour later , a W.F.P. The correction for diffusion is then practically negligible in the case were as follows:\#151 ; 366578 x 10-8 Mean ... .
: coefficient ; the values for the latter 366601 xlO"8 366593 366598 366581 366573 366587 xlO"8 Sulphur on the Thermodynamic Scale .
201 ?
The following values for the S.B.P. were obtained Corrected for diffusion .
o 0 Up . .
444-1261 ( 68 hours ) 444-1801 Down ... . . .
444-390 J ( 4 \#187 ; ) 444-387 J Up . .
444-3261 ( 20 \#187 ; ) 444-3421 Down ... . . .
444-437J ( 4 \gt ; \gt ; ) 444-434J Up ... 444-281 ( 44 \#187 ; ) 444-3161 Down ... .
... 444-336J ( 4 }\gt ; ) 444-333 J Up ... 444-319- ) , ( 20 ) 444-3351 Down ... ... 444-342J ( 4 ) ) ) 444-339 J Up ... 444-2681 i ( 92 ?
\gt ; ) 444-3421 Down ... ... 444-345J 1 ( .
4 ) 444-342J Up ... 444-332* ' L ( 20 \#187 ; ) 444-3481 Down ... ... 444-429- 1 ( 4 V ) 444-426 J Up ... 444-368 ' l ( 72 }\gt ; ) 444-4261 Down ... ... 444-371- J ( 4 i ) ) 444-368 J Up ... 444-330 ' l ( 20 V ) 444-346 1 Down ... ... 444-366- J ( 4 )\gt ; ) 444-363 J Up ... .
444-332 ' 1 ( 44 f ) ) 444-3671 Down ... ... .
444-350 .
1 ( 4 J5 ) 444-347 J Up ... .
444-286 ' l ( 20 \gt ; \gt ; ) 444-3021 Down. .
... .
444-307 .
J ( 4 \gt ; ?
) 444-304J Mean ... ... . .
444*356 ' .
As a rule , in manipulating the oil-gauge , it was possible to prevent any of the small air-bubbles that formed at the tap from passing up into the thermometer .
Occasionally ( especially over the week end ) this was not easy , and for the first S.B.P. there is a note in my book that some minute bubbles passed up both sides .
I have , therefore , taken no account of it , in obtaining the mean .
After this , the nitrogen was removed , and the thermometer filled with air that was passed through potash solution , a phosphoric anhydride tube , and a spiral immersed in liquid air .
The following observations were made :\#151 ; Coefficient\#151 ; Pressure = 81*1 cm .
Up .
367098 xl0-8l Down 367083 x10"8 Down .
367043 J Up 3671071 Down .
367069 Mean Down 367079 x 10-* .
367076J 202 Mr. N. Eumorfopoulos .
The Boiling-point of S.B.P.\#151 ; o O Up .
444-1731 Up ... 444-1561 Down .
444-105 J Down ... ... 444-077 J Up 444-1361 Down 444-079 J The variations are much larger than in the case of nitrogen at a corresponding pressure .
Further , the " down " values of the S.B.P. are in all three cases about 0*07 ' smaller than the " up " values .
The difference is in the same direction in the two similar coefficient values .
After the last W.F.P. the bulb was heated up to 100 ' for 10 or 15 minutes , letting out mercury so as to keep the pressure constant , then it was cooled to 0 ' for a similar period , mercury being introduced .
This was done three times in all in succession .
The final result was , after all corrections are made , as if 22 mgm .
were driven out in the process .
This would make a difference of 30 units ( with six significant figures ) in the coefficient , and is in the same direction as the above-noted differences .
After this experience I thought it was useless to proceed further with air .
I had intended carrying out some measurements with argon , but it seems preferable first to clear up the source of the air discrepancies , and a search must also be made for an oil which dissolves less gas than the Fleuss pump oil .
The following coefficients have thus been found for nitrogen:\#151 ; Pressure .
Coefficient x 108 .
Pressure .
Coefficient XlO8 .
78-4 367019 39-5 366561 75-7 366991 41-5 366587 79-2 367005 To compare these , we shall reduce the first three to a common pressure 79*2 and the last two to 41*5 , with the following result : Pressure .
\lt ; Coefficient x 10s .
Pressure .
Coefficient x 108 .
r 367028 a , tr f366583 79-2 \ 367030 41 O 1366587 1-367005 I consider that the last value of each set is the more reliable .
If we adopt these , the value of the coefficient for a pressure of 0 cm .
is 0*00366127 , the reciprocal of which is 273*129 .
Sulphur on the Thermodynamic Scale .
203 The values for S.B.P. are as follows :\#151 ; Pressure .
Temperature .
Pressure .
78-4 444T40 39-5 75-7 444-125 41-5 79-2 444-125 Temperature .
444*392 444*356 Reducing these , as in the case of the coefficient , we have Pressure .
79*2 Temperature .
r 444*135 J 444*104 1444*125 Pressure .
Temperature .
41*5 f444*380 1444*356 Adopting as before the last values , we get for the value of S.B.P. on the thermodynamic scale 444*61 ' .
Experiments that are now being carried out on the Joule-Thomson effect will , it is hoped , throw further light on these reductions to the thermodynamic scale .
There is no need to collect here previous results on the S.B.P. , as this has been admirably done by Waidner and Burgess , and by Day and Sosman.* The principal results of this research are as follows :\#151 ; 1 .
The thermodynamic correction to the S.B.P. has been obtained experimentally , and the value of the S.B.P. on the centigrade thermodynamic scale found to be 444*61 ' .
2 .
The value of the melting point of ice on the absolute centigrade thermodynamic scale is 273*13 ' .
I wish to thank the Government Grant Committee of the Royal Society for a grant towards the expenses of this research .
* Waidner and Burgess , \#163 ; Bulletin of the Bureau of Standards , ' vol. 7 , No. 1 ( 1910 ) ; Burgess and Le Chatelier , The Measurement of High Temperatures , ' 1912 ; Day and Sosman , 'Amer .
Journ. of Science , ' vol. 33 , p. 529 ( 1912 ) .

