]\amp ; gt ; 105 ] : : : : II1 .
On the Ieasurement of Temperature .
Part I.\amp ; mdash ; On the Pressure Coefficients of Hydrogen and Helinm at Constant Votume and at rent Initiat Pressures .
Part II.\amp ; mdash ; On the Vapour Pressures of Oxygen at Temperatures below its Boiling\amp ; hyphen ; Point on the Constant\amp ; hyphen ; Votnme and Hetium Scates .
Part III.\amp ; mdash ; On the Vapour Pressures of Liqnid Hydrogen at Temperatures below its Boiling-Point.on the Constant\amp ; hyphen ; Hydrogen and Helium Scates .
MORRIS W. TRAVERS , D.Sc .
, Fellow of University College , London , GEORGEi SENTER , B.Sc. , ADRIEN JAQUEROD , D.Sc .
by Professor WILLIAM RAMSAY , F.R. Received June June 19 , 1902 .
PART I. On the PrJssure Coefficients of Hydrogen Helium at Constcmt Volume and Initial Pressures .
MORRIS W. TRAVEnS , D.Sc .
, Fellow ofUniversity , London , and ADRIEN JAQUEROD , D. Sc. CONTENTS .
Page 1 .
Introduction . . . . . .
106 2 .
Suitable thermometric substances . . . . . .
107 3 .
Previous measurements of the pressure coefficients of hydrogen and . . .
108 4 .
Accuracy of the results . . . .
109 5 .
Apparatus employed in the research .
The )arometer . . . . . . . . .
109 6 .
The constant\amp ; hyphen ; volume thermometer . . . . . . . . .
111 7 .
The calibration of the ) . . . . . . . . . . . . . .
113 8 .
The dead\amp ; hyphen ; space . . . . . . . . . . . . . .
lld 9 .
The stem of the thermometer . . . . . . . . . . . . . . . .
The volumes of the , stem , dead\amp ; hyphen ; spacc . . . . . . . . . .
Ihe manometer and its conncction with the , \amp ; c. . . . . . . . .
116 12 .
The introduction of gas into the . . . . . .
117 13 .
The scale . . . . . . . . . . . . . .
118 14 .
The temperature of the me1cu1y columll , . . . . . . .
1 15 .
The measurement of the essure on tho in the the1momete1 . . . . .
16 .
The ice\amp ; hyphen ; point . . . . . . . . . . . . . .
VOL. .
1 ' Page 17 .
The boiling\amp ; hyphen ; point .
18 .
Calculation of the results 19 .
The details of one experiment . . . . . . . . . . . .
123 20 .
The pressure coefficient of hydrogell . . . .
125 21 .
The pressure coefficient of helium . . . . . . . . . . . . .
128 22 .
The accuraoy of the results . . . . .
131 23 .
The final values for the pressure coefficients of hydrogen and helium . . . . .
131 24 .
The pressure coefficients of hydrogen and helium at lower initial pressures .
133 25 .
General conclusions 1 .
Introduction .
how much depends upon our knowledge of the relationship between the thermodynamic scale of temperature and the various scales of temperature as determined by means of gas thermometers , it is surprising that so little attention has been paid to the matter in recent years .
Indeed , instead of attempting to arrive at the necessary experimental data for the solution of the problem , it appears as if chenusts and physicists were prepared to accept scales of temperature , which differ considel.ably one from another .
Since the problem was cked by JOULE and LVIN , more thau halt a century , it has been made the subject of investigation by more than one mathematiciau .
Recently ( ' Phil. Mag 1901 , vol. 2 , p. 130 ) J. has succeeded in developi1g some fairly simple equations by means of which the differeuoe between the thermo\amp ; hyphen ; dynamic scale of temperature and the tempelature determined by means of gas thelmometers can be calculated .
The data necessary in the case of gas are the following:\amp ; mdash ; The coefficient of increase of pressure at constant volume .
The coefficient of increase of volume at constant pressure .
The variation of " " \amp ; ldquo ; with pressure at different temperature .
The Joule\amp ; hyphen ; Thomson effect and its variation with temperature .
Applying existing data to his equations , \amp ; hyphen ; INNES arrived at resuIts which , as the following table shows , are anything but satisfactory : \amp ; hyphen ; Correction of the Constant\amp ; hyphen ; Volume Thermometer .
ON THE MEASUREMENT TEMPERATURE .
107 Correction of the Constant\amp ; hyphen ; Pressure Thermometer .
Of the data employed in making the foregoing calculation only the pressure coefficients of hydrogen and nitrogen , which have been determined by P. CHAPPITIS , of the " " Bureau International des Poids et Mesures and the variation of " " \amp ; ldquo ; with pressure at the normal temperature , which has been recently investigated by ( ' Leiden Laboratory Reports , ' 1901 ) in the case of hydrogen , can be considered as being known with sufficient certainty .
AMAGAT has determined the variations of " " \amp ; ldquo ; for nitrogen , but the experiment was made at high pressure and not at the pressure under which thermometric measurements are usually made .
In view of the fact that both air and oxygen appear to obey BOYLE'S law at pressures below 150 millims. of mercury RAYLEIGH , Phil. Trans , extra\amp ; hyphen ; polation of AMAGAT'S results at high pressure does not seem justifiable ; and to obtain the data necessary for the correction of the gas thermometer the variations of " " \amp ; ldquo ; should be carefully investigated at pressures below two atmospheres .
REGNAULT'S values for the volume coefficients of nitrogen and hydrogen are not to be upon .
The Joule\amp ; hyphen ; Thomson effect for air and hydrogen was detel'mined by JOULE and Lord KELVIN more than half a century ago .
Without discussing their results , it may be pointed out that by employing electrical methods for the measurement of temperature , it should now be possible to measure the effect with a much higher degree of accuracy , and over a wider range of temperature than was then possible An investigation in this direction has been commenced by Dr. .
A. LEHFELDT conjointly with one of us .
2 .
Snitable mometric Substcmces .
For the reason that the composition of atmospheric air varies with the prevailing condition , .
is not a suitable substancs for use in thermometry .
Atmospheric nitrogen appears , however , to contain a consfant quantity of argon and its companions , and may be employed in thermometry over a range , limited , perhaps , by the tem\amp ; hyphen ; peratures at which it condenses on the one hand and dissociates on the other .
Until recently , hydrogen appeared to be the most important thermometric substance .
DR. M. W. TRAyERS , MR. G. , AND DR A. JAQUEROD The constant\amp ; hyphen ; volume hydrogen scale peared to approximate closely to the absolute scale of temperature , and the gas could be employed in any measurements to which thermometers could be applied .
For this reason the Commit \amp ; eacute ; International adopted as the normal scale of temperature the scale of a constant\amp ; hyphen ; volume hydrogen there \amp ; hyphen ; mometer , in which the pressure at the ice\amp ; hyphen ; point was 1000 millims. of mercury .
Recently , however , we have succeeded in reaching temperatures at which the vapour pressure of hydrogen itself is extremely small , and which cannot be investi\amp ; hyphen ; by means of the hydrogen thermometer with any degree of accuracy ( see p. 179 ) .
there , it has now been ascertained that at high temperatures hydrogen will reduce glass , porcelain , and even silica ; and as the gas diffuses readily through the walls of platinum vessels , its application in this direction must be considered limited .
It has been pointed out by one of us Experimental Study of Gases p. 156 ) that pure helium appears to be a much more perfect thermometric shbstance than hydrogen .
It is chemically inactive ; and , so far as we can predict , is incapable of undergoing dissociation at high temperatures .
Hence it is probable that measurements of high temperatures , made by means of quartz thermometers filed with helium , are alone to be relied upon .
For the measurement of low temperatures it presents similar advantages ; for as it appears probable that its critical point lies below , it may be considered as remaining a very perfect gas down to the lowest temperatures which it has hitherto been possible to reach .
Further , as the following results show , the hydrogen and helium scales agree closely between and 10 C. , and consequently it may be assumed that at temperatures above C. the difference between the thermo\amp ; hyphen ; dynamic scale of temperature and the scale of a constant\amp ; hyphen ; volume thermometer filled with one of these gases is small .
At lower temperatures the divergence between the two scales is small , but the contraction of the helium being always less than that of the hydrogen indicates that the former remains the more perfect gas .
Unless some hitherto unknown objection is attached to the use of the helium thermometer , the helium scale should the hydrogen scale as the normal scale of temperature .
3 .
Previous Measurements of the Pressure Coefficient of Hydrogen and Helium .
The first step in the series of researches on which we have embarked consisted in the determination of the pressure coefficients at constant volume of hydrogen and helium between the melting\amp ; hyphen ; point of ice and the boiling\amp ; hyphen ; point of water under standard pressure , which , in the basement of University College , is equivalent to millims. of mercury at C. The coefficient for hydrogen , at a pressure of 100 millims. of mercury at the ice\amp ; hyphen ; point , is given by CHAPPUIS Travaux et M\amp ; eacute ; moires du Bureau International des Poids et Mesures as , the mean of seven determinations , of which the highest is and the lowest , and on this deter\amp ; hyphen ; ON THE MEASURffiMENT OF TEMPERATURE .
mination the normal scale of temperature of the Commit \amp ; eacute ; International is based .
Kammerlingh ONNES ( 'Leiden Communications , ' 1901 , No. 60 ) gives as the value of the coefficient , the actual numbers being .
The result obtained by REGNAULT , , need not now be considered .
It has hitherto been assumed that the pressure coefficient for hydrogen is independent of the initial pressure , and does not , as in the case of nitrogen , decrease and at length attain a limiting value .
As important deductions have been based on this statement , we decided to measure the coefficient for our standard gases at initial pressures corresponding to 350 , 500 , and 700 millims. of mercury .
An approximate measurement of the pressure coefficient of helium was made by KUENEN and RANDALL in 1895 ( ' Phil. Mag In employing the helium thermometer for the measurement of low temperatures , OLSZEWSKI assumed that the coefficient is the same as that of hydrogen .
4 .
of the\amp ; hyphen ; Results .
In order to deternune the fifth significant figure in the pressure coefficient , the total error must not exceed 1 part in 40,000 , a degree of accuracy which it is practically impossible to attain in individual experiments .
As will be seen later , the greatest error in our observations may be attributed to inaccuracy in the scale which , over the whole length , may be considered accurate to at least millim. , or to 1/ 25,000 of the increase of pressure on the gas between and 10 C. at an initial pressure of 700 millims. 5 .
employed in the .
The Since at the boiling\amp ; hyphen ; point of water a change of pressure of millim. of mercury corresponds to a change of temperature of C. , or 1/ 20,000 of the difference between the boiling and freezing\amp ; hyphen ; points of water , it was sufficient to observe the atmospheric pressure with this degree of accuracy .
The barometer employed is shown in fig. 1 ; it was constructed by us specially for this research .
The upper chamber was 20 millims. in diameter and 200 millims. in length .
It was sealed to the stem , which was drawn out at the lower end to a capillary tube ofan internal diameter of about 1 millim. In filling the barometer , the capillary end of the tube was sealed to a tube ( fig. 2 ) , of which the branch led to the mercury pump , and the branch projected downward and was drawn to a fine point , which was sealed .
The A and of the apparatus were heated , and , when the apparatus was thoroughly exhausted , the point of the tube was broken below the surface of some warm mercury a basin .
The mercury entered slowly , and , passing through the constriction , ran down into , which was heated sufficiently to boil the liquid .
When the apparatus was completely filled , the tube was cut through the constriction , and the small bubble of air which remained in was replaced by mercury .
Both the upper chamber and the limb of the tube which formed the lower chamber were set by means of plaster of Paris into brass mounts , which were soldered to plates II ' , shown in section in the drawing .
In the case of the upper chamber this was done after it had been filled with mercury .
In setting up the barometer the tube FG was first fixed to the piece of wood which supported the instru\amp ; hyphen ; meant , and was then filled with mercury .
The upper portion was then inverted , the open end plunged into the limb of the tube , and the brass plate I screwed to wooden support L. A dish was placed below the apparatus to catch the mercury overflowing from ; finally some of the mercury in the tube was removed by means of a syphon , so that the surface of the mercury in the upper and lower chambers appeared as in the figure .
The scale , which was placed in front of the barometer , consisted of a piece of plate\amp ; hyphen ; glass to which were joined , by means of Canada balsam , two short scales 100 millims. in length and divided in millimetres .
To prevent them shifting , through the flowing of the cement , they were further fixed by a binding of copper\amp ; hyphen ; wire covered with plaster of Paris at the top and bottom .
The scale was compared with a standard scale from the Physikalische Reichsanstalt at Charlottenburg , and afterwards at the National Physical Laboratory with a standard metre ( p. 119 ) .
The temperature of the column of mercury was measured by means of a thermometer divided to C. placed half\amp ; hyphen ; way between the top and bottom .
ON THE MEASUREMENT OF .
MPERATURE .
Readings of the barometer were made by means of two telescopes placed on a stand opposite the instrument .
The barometer itself was attached to a pillar so that a card placed behind ( fig. 3 ) it at an angle of to the line of sight gave a good iUumination from a window at the side .
The upper half of the card was black , the lower half white , and before taking a reading the card was adjusted so that the diyiding line appeared about 1 millim. above the surface of the mercury .
By this arrangement the meniscus appeared perfectly sharp , as all surface reflections were completely eliminated .
Mercury Thermometers.\amp ; mdash ; The mercury thermometers em\amp ; hyphen ; ployed in these obseryations were compared with the thermometer employed to measure the temperature of the dead\amp ; hyphen ; space .
This thermometer was itself compared with the gas thermometer and the error oyer the small range of temperature required , viz. , to C. , was known to within .
The weight of such an .
will be discussed later .
6 .
The Constant\amp ; hyphen ; Volume Thermometer .
The atus employed in the determination of the pressure coeflicie1lts of gases diffel .
S in many respects from that employed by other obsel .
verse .
HAPPUIS employed thermometers of hard glass and of platino\amp ; hyphen ; iridium , and ONNES bulbs of hard glass ; in both cases the connection between the thermometer bulb and the dead\amp ; hyphen ; space was made by means of a steel tube co1mected with the glass by means of cement .
Beyond the fact that hard glass has a slightly lower coefficient of expansion than soda glass , there is no advantage in employing it ; on the other hand , soda glass is more easily worked , and as by using it the bulb , stem , and dead\amp ; hyphen ; space can ] be made .
in one piece , considerable advantage is to be gained .
The bulb of our instru\amp ; hyphen ; meant ( fig. 4 ) was made by sealing a piece of glass tube 80 millims. in diameter to it capillary tube millim. in internal diameter , which for conyenience in calibration was divided in millimetres for a portion of its length .
The stem was bent at right angles at ( fig. 4 ) , and ultimately sealed at to the portion of the stem leading to the dead\amp ; hyphen ; space .
ON ASURhMFN T OF 7 .
Calibration of Bulb .
The volume of the bulb was determined by it empty , and afterwards filled with water at the melting\amp ; hyphen ; point of ice .
A glass tube was connected with the stem by a piece of rubber tube to prevent loss water when the temperature rose .
The following are the results:\amp ; mdash ; The coefficient of expaloeion of the glass of which the thermometer bulb was made was determined by the \amp ; hyphen ; thermometer method .
The \amp ; hyphen ; thermometer was made the same glass tube as the thermometer bulb .
The value the coefficient was found to be The composition of the glass employed , which was obtained from .
C. M\amp ; Uuml ; LLER \amp ; Co. , of High Holborn , appears to be very constant ; for the coefficients of expansion of samples of it determined by KUENEN and RANDALL ( loc. cit. , p. 109 ) in 1896 , and by one of us in 1901 , have the same value as that last determined .
The change of volume of the bulb due to change of the pressure on the gas inside it was also measured .
It was determined by filling the bulb with water to a point on the stem and measuring the change in the position of the tneniscus when the was reduced by means of a water pump .
The change of volume amounted to cub. centim. per atmosphere .
8 .
Both and ONNES employed a glass tube with a steel cap attached to it by cement , and enclosing a steel for the adjustment of the merculy meniscus .
In the construction of our instrument we have entirely dispensed with steel and cement connection , and have employed an opaque glass point sealed into the tube , which formed the dead\amp ; hyphen ; space at its point of junction with the stem .
The tube for the dead\amp ; hyphen ; space cut from a piece of glass tube of 9 millims. internal diameter , selected for its straightness and regularity of bore ; the manometer tube was cut from the same piece .
In making the junction with the capillary tube great care was V0L .
DR. .
TRAyERS , MR. G. SENTER , AND .
A. JAQUEROD taken to heat only about 3 millims. of the wider tube , so as not to disturb the regularity of its surface at the level at which the mercury meniscus was afterwards observed .
The operation of sealing the glass point into the tube not an easy one .
After bending the capillary tube as in the figure , the volume of the dead\amp ; hyphen ; space was determined in the following manner .
The wide tube was cut about 30 millims. below its junction with the tube and sealed to a capillary stop\amp ; hyphen ; cock drawn out to a point at its other end ( fig. 5 ) .
The aratus was fixed ni a vertical position , a tap\amp ; hyphen ; funnel and rubber tube were attached to the stop\amp ; hyphen ; cock , and mercury was run into the tube till its surface came exactly into contact with the point ; the final adjustment was made by closin the tap on the funnel and compressing the rubber tube with a screw\amp ; hyphen ; clip .
The stop\amp ; hyphen ; cock on the dead\amp ; hyphen ; space was then closed , the rubber tube remoyed , and the mercury in the dead\amp ; hyphen ; space run out into a porcelain crucible and weighed .
The rubber tube was then again attached to the apparatus , which was next filled with mercury to a mark on the stem .
This mercury was weighed , and from the difference of the two weighings the volume of gas in the dead\amp ; hyphen ; space , between a mercury meniscus in contact with the point and a mark on the stem , was calculated .
The experimental results are as follows:\amp ; mdash ; In making an actual experiment , the mercury in the dead\amp ; hyphen ; space was brought to within a short distance of the point ; this distance was measured by means of the telescope and micrometer eye\amp ; hyphen ; piece used in measuring the pressure ( p. 120 ) .
As the ON THE MEASUREMENT OF TEMPERATURE .
115 volume of the dead\amp ; hyphen ; space was thus somewhat greater than that determined by the method just described , it was necessary to apply a correction amounting to where is the distance between the surface of the mercury and the ) , and is the radius of the tube ( see p. 121 ) ; : As will be shown later , it was necessary in measuring the pressure to apply a correction for the heights of the mercury meniscus in the deaCl\amp ; hyphen ; space and mano\amp ; hyphen ; meter tube of the thermometer , which varied slightly in every experiment .
If the meniScus in the dead\amp ; hyphen ; space be considered to have the form of a semi\amp ; hyphen ; ellipsoid of height and of radius , the change of volume of the dead\amp ; hyphen ; space with change of height of the meniscus will be given by The mean height of the meniscus was found to miHims .
, and the extreme variation for any observation never exceeded millim. , which corresponds to a change of volume of cub. centim. in the dead\amp ; hyphen ; space .
As such an error had a considerable effect on the value of and , we were obliged to take it into consideration .
9 .
The Stem .
In our earlier experiment we had found that , if by any accident the mercury reached the top of the dead\amp ; hyphen ; space , minute globules of it adhered to the glass and were in time cal.ried over into the b\amp ; uacute ; lb. As the vapour pressure of mercury is millim. at C. , this gave rise to serious errors in our measurements .
To obviate this difficulty we introduced into a portion of the stem between and a closely\amp ; hyphen ; wound coil of silver millims. long and slightly over 2 millims. in diameter .
The tube was made slightly larger at this point .
The volume of the space was easily determined by weighing the tubes empty and full of water .
10 .
Volumes of Bulb , Stem , Dead\amp ; hyphen ; 6p ace .
the point in the dead\amp ; hyphen ; space , the whole of the internal space was considered as divided When the instrument was set up permanently and the mercul.y brought close to into three regions .
I. The bulb and stem to within 40 millims of the horizontal portion .
Volume at ice\amp ; hyphen ; point cub. centims. Volume at boiling\amp ; hyphen ; point cub. centim , Q. .
The stem from the point already referred to to the middle of the wall of the water space enclosing the barometer tube .
The temperature of the stem was determined with sufficient accuracy by means of a thermometer with the bulb placed between and Volume cub. centim. In some of our earlier experiments , about 3 centims. of the stem outside the meter was enclosed between sheets of asbestos , and the temperature was determined by means of a mercury thermometer .
The mean temperature of this portion of the stem was usually about C. , and , as its volume was only 0182 cub. centim. , its temperature could be taken as 10 C. without affecting the value of III .
The dead\amp ; hyphen ; space and remainder of the stem .
The dead\amp ; hyphen ; space and barometer were enclosed in a water jacket , as will be described presently ; the temperature was determined by means of the mometer v Volume } , where is the distance in centimetres between the point and the surface of the mercury , and is the height of the meniscus .
11 .
Ianometer its the \amp ; hyphen ; Space , The pressure on the gas in the thermometer was determined directly by observing the difference of the level of this mercury meniscus in the dead\amp ; hyphen ; space and in a manometer .
By employin this method it was necessary to take only two readings , and the observed pressures were entirely independent of the atmospheric pressure .
The tube which formed the dead\amp ; hyphen ; space after removing the stop\amp ; hyphen ; cock was sealed at its lower end to a wider tube , which was connected with the manometer tube and with a tubs , which served to catch any air bubbles which might enter the apparatus the stop\amp ; hyphen ; cock and the rubber tube leading to a reservoir .
A tube 5 millims. in diameter , sealed into the side of the tube , was connected to a stop\amp ; hyphen ; cock of capillary bore , through which the gas could be introduced into the apparatus .
The tube terminated above in a narrow glass tubs , connected by a piece of india\amp ; hyphen ; rubber pressure tubing with one limb of a glass stop\amp ; hyphen ; cock , through which traces of air entering through the tube could be expelled .
By closing the stop\amp ; hyphen ; cock and compressing the rubber tube by the pinch\amp ; hyphen ; cock the final adjustment of the level of the mercury in the dead\amp ; hyphen ; space could be effected .
The manometer tube was carried vertically upwards from the bottom of and above the level of the dead\amp ; hyphen ; space , so that its axis coincided with that of the latter .
In observing the pressure on the gas in the thermometer , it was only necessary to read the difference of level between surfaces of the mercury in the dead\amp ; hyphen ; space and in the manometer tube which lay in the same vertical line .
The top of the ON THE IsIEASUREMENT OF PERATURE .
ltlanometer tube was millims. above the level of the dead\amp ; hyphen ; space , so that when the pressure on the gas in the thermometer was 1000 millims. there was still a considerable vacuous space above the mercury .
The manometer tube was boiled out and the apparatus filled with mercury before the thermometer was fixed to its stand , or the bulb sealed to the capillary tube leading to the dead\amp ; hyphen ; space .
For this purpose the apparatus was set up temporarily in clamps , and the top of the manometer tube , which was drawn out to a fine capillary ( fig. 4 , A ) , was sealed to a tube leading to the mercury pump .
The capillary tube connected with the dead\amp ; hyphen ; space was also closed .
After thoroughly exhausting the apparatus , and heating it to remove trace of water condensed on the glass , mercury was admitted through the stop\amp ; hyphen ; cock .
As the mercury entered the manometer tube it was boiled by heatin the tube with a Bunsen burner .
By this method , which we have frequently employed , the rising surface of the mercury remains in contact with mercury vapour only , and as the exhaustion is continued throughout the operation a very perfect barometric vacuum is finally obtained .
By admitting air to the ) through the stop\amp ; hyphen ; cock , and lifting the reservoir attached to the tube , the mercury in the manometer tube was raised nearly to the top .
By heating , mercury vapour was made to pass through the constriction , and condense in the descending tube .
The constriction was then iu the blow\amp ; hyphen ; pipe , and the manometer as sealed .
The mercury was then allowed to fall in the manometer tube .
The bulb was not sealed to the rest of the apparatus till this operation had been completed , as it was found that it was impossible to prevent mercury vapour condensing in it .
The connection was now made by sealing the stem with a mouth blow\amp ; hyphen ; pipe .
It was unnecessary to make any correction for the distortion of the stem , as the volume of 1 millim. of its length amounted only to cub. centim. The instrument was now fixed to its permanent support , as in fig. 4 .
This consisted of a stout wooden plank , 3 centims. thick and 15 centims. wide at the bottom , cut away for 6 centims. of its width from the level of the dead\amp ; hyphen ; space upwards to allow of the illumination of the scale from behind .
The support was stayed rigidly at the top to the wall and to a beam running transversely across the room , so as to prevent all lateral movement .
We do not intend to give an account of the experiments which we carried out before we commenced the series of measurements on which we base our results .
It will suffice to state that about four months were spent and sources of error such as that due to lack of rigidity in the apparatus .
12 .
Introduction of the the ) atus .
The gases employed in these experiments were hydrogen and helium , and it will be convenient to consider the method by which they were prepared and purified at a later stage .
The gas was introduced into the the Dll .
TRAVERS , MR. G. SENTER , AND DR. A. JAQUEROD which this purpose was connected through a bulb and a tube to the apparatus containing the gas , and to a mercury pump .
The tube was intended to receive the mercury contained in the portion of the side tube above the stop\amp ; hyphen ; cock .
By lowering the reservoir connected by the rubber tube with the stop\amp ; hyphen ; cock the level of the mercury in could be made to fall below the point of entry of the side tube connected with .
Then by carefully opening the cocks and alternately lnanipulating the mercury pump , the gas in the apparatus could be completely removed .
When this was effected the mercury in the tubes , and stood at the same level .
Before introducing the gas the bulb was heated in an air\amp ; hyphen ; bath for about hour to 15 C. to move all traces of moisture , was washed out several times with the dry gas .
The operation of filling the apparatus was the reverse of that of emptying it .
The gas was admitted slowly by opening the stop\amp ; hyphen ; cock , and at the same time mercury was admitted through the stop\amp ; hyphen ; cock .
The quantity of gas required to fill the apparatus to the desired pressure was at first found by trial , but after the first experiment it was known that when the level of the mercury in the tube lay just below the tube leading to , the mercury rose to a certain height in the manometer tube .
When sufficient gas had been admitted the stop\amp ; hyphen ; cock was closed and more mercury was allowed to enter the apparatus through .
The gas in the portion of the side tube above was easily displaced by mercury .
( See also Part III .
, p. 171 , et seq. ) 13 .
The Scale .
The measul.ement of the pressure in the thermometer bulb could have been easily and effectively made by employing a well constructed cathetometer with a standard scale attached to it .
As , however , we had no means of obtaining such an instrument , we were obliged to employ the following method .
On the surface of a piece of plate\amp ; hyphen ; glass short scales ruled in millimetres by ZEISS of Jena , were cemented by means of Canada balsam .
The scale so constructed was fixed in front of the 1nercury column and dead\amp ; hyphen ; space , so that when the lowest scale , which was 50 millims. in length , was in front of the dead\amp ; hyphen ; space , the four remaining scales were 100 millims. in length .
With this arrangement it was possible to fill the thermometer at three initial pressures , , at 350 , at 500 , and at 700 millinR of mercury , and to determine the pressure coefficient of the gas at each pressure .
The method of observation will be described later .
The distances between the points on the scale at which observations were made were first of all determined by us by means of a mi rometer apparatus , designed by Mr. HILGER for measuring spectro\amp ; hyphen ; photographs , with a screw of 1\amp ; hyphen ; millimetre pitch drum divided into 100 parts .
By means of this instrument it was possible to determine lengths of 150 ] lims .
to millim. At the end of our research the ON THE MEASUREMENT OF .
119 scale was sent to the National Physical Laboratory , and was there compared with a standard brass metre scale ( Soci\amp ; eacute ; t\amp ; eacute ; Genevoise , No. 59 ) ; the comparison was guaranteed to millim. For this we wish to express our thanks to the Director , sets of comparison are very concordant , except so far as .
the distance between scale I. and scale III .
is concel .
Iled .
With regard to the discrepancy of millim. which exists here , we wish to state that the values of the pressure\amp ; hyphen ; coefficients first obtained for hydrogen at initial pressures of.350 and \amp ; sect ; ; 500 millims. , indicated that there was probably small in the scale at point .
This has been corrected .
In the following table column A gives the distances between the line 9 on the lowest scale ( I. ) and the middle points on the four remaining scales , as determined by us ; column gives the vahJe stated in the certificate furnished by the National Physical Laboratory , while the numbers in column are calculated from column by multiplying the latter by the factor Line , I. , to line , scale yIVII .
C. It need not be pointed out here , that in the measurements of the pressure\amp ; hyphen ; coefficient it is only necessary to know the absolute length of the scale approximately .
We have accordingly taken the figures in the second column ( B ) as representing the distances between the standard points on our scale .
14 .
The Temperatvre of the Column .
Since in these experiments we were attempting to measure the pressure with a degree of accuracy approaching millim. , it was necessary to take special precaution with regard to the temperature of the mercury column .
In our first experiments we attempted to measure the temperature of the column by means of four thermometers , the bulbs of which were enclosed in glass tubes , of the same diameter as the barometer column , filled with mercury .
We found , however , that the temperature could not be read with certainty to C. , nor even with that degree of accuracy when the steam point was being determined .
As an error of in the temperature of the column involves a corresponding error of millim. in millims we enclosed the dead\amp ; hyphen ; space and mercury columns in a water jacket . .
G. , AND Dh ' .
A. JAQUFROD manner in which this was effected is shown in .
The dead\amp ; hyphen ; space and manometer column year enclosed between two ) ieces of plate\amp ; hyphen ; glass , and pieces of plate\amp ; hyphen ; glass in front forming the scale ; strips of wood , .
and , formed the side and bottom of the waterjacket , the hole being made water\amp ; hyphen ; tight by means of a lnixture of red\amp ; hyphen ; lead and gold size .
The arrangement was carried on brass brackets screwed to the of the main support .
The two plates of glass were clamped to the wooden strips by brass bands and wooden wedges ; the brass ands were screwed at one end to the main support .
Water entered the jacket at the bottom , and escaping at the top by a rubber ) , flowed over the tubes and , and the portion of below the water jacket .
This of the apparatus was swathed in cotton cloth , and was maintained at a , though indefinite temperature , by means of the current of water .
The temperature of the water in the jacket remained constant within for a sufficient time for the measurement of the pressure , and rarely varied by during the period of one hour , necessary for four consecutive observations .
The error due to temperature on ( 950 millims. ) is thus reduced to millim. which is less than the error of observa\amp ; hyphen ; tion .
It may also be noted that , as the temperature of the dead\amp ; hyphen ; space was made with the same degree of accuracy , and an error of in determining it makes rence of millim. on , this source of ) also disappears , The height of the column of which the temperature varied between and reduced in every case to the height at C. It is obvious that as the manometer tube and the glass scale in front of it are not parallel , and the space between them is filled with water , a refraction error may be introduced which will have a constant influence on the value of and .
If the deviation of the manometer tube amounts to 2 millims. in its total length , the error in reading the height of the column would be roughly millim. By observing the height of the column when the jacket contained water , and then allowing the water to escape rapidly , we could , however , detect no change the position of the meniscus .
Further , since the manometer tube was sufficiently straight for the error to and equally , the error due to refraction can be neglected .
10\amp ; hyphen ; .
of Pressmre on the the : The Vations made by means of a telescope aced at a distance of one metre from the scale .
The telescope was fitted with a Hilger ocular micrometer , with a screw , divided into 100 parts , each corresponding to a movement of the cross\amp ; hyphen ; wire of millim. , or to millim. on the scale .
Each meniscus was illuminated from behind by means of an electric glow\amp ; hyphen ; lamp covered with white tissue paper ; cards placed behind the glass plate , cut off the light at a height of 1 or 2 millilns .
bove the level of the mercury ( cf. p. 109 ) .
\amp ; dollar ; ON THE MEASURLhlENT OF IPERATURE .
121 When the temperature of the bulb and column were steady , the mercury in the dead\amp ; hyphen ; space was adjusted first of all by means of the mercury reservoir and stop\amp ; hyphen ; cock , and finally by means of the pinch\amp ; hyphen ; cock .
The mercury was always brought to within about millim. of the point .
After tapping the support with the hand , .
and a short time the mercury to settle down , the seryations were made .
The spider\amp ; hyphen ; line of the mlcrometer was first brought into coincidence with the line on the scales next above the meniscus , and then with the top of the meniscus itself , readings being taken in both positions .
A light was then held in front of the ' instrument , and the height of the meniscus was determined by bringing the spider : line into coincidence with the point at which the vertical bright line , caused by the reflection of the light from the curved surface of the mercury column , came to an end .
The height of the mercury meniscus was applied to the calculation of the correction to be applied for capillarity ( KOHLRAUSCH ) .
The determination of the position of the point with regard to the scale was also necessary for the calculation .
of the distance of the mercury from it , and from this the volume of the dead\amp ; hyphen ; space .
Its value , and the value of one division of the micrometer screw in terms of a division of the scale , were measured separately .
Observations of the pressure were invariably made by each one of us altel.nately , the other reading the thelmometer encl sed in the water jacket , and the..thermometer laced close to the horizontal portion of the stem .
position of the lowel meniscus was observed , then that of the upper meniscus , clnd finally the lower meniscus again .
The level of the mercury in the appal.atus was usually eadjusted .
of measul.ements . .
The \amp ; hyphen ; Point .
The bulb was immersed to within 40 millims. of the ontal portion of the stem in an inve ted glass bell\amp ; hyphen ; jar , and sul.rounded with broken ice packed tightly round it .
The bell\amp ; hyphen ; jar was filled nearly to the top of the vessel with distilled water .
Samples of the ice were on two occasions melted and the liquid evaporated ; in neither case was it found to contain more than a minute trace of solid matter .
A piece of wood , with slit cut in it the stem of the thermometel , was placed on top of the bell\amp ; hyphen ; .
containing the ice .
17 .
The Boihng\amp ; hyphen ; Point .
: The bulb of the thermometer was surrounded with dry steam by means of a double\amp ; hyphen ; walled copper hypsometer of the usual form .
The stem of the thel.mometel .
passed through a hole in the cover of the hypsometer , and the ture fairly steam\amp ; hyphen ; tight by means of a small strip of wet wound ound the sten ] .
No correction of the difference of pressure side and outside the V0L .
To shield the apparatus against radiation from the heated vessel a piece of wood , kept wet by a stream of water , was used as a screen .
A similar piece of wood , with a slit cut in it for the stem , was held in a clamp horizontally about 1 centim. above the top of the hypsometer to screen the horizontal portion of the stem .
The temperature of the short length of stem between the hypsometer and the screen was usually about , a difference from too small to be considered .
During each experiment the barometer was always read three times ; at the beginning , at the end , and between the two pairs of observations .
18 .
Calculation of the Results .
In calculating the value of the pressure coefficient from the results of our expel.iments , we employed the usual methods .
The pressure and which the would exert at the temperature of melting ice , and at that of saturated water pour at the nol.mal pressure , if confined in a space corresponding to the volume of the thermometer at C. , were first calculated .
The pressure coefficient was then obtained from the equation In and it is of course necessary to assume some value for the essure coefficient in ] order to apply the necessary correction for the gas contained in the stem and dead\amp ; hyphen ; space ; and in reducing the tempel.atul.e of the bulb to 10 C. in ting P. These corrections are , however , small ; and no appreciable error is introduced by taking the coefficient as 1/ 273 .
The following equations were employed in calculating and .
Where is the observed pressure at the is the obsel.ved pressure at the steam\amp ; hyphen ; point .
is the temperature of steam corresponding to the baroluetric pressure .
and are the temperatures of the stem .
and the temperatures of the , is the volume of the bulb at the ice\amp ; hyphen ; point .
is the volume of the bulb at the steam\amp ; hyphen ; point .
is the volume of the stem .
is the volume of the dead\amp ; hyphen ; space .
, is the total volume of the method by which the temperature , pressure , volume were calculated .
COl.rectedhas already been dealt with .
ON THE OF TEMPERATURE .
19 .
Details the Results of One Experiment .
As we have already stated , each determination of the ice\amp ; hyphen ; point or boiling\amp ; hyphen ; point consisted of four separate observations of the height and temperature of the barometer column , of the temperature of the stem , \amp ; 0 .
The readings were taken by each of us alternately .
In determining the boiling.point the barometer was also read three times , at the beginning , at the end , and in the middle of each set of experiments , and the temperature of the bulb was calculated from the mean corrected pressure .
It would be useless to set down the whole of the measurements involved in every experiment , but in order to give an idea of the accuracy of the observations , we give the full details of one complete set of measurements .
ON THE MEASURLMENT OF 22 .
Accuracy of In describing the construction of the apparatus and the manner in which the experiments were carried out , we have already pointed to the errors which may occur and their effect on the accuracy of the final result .
It will have been noticed that with the exception of the errors in the barometer reading , and in the tempera\amp ; hyphen ; ture of the steam in the hypsometer , the errors of observation affect and almost equally , and hence have little effect on the value of the coefficient .
Further , since an error of millim. in the barometer reading only changes the value of ( 950 millims. ) by millim. , the effect of such an error on the value of the coefficient is less than five units in the sixth significant figure .
With regard to the measurement of the pressure , as we have already stated , the errors in the scale were certainly less than millim. An error in the difference of this magnitude would involve an error of one unit in the fifth significant figure .
The actual extreme difference between the value of in the set of three observations which constitute each series usually amounts to to millim. , but these diflerences appear to be unbiased and do not influence the final result in one direction or in the other .
The value of the temperature correction practically disappears , for , as we have pointed out , the temperature of the column could be determined to C. , and in successive determinations of and the tempera\amp ; hyphen ; tures were practically identical .
The capillarity correction involving the height of the meniscus we considered at first to have a doubtful character , but as the height of the meniscus varied by millim. in different observations , and the application of the correction brought the observed pressures to the same value , we have come to the conclusion that it is sound .
The errors in the measurement of the volume and temperature of the dead\amp ; hyphen ; space are small .
Their effect on the value of the coefficient ' has already been sufficiently discussed .
23 .
The Final Values for the Pressure oefficient of ) und As the tables in Sections 20 and 21 show , we have two series of deternlinations of the pressure coefficient both in the case of hydrogen and of helium at initial pressure of about 700 millims. of mercury .
The results are as follows:\amp ; mdash ; Each of the observed values is based , as the tables show , on eight measurements of the pressure on the , the tempsrature of dead\amp ; hyphen ; space , \amp ; c. , when the bulb of the thermometer was immersed in ice , and twelve similar measurements where the bulb was surrounded with steam .
The agreement between the observations is satisfactory .
As has already been pointed out at the commencement of this paper , the best determinations of the pressure coefficient for hydrogen are those of CHAPPUIS and ONNES .
' work is to be found in the et Me'moires du Bureau International des Poids et Mesures , ' vol. 6 ( I888 ) , and in the subsequent volumes ; an account of his work is also to be found in the 'Rapports du Congr\amp ; egrave ; s International de Physique ' ( Paris , 1901 ) .
The results obtained in 1887 with his large thermometer , which has a platinum\amp ; hyphen ; iridium bulb of nearly 1 capacity filled at an initial essure of 1000 millims. of mercury , are as follows:\amp ; mdash ; The mean value of four determinations cal.ried out in 1899 with the was , and of five determinations made in 1895 with a bulb of " " verre dur\amp ; ldquo ; was .
The mean value of the seven determinations made in 1887 is , however , retained as the ] ) robable value of the coefficient , which forms the basis for the definition of the so\amp ; hyphen ; called normal scale of temperature .
The three determinations of the coefficient by ONNES ( ' Communications from the Physical Laboratory of the University of ' No. GO ) are as follows:\amp ; mdash ; As in the of CHAPPUIS ' experiments , the initial in the thermometer was about 1000 millims. of mercury .
Though the highest initial pressure at which our measurements were made was 700 millims. of mercury , there is , as we shall presently show , no reason for assuming that the coefficient varies with the pressure , and the agreement between CHAPPUIS ' results and our own may be considered as confirmatory of the latter .
The combined results may prol ) ably be considered correct to the fifth significant figure .
ON THE MEASUREMENT OF TEMPERATURE .
24 .
The Pressure Coefficients of Hydrogen Hdium at Lower Pressures .
Though the accuracy of the determinations of the pressure coefficients at lowel ' pressures is reduced in proportion to the initial pressure , the results we have obtained indicate that the pressure coefficient for the two gases is practically indepen ent of the pressure .
The value of the coefficient found for hydrogen at an initial pressure of 500 millims. is , and for helium and ; if the coefficient attained a limiting value corresponding to at zero pressure , the value at 500 millims. should be The determination of the pressure coefficient for hydrogen at an initial pressure of 350 millims. is practically valueless ; only one set of measurements was made , and we found that our method was not sufficiently accurate to investigate the coefficient at such a low pressure .
has determined the pressure coefficients of dine ] ) with the following results:\amp ; mdash ; On plotting the values of the coefficient against the initial pressure , the value of the coefficient at zero initial pressure is found on linear extrapolation to lie between and As DANIEL BERTHELOT ( Comptes Rendus , ' 1898 ) has out , the specific volumes of the common gases at zero pressure , calculated from their densities under normal conditions , and the variation of " " \amp ; ldquo ; with pressure , are the same .
Their densities under the same conditions are proportional to their chemical atomic weights .
If this is the case , they should , at zero pressure , behave as perfect gases , and the temperature scale on a thermometer filled with any gas at very low ) ressure sbould be coincident with the absolute ] of temperature .
ON THE MEASUREMENT 1.35 PART II .
On the Vapour Pressures of Liquid at nperatures below its Boiling\amp ; hyphen ; Point on the Constant\amp ; hyphen ; Volume Hydro Helium MORRIS W. TRAVERS , D. Sc. , of University , London , EORGE SENTER , B. Sc and JAQUEROD , D. CONTENTS .
1 .
Introduction . . . . .
2 .
Previous measurements of the of liquid ) gen . . .
136 3 .
The coefficients of hydrogen helium ) etween 0 and 100 138 4 .
The coefficient of expansion of glass between and C. . . .
138 5 .
Method of experiment . . . . . . . . .
1.38 6 .
The large constant\amp ; hyphen ; volume thermometer . . . . .
1.38 7 .
The smaller constant\amp ; hyphen ; volume thermometers . . . . . .
141 8 .
Calibration of the thermometers . . . . .
14 9 .
Calculation of the resuIts . . . . .
14 10 .
The vapour pressures of liquid oxygen on the constant\amp ; hyphen ; vohnne hydrogon scade 145 11 .
The vapour pressures of oxygen on the \amp ; hyphen ; volume helium .
148 12 .
Treatment of the resuIts . . . . .
151 13 .
Tabulated vapour pressures of liqu oxygen . . .
) 14 .
Discussion of the resnlts . . . . .
152 15 .
The probable accnracy of the results . . . . . .
1 1 .
AFTER determining the pressure coefficients of hydrogen and betweGlt the melting\amp ; hyphen ; point of ice and the boiling\amp ; hyphen ; point of water , and so obtaining the fundamental intervals of our thsrmometer scales , we proceeded to apply our results to the determination of the boiling\amp ; hyphen ; point and vapour pressures of liquid oxygen .
In these experiments we have employed three different thermometers ; the coefficient of expansion of the glass at low temperatures has been specially determined ; every precaution has been taken in preparing the different samples of hydrogen and helium employed ; and the errors due to impurity in the oxygen and to the superheating of the liquid have , we believe , been eliminated .
The ] lowing account contains full details of our experimental method and results . .
G. SENTER , AND DR. A. JAQUEROD ; Tln .
: relIlent to which any importance can be attached is that in 1888 Wiener ichte , Abth . .
ture of \amp ; hyphen ; point of oxygen was measured on the constant\amp ; hyphen ; volume , but details given in the paper , which is entitled " " Die keit des stoffs , \amp ; ldquo ; and was shed after his death .
The given by him is C. SZEWSK I appears to have measured the boiling\amp ; hyphen ; point of oxygen on several ccasions .
In 1806 he published in ture ' ( vol. 54 , p. 377 ) an account of his com\amp ; hyphen ; the and helium thermometers at temperatnres corresponding to ressures below 741 millims. mercury .
By extrapolation the temperature of the boiling\amp ; hyphen ; point on either thermometer appears to be ) paper on the liquefaction of gases he gives as the \amp ; hyphen ; point .
The capacity of the bulb of his thermometer was only 2 cub. centims. ( ' Phil. ( 5 ) , vol. 40 , p. ) also made a large number of measure\amp ; hyphen ; ents .
He made twenty\amp ; hyphen ; five measurements of the temperature corresponding to nillims .
, and gives C. as the mean of the results .
The boiling\amp ; hyphen ; point calculated by extrapolation is C. being per .
worked in OLSZEWSKI'S laboratory , and employed the same instrument the latter .
WITKOWSKI Phil. gives the \amp ; hyphen ; point of oxygen as C. , the value for the coefficient of expansion of and compared the hydrogen and ometers at the temperature of liquid air , and \amp ; hyphen ; und that the readings of the air thermometer were lower than those of the hydrogen thermometer .
Later , HOLBORN ( ' Ann. der Physik , ' ( 4 ) , 1901 , vol. 6 , p. 242 ) dardised a platinum lesistance thermometer by meanls of an air thermometer , and by means of it measured the boiling\amp ; hyphen ; point of pure oxygen .
The temperature referred to the constant\amp ; hyphen ; volume hydrogen scale is C. Full details of the work , which was carried out in the Physikalische Reichsanstalt at , are given in the paper .
NBURG and KRITGEL ( ' Ber .
, ' 1900 , vol. 32 , p. 1818 , vol. 33 , p. 637 ) found for the temperature corresponding to a pressure of 745 millims. ( .boiling\amp ; hyphen ; point .
The measurement was made by means of a thermo\amp ; hyphen ; electric junction , at the tempel.ature of liquid air , which was supposed to be DEWAR , employing constant volume thermometers filled with hydrogen and oxygen , obtained the following values .
The numbers in third value of the following table are calculated from the mean value ON THE MEASUREMENT OF TEMPERATURF , .
Hydrogen Thermometer .
The coefficient of expansion for hydrogen is taken as .
Taking the same coefficient for oxygen , a proceeding which does not appeal to be justified , he obtained the following results:\amp ; mdash ; Oxygen Thermometer .
BALY ( ' Phil. Mag 1900 , vol. 49 , p. 517 ) measured the vapour pressures of oxygen on the constant\amp ; hyphen ; pressure hydrogen scale , employing a Callendar thermometer .
The results appear to be affected by a constant error ; they will be discussed elsewhere .
So far as the evidence contained in the original papers is concerned , it appears to us that greatest weight is to be attached to the work of HOLBOBN and WIEN .
It is true that the results of WROBLH : WSK1 , OLSZEWSKI , and ESTREICHER show a close agreement , but none of these experimentalists furnish sufficient data to make criticism possible .
In all the experiments we have referred to , the boiling\amp ; hyphen ; point or vapour pressure of oxygen has been measured by immersing the thermometer in the liquid and measuring the pressure at which it was boilin .
ESTREICHER and measured the vapour pressures at lower temperatures by connecting the vacuum vessel containing the liquid with an exhaust pump , and making the liquid boil under reduced pressure .
The difficulty of making accul.ate measurements by this method is twofold .
In the first place , it is not easy to liquify a sufficient quantity of pure oxygen ; and in the second place , the liquid oxygen tends to become superheated , and does not boil steadily .
The impurity in the oxygen would most probably be air , which would lower the boiling\amp ; hyphen ; point ; other impurities would , however , have the opposite effect .
In every case it appears to have been assumed that the coefficient of expansion for glass remains constant over the whole range of temperature .
VOL. CC .
\amp ; mdash ; A. .
M. W. TRAyERS , MR. G. SENTER , AND DR. A. JAQUEROD 3 .
The Pressure Coefficients of Hydrogen and Helium between and 10 C. As has already been pointed out , there is no appreciable difference between the pressure coefficient of the two gases , which may be considered as approx mat ng very closely to .
The reciprocal of this number is , and though an error of only is incurred by taking 273 as the temperature of melting ice on the scale of either .
thermometer , we have in calculating our results employed the true coefficient .
4 .
The of Expansion of Glass and C. The coefficient of expansion of the glass was determined in the following manner .
The inner tube of a cylindrical vacuum vessel , which was 30 millims. wide and 1000 millims. long , was graduated for short distances close to its two ends .
The vacuum vessel was enclosed in a water\amp ; hyphen ; jacket , and was set up vertically in front of two telescopes , of short focus , fitted with micrometer eye\amp ; hyphen ; pieces .
The cross wires of the micrometer eye\amp ; hyphen ; pieces were first brought into coincidence with marks at the two ends of the vacuum vessel .
The vessel was then filled with liquid air and the distances through which the marks on the vacuum vessel moved were observed .
The following result for the volume coefficient is calculated from the linear con\amp ; hyphen ; traction of the glass :\amp ; mdash ; Coefficient of expansion to C. ( by weight thermometel ) to C. ( from linear contraction ) , , 5 .
Method of Experiment .
In our experiments we have overcome the difficulty last referred to by immersing our thermometer , together with a bulb in which pure oxygen could be liquefied , and which was connected with a ) manometer , in liquid air or oxygen contained in a vacuum vessel .
The liquid in the vacuum vessel could , if necessary , be made to boil steadily , by passing through it a current of air or hydrogen , and could be maintained at constant temperature by enclosing the vacuum vessel in another vacuum vessel oflarger dimensions also containing liquid air .
Readings of the thermometer and 01 the manometer connected with the bulb containing the pure liquid oxygen were taken simultaneously by two observers .
6 .
The Large Constant\amp ; hyphen ; Volume Thermometer .
The thermometer employed in determining the pressure coefficients of hydrogen and helium between and lo C. has been fully described in the previous memoir ( Part I The same instrument , and three others of smaller dimensions , were ON THE OF ATURE .
employed in these researches , the largel thermometer being only used to determine four points on the pressure curve for liquid oxygen on the hydrogen scale as a check on the measurements made by means of the smaller instruments .
It will be remembered that in the large instrument the mercury column and dead\amp ; hyphen ; space were enclosed in a water\amp ; hyphen ; jacket , and the pressure was measured by means of a scale of special construction , which formed the first surface of the water\amp ; hyphen ; jacket .
In these measurements the ) of the there \amp ; hyphen ; mometer and ut 20 millims. of the were immersed in a liquid , consisting mainly of oxygen , contained in a vacuum vessel holding about 450 cub. centims. , which was enclosed in a larger vacuum vessel containing a little liquid air .
Beside the there \amp ; hyphen ; mometer was placed , as in fig. 1 , a glass tube , in which pure oxygen , obtained by heating potassium permanganate , could be liquefied .
This tube com\amp ; hyphen ; municated with the pump , with the apparatus for generating the oxygen , and with a manometer of the type shown in fig. 2 .
During an experiment a stream of hydrogen was passed through the liquid in the inner vacuum vessel , to prevent superheating , and to stir \amp ; iacute ; t thoroughly .
In calculating the results it was possible to con\amp ; hyphen ; cider the whole of the stem from the top of the bulb to the level of the top of the vacuum vessel as at the temperature ofliquid air .
The portion not immersed in the liquid was only about 30 millims. long , and the error so introduced would be considerably less than C. The coefficient of expansion of the glass between the \amp ; hyphen ; point of water and the temperature ofliquid air was taken as , a number which we obtained by actual experiment ( p. 138 ) .
The thermometer was filled with pure hydrogen by the method described the second appendix to Part III .
of this paper .
The freezing\amp ; hyphen ; point of water on the constant\amp ; hyphen ; volume hydrogen scale is taken as , the reciprocal of the pressure coefficient The formula employed in calculating the results is as follows:\amp ; mdash ; where is the pressure on the gas when the bulb is surrounded with melting ice ; is the pressure on the gas when the bulb is at the temperature to be measured ; V is the volume of the bulb to ( fig. 1 ) , at C. ; is the volume of the dead\amp ; hyphen ; space and the portion of the stem within the water\amp ; hyphen ; jacket ; is the volume of the remaining portion of the stem ; is the coefficient of expansion of glass ; is the temperature to be measured ; and are the temperatures of the stem and dead\amp ; hyphen ; space respectively in degrees Centigrade .
The left\amp ; hyphen ; hand side of the equation the of the constant for the thermometer .
The Ice\amp ; hyphen ; Point .
Volume of ) ) at C. . . . . .
94 2096 cub. centims. , , dead\amp ; hyphen ; space . . . . . .
, , stem . . . . . .
, , , Pressure on gas corrected to C. . . . .
millims. Temperature of dead\amp ; hyphen ; spacc . . . . .
ll C. , , stem . . . . .
C. Constant . . . . . . . . .
Vapour Pressures of Liquid ) xygen .
Volume of ) ) . . . . . . . . .
93 8336 .
The figures in the last columns are obtained from the smoothed results of the measurements of the vapour pressure of oxygen obtained by means of the small thermometer .
There appears to be a constant difference of C. In discussing the accuracy of the result it is necessary to point out first of all that it is extremely difficult to maintain a large mass of liquid air or oxygen in a steady state of ebullition ; and for this reason it is probable that the results obtained by means of the smaller thermometers are the more accurate .
In the case of liquid air we have to deal not only with the difficulty which arises from the superheating of the liquid , but , as the liquid evaporates , its temperature rises rapidly .
Liquid oxygen is not easy to obtain in large quantities , and it is extremely difficult to make it boil steadily ; it may , in fact , become superheated to the extent of a whole degree Centigrade .
ON THE MEASUREMENT OF TEMPERATURE .
In the determination of the ice\amp ; hyphen ; point the pressure could be measured without difficulty to millim. , the temperature of the dead\amp ; hyphen ; space to C. ; as has been pointed out in Part I. of this paper , the accuracy of this measurement was of the order of 1 part in 20,000 .
The errors are thus confined to the measurement of the temperature of the liquid oxygen .
In part of the experiment the pressure could not be measured with tainty to nullim .
, though the error certainly never exceeded millim. , which is equivalent to C. Such errors as exist , must be attributed to the unequal heating of thermometel bulb , and of the bulb containing the liquid oxygen .
7 .
The \amp ; hyphen ; Volume Thermometers .
As has already been stated , the use of a thermometer with a large bulb for the measurement of low temperatures is open to serious objections .
In investigating the whole range of vapour pressure of liquid oxygen and of liquid hydrogen , three instruments were used with bulbs of capacities corresponding to 12 , 26 and 27 cub. centims. These instruments ( fig. 3 ) did not differ in any important } ) articularfi.om the large thermometer already described , but as it not necessary to determine the pressure with a degree of accuracy greater than millim. , the manometer column , which was 7 millims. in diameter , was not jacketed with water , and the distance between the surface of the mercury in and in the dead\amp ; hyphen ; space was read directly on a glass scale , placed behind the instrument , by means of two telescopes ; the readings were corrected to C. The temperature of the dead\amp ; hyphen ; space , and of the stem as far as , was measured by means of a thermometer placed close beside .
The temperature of the manometer column was taken at the mean between the temperature of the dead\amp ; hyphen ; space and the temperature indicated by a thermometer placed at the level of the surface of the mel.cury in the tube The thermometer bulb , which had a capacity of 12 , 26 or 27 cub. centims. in the three thermometers , was sealed to the stem as indicated in the figure , so that if any mercury were accidentally introduced into the bulb , it could be completely removed from it by simply opening the stop\amp ; hyphen ; cock and lowering the resel'voir con\amp ; hyphen ; nected with it .
In making a measurement of temperature by means of constant\amp ; hyphen ; volume gas thermometer , the temperature of the bulb and of the dead\amp ; hyphen ; space are to be constant ; and when the latter is measured by means of a mercury thermometer placed beside it , the former can be calculated from the pressure exerted by the gas .
The temperature of the vertical portion of the stem is , however , uncertain , and is usualy taken to be the mean of the extreme temperatures .
In measuring very low temperatures , the density of the gas in the dead\amp ; hyphen ; space becomes so low that its mass becomes nearly negligible , while the amount of gas in the stem becomes sufficiently ON MEASUREMENT OF TEMPERATURE .
To measure the temperature of this portion of the stem we employed the device which is shown in the figure .
A second thermometer similar to the one already described , with a cylindrical bulb of the same length as the portion of the stem , was mounted opposite the luain thermometer .
From the readings of this auxiliary thermometer , taking the temperature of the dead\amp ; hyphen ; space and mercury column to be the same in the case of the main thermometer , the mean temperature of the stem could.be calculated with sufficient accuracy for our purpose .
This device has already been applied by other investigators to similar measurements .
The stem of the main and auxiliary thermometers passed at through a rubber stopper not shown in the figure .
A tube , terminating in a bulb at the lower end , also passed through the rubber stopper , and communicated with a manometer of the type shown in fig. 2 , and with an apparatus generating pure oxygen .
The volume of the bulb and stem to the point , of the stem from to , and of the remainder of the stem and dead\amp ; hyphen ; space , determined by the methods described in Part I. of this paper .
As in the previous case , the stem was divided in millimetres between and 8 .
libration of the hermometers .
Calibration of the stem.\amp ; mdash ; The volume of one division of the stem was determined before the bulb was sealed to it , by introducing into it a thread of mercury which was measured and subsequently weighed .
Calibration of the dead\amp ; hyphen ; space.\amp ; mdash ; The volume of the dead\amp ; hyphen ; space to the zero point on the stem was determined by the method descl.ibed in Part I. of this paper .
A capillary stop\amp ; hyphen ; cock was sealed to the tube which joined the lower part of Mercury was introduced through the stop\amp ; hyphen ; cock by means of a rubber tube and mercury reservoir , till it was brought into contact with the point .
The stop\amp ; hyphen ; cock was then closed , the rubber tube removed , and the mercury run out of the apparatus and weighed .
The rubber tube was then re\amp ; hyphen ; attached , and , by raising the reservoir , mercury was introduced so as to fill the dead\amp ; hyphen ; space and to the zero point on the scale .
This quantity of mercury was weighed as before .
Volumes of the Bulbs .
Volumes of the Stems .
A. millims. of stem , .
2865 gramme of mercnry ; 1 millim. ' 00319 gramme 1 millim. , , cub. centim. BuII ) B. 168 8 millims. , , .
65529 millim. 1 millim. , , .
000286 cub. centim. Bulb C. The stem was of the same diameter as that of bulb B. Volume of the Dead\amp ; hyphen ; Space .
In calculating the results , the thermometers were considered as divided into three sections .
The Volume of the Auxiliary Thermometer .
The volume of the bulb and dead\amp ; hyphen ; space were determined by the methods already referred to .
As great accuracy was not required , details need not be given .
Volume of bulb to ( fig. 3 ) . . .
2.124 cub. centims. Volume of dead\amp ; hyphen ; space and stem .
ON THE MEASUREMENT OF TEMPERATURE .
9 .
Calculation of the Results .
The results were calculated by means of the formula : Where is the pressure on the gas when the bulb and stem were immersed in ice , is the pressure on the gas when the bulb is at the tempel.ature to be measured , V is the volume of the bulb at C. , V , is the volume of the stem from to is the volume of the dead\amp ; hyphen ; space , is the coefficient of expansion of the glass , is the melting\amp ; hyphen ; point of ice on the gas scale , the reciprocal of the pressure coefficient of the gas with which the thermometer is filled , is the temperature to be measured , and are the temperatures of the dead\amp ; hyphen ; space in degrees Centigrade , is the temperature of the stem calculated from the readings of the auxiliary thermometer .
In calculating , the following formula was employed : Since for each series of vations the ice\amp ; hyphen ; point remained constant , the left\amp ; hyphen ; hand side of each equation could be expressed by a constant for each filling of the thermometer .
10 .
Pressures of Liquid Oxygen on the Constant\amp ; hyphen ; Volnme Hydrogc Scale .
March , 1901 .
Thermometer ( cub. centims. bulb ) .
Ice\amp ; hyphen ; Poiut of Thermometer .
VOL. CC.\amp ; mdash ; A. The constant for the auxiliary thermometer was in this experiment determined by measuring the pressure when the bulb and dead\amp ; hyphen ; space were at the same Vapour Pressures of Liquid Oxygen .
( II .
) Thermometer , refilled with hydrogen .
Ice\amp ; hyphen ; Point , March 19th , 1902 .
DR. M. W. TRAVERS , MR. G. SENTER , AND Db ' .
A. JAQUEROD ( III .
) Thermometer ( cub. centims. bulb ) , filled with hydrogen , June 2nd .
Ice\amp ; hyphen ; Point of Thermometer .
Vapour Pressure of Liquid Oxygen .
June 3rd , 1902 .
This result is the mean of two very concordant observations .
11 .
The Pressures of Oxygen on the \amp ; hyphen ; Volume Helium Scale .
( I. ) Thermometer ( cub. centims. bulb ) .
The helium , which had been by heating cl\amp ; egrave ; veite , was purified by sparking with oxygen , and was afterwards passed through liquid oxygen condensed in a bulb and cooled to below C. The gas was passed through a bulb immersed in liquid air on its way into the thermometer .
Ice\amp ; hyphen ; Point of the Thermometer .
The auxiliary thermometer was still filled with hydrogen .
ON THE MEASUREMENT OF TEMPERATURE .
Vapour Pressure of Liquid Oxygen .
( II .
) Thermometer A. The thermometers were filled with helium purified by passing it through a coil immersed in liquid hydrogen at its boiling\amp ; hyphen ; point ( Appendix III Ice\amp ; hyphen ; Point of Thermometers , April 17th and March 25th , 1902 . .
Pressure of Liquid Oxygen .
( II1 .
) Thermometer ( cub. centims. bulb ) .
The thermometers were filled helium purified by passing it through a coil cooled to by means of liquid hydrogen boiling under reduced pressure ( Appendix III Ice\amp ; hyphen ; Point of Thermometer , June 4th , 1902 .
Vapour Pressure of Liquid Oxgen ( IV .
) Thermometer ( cub. centims. bulb ) .
The helium was from the same sample as was used in the last experiment .
Ice\amp ; hyphen ; Point of Thermometer , June 11th , 1902 .
The auxiliary thermometer was not refilled .
ON THE MEASUREM]]]NT OF TEMPERATURE .
Vapour Pressure of Liquid Oxygen .
this experiment the vacuum vessel surrounding the thermometer bulb , \amp ; c. , was first of all filled with nearly pure liquid oxygen .
The liquid made to boil steadily by passing a current of gaseous oxygen into the liquid .
Liquid air was added to the liquid oxygen to obtain the lower temperature .
Between the first and second observations the manometer and bulb contaming the pure oxygen were exhausted and refilled .
12 .
Treatment of the ldesults .
The figures given in the last column of the foregoing tables are the temperatures , corresponding to the observed pressures , taken from the smoothed vapour pressure curves shown in Plate 1 .
It will be observed that the points obtained by direct qervation lie in every case very close to the curve , and that the difference between the observed and calculated temperatures rarely exceeds two or three hundredths of a degree .
The results obtained by direct observation of the vapour pressures were smoothed by the method of RAMSAy and Phil. NJag .
This method consists in calculating the ratios of the absolute , \amp ; c. , for any pair of substances A and corresponding to vapour pressures \amp ; c. , and plotting the ratios \amp ; c. , against the temperature , as rectangular coordinates .
The points so defined lie on a straight line , from which the temperature corresponding to any pressure for the substance can be calculated , by first finding the temperature corresponding to that vapour pressure for the substance , determining the value of the ratio dividing the value of by it .
In smoothing our results we took water as the second substance A. Two sets of ratios were obtained by this method , corresponding to the tempera\amp ; hyphen ; tures of liquid oxygen on the hydrogen and helium scales .
When the results were plotted against the absolute temperatures of water two parallel straight lines were obtained .
From these lines the smoothed values of the ratios corresponding to those temperatures at which the vapour pressures of water have the values expressed in the following table were determined , and from them the corresponding temperatures of liquid oxygen on the scale of the two thermometers were calculated ( see Plate 1 ) .
13 .
Vapour Pressures of Liquid Oxygen .
14 .
Discussion of the Results .
When the results which are tabulated in the preceeding .
table are plotted on a diagl.am , it will be observed that the vapour pressures of liquid oxygen on the scale of the two thermometers are expressed by two curves , and that the temperature corresponding to any particular pressure is always higher on the helium scale than on the hydrogen scale .
As we shall show later ( Part IIL , p. 169 ) , the divergence becomes still greater at lower temperatures .
Though the pressure coefficients for hydrogen and helium between the melting\amp ; hyphen ; pomt of ice and the boiling\amp ; hyphen ; point of water do not appear to differ appreciably , and though at the normal temperature these gases ma , be considered as nearly perfect , it ON THE MEASUREMENT OF TEMPERATURE .
is not surprising that they exhibit a difference at lower temperatures .
It must be remembered that while the critical point of hydrogen is about ) , it is probable ( Part III .
, p. 177 ) that the critical point of helium does lie from .
The temperature of liquid air , about ) ressed as a of critical tem\amp ; hyphen ; perature of the gas , is 8 with regard to heliunl , and with regard to hydrogen , which is at that temperature in a state corresponding to that of oxygen , of the critical point is , at the boiling point of water .
At this we have no reason to believe that oxygen behaves in any way as a perfect gas .
Further , as was first shown by WROBLEWSKI ( loc. cit. , p. 136 ) , and latel by two of us SENTER , Brit. Assoc , the coefficient of expansion of hydrogen at constant pressure between the normal temperature and the temperature ' ofliquid air increases rapidly with rise of pressure .
This tends to confirm the result which we now bring forward , that nitial pessure of 1 metr of mercury at the melting\amp ; hyphen ; point , the constant volume hydrogen a2 by at the boiling\amp ; hyphen ; point of liquid .
At lower initial pressures the the two scales of temperature might possibly be smaller .
15 .
The Probable Accura ) esnlts .
It may at once be pointed out that experiments of this kind the probable accuracy cannot be arrived at by any lnethod of calculation .
So far as the actual measurements of temperatures and pressures are concerned , they may easily be made to 1 in 10,000 .
In discussing the actual experiments we have , however , to consider such sources error as may be due to the superheating of the liquid of which the vapour pressure is being determined ; ities in the temperature of the liquid in which the thermometer was immel.sed ; and errors due to the liquids employed .
It will be observed that in the course of our work four rent thermometers , the large thermometer used for determining the pressure coefficients of the gases betweel and \amp ; dollar ; three smaller instruments ) which the bulbs had capacities of 12 ) .
centims. ( A ) , 26 cub. centimH .
( B ) , and cub. centims. ( C ) , respectively .
Though the ] ) el.lnometel.sl } give the most accurate results , and , indeed , differentiation of the equations } ) loyed in culating the results indicates that the measurement of a steady temperature of about .
should be accurate to , it appears to be impossible to maintain so large a thermometer bulb at a uniform and steady tnre .
We have considered that the results obtained means of the } thermometers , which are very concordant among themselves , and diffe ] .
only by results obtained by means of the large thermometer , are the most accurate .
The values of the boiling\amp ; hyphen ; point and vapour pressures of liquid hydrogen deterlnined by of 1 . .
I. W. .
G. , AND DR. A. JAQUEROD the sanlethel.mon meters show an even closer agreement ; this , as we shall presently show , is due to the fact that liquid hydrogen , unlike liquid oxygen , does not tend to become superheated .
It is somewhat remarkable fact that pure oxygen when liquified can only with difficulty ) made to boil steadily .
By passing a rapid current of oxygen or through the liquid a fairly steady temperatul'e may be maintained , but if the current is the temperature may more than one degree .
This probably accounts the fact that the temperatures found by us are in case somewhat lower than those of other observers , who , without exception , measured the pressure on the mass of liquid in which the thermometer was immersed .
This source of error we have , however .
great pains to eliminate ; for the accuracy of the actual surelIlents , the concordance of the observation made by means of the different the moJneters furnishes a sufficient guarantee .
ON THI ASURLM OF I IIL On the Vapour Pressures of Liquid Hydrogen at below its BoilicPoint on the Constant\amp ; hyphen ; Volume and MORRIS W. , D.Sc .
, Fellow of iversity College , ADRIEN JAQLTEPOD , D. CONTENTS .
1 .
Previous investigations . . . . . . . .
155 2 .
The liquefaction of hytlroge1l . . . . .
157 3 .
The mttnipulation of liqnid hydrogelt . . . . .
4 , Method of determining the vapour ) ressnres of ] geIl . . . . .
5 .
Calculation of the results . . . . .
16 6 .
The vapour pressures of liquid hydrogen on the constant\amp ; hyphen ; volume hydrogen scale 7 .
The vapour of liqnid hydrogen on the helium scale 164 8 .
Treatment of the resnlts\amp ; mdash ; their ) . . . . .
167 9 .
Tabulated vapour pressures of liquid hydrogen . . . . . .
169 10 .
Discussion of the results . . . . .
169 APPENDIX I.\amp ; mdash ; On the melting\amp ; hyphen ; point of hydlogen . . . . .
170 II.\amp ; mdash ; On the prepa ] ation of pnre hydrogen . . .
171 III.\amp ; mdash ; On the preparation of helium . . . . .
\amp ; mdash ; Note on the vapour pressures of solid neon . . . . .
.\amp ; mdash ; On the probable vnlues of the criticnl boiling\amp ; hyphen ; points of helinm , and on an attempt to liquefy . . . . .
1 .
THE of liquid hydrogen has been measured directly by indirectly by OLSZEWSKI .
The former employed constant\amp ; hyphen ; volume hydrogen and helium thermometers , the lattel , however , containing a mixture of helium and neon , obtained by a method which will be discussed later .
) .
The results il ] as follows Proc .
M. W. TRAVERS , MR. G. SENTFR , AND DR. A. JAQUEROD .
liquid \amp ; hyphen ; ygell was measured in every case a check on the results .
The results obtained by means of the hydrogen thermometer filled under a pressure of 739 lnillims .
of mercury differ from the mean of our observations by .
The diifel'ence between this result and those obtained in experiments , and ( c ) is certainly due to experimental error , since the temperatures determined by means of should rise when the pressure on the gas in the thermometer is reduced .
Since the on the gas in the thermometer was in each of these iments very lnall , the experimental was proportionally , and much ' weight must be attached to the last measurement .
The results obtained by means of the helium thel.mometer are lower than those ined by us , though the difference is not constant .
OLSZEWSKI ( ' Phil. Mag. , ' 1895 ( 5 ) , vol. 39 , p. 199 ; vol. 40 , p. 202 ) employed the following method : Hydrogen was compressed to 170 atmospheres into a steel cylindel , with a loose glass lining enclosing the platinum coil of a resistance thermo\amp ; hyphen ; meter , of which the resistance had been determined at and C. The cylinder was cooled in liquid air , boiling ?
, and the gas was then to escape from the cylinder .
When the pressure fell to one atmosphere the resistance the coil was measured , the temperature determined by linear extrapolation .
As the esult of numerous trials , he found the temperature to be Though the method gave accurate results for the boiling\amp ; hyphen ; point oxygen , the results are of no value so far as the temperature of liquid hydrogen is concerned , as the variation of the esistance of platinum with the temperature is not linear .
Beyond determining the boiling\amp ; hyphen ; point of liquid hydrogen on the scales of the two thermometers , DEWAR Proc finds the melting\amp ; hyphen ; point on the helium scale to be ; an earlier determination by him gives the melting pressure as 55 millims. find millims. respectively , He also gives the following values the temperatures of solid hydrogen corresponding to very low ( loc. cit.):\amp ; mdash ; ON THE OF TEMPERATURE .
1 b7 As has already b\amp ; egrave ; en pointed out ( p. 1561 , the pressure on the in thermo\amp ; hyphen ; meters was very small , and little weight can be attached to the results .
2 .
The of The liquid hydrogen employed in these researches was obtained by means an apparatus devised and constructed by one of us two years ago .
The apparatus as i was originally designed , has been described in the 'Philosophical Magazine ' ( 1901 , vol. 17 , 411 ) , and such slight modifications have since been introduced are referred to in the treatise , TRAVERS on ' The Experimental tudy of Gases , ' p. 206 .
The hydrogen was obtained by the action of dilute olllnercial granulated zinc , contained in a lead vessel , of which the joints had soldered in the oxyhydrogell .
The by sing it first through a solution of potassium permauganate ; ough two glass towers , each metres in length , containing pumice ovel which acid solution of potassium omate 1 continuously ; a third glass tower containing pumice kept moist with silver solution ; and finally , through a wash bottle containing caustic potash tion , into the gasometel .
In this it was possible to obtain hydrogen from which but little solid } during process of liquefaction .
* In one experiment in which we failed to obtain liquid hydrogen , owil to the formation of solid air in the vacnum vessel closing the coil , we found that the packin of the low\amp ; hyphen ; pressure ) of the comple ; sor become ) worn that air was being taken into the cylinder with the hydrogen .
With compressors of the " " Whitehead\amp ; ldquo ; type , in which the piston\amp ; hyphen ; rod does not pass through a gland , and the compression is only affected by the inward motion of the piston , special care must be taken that the piston packing or cup\amp ; hyphen ; fibre fits the cylindel In connection with the work on liquid hydrogen , we wish to express thanks to Mr. HOLDING , Mechanic in the Chemical Department of University College , for his invaluable services .
* In connection with the preparation of tho hydrogen , ious observutirj1l was mnde which has not been explained .
The silver nitrate solution which escaped at the ) ottom of the tower appeared almost immediately to deposit pure silver in the form of small bright metallic .
Ve ry little silver appeared to form inside the tower , and it seems as if some compound were produeed whioh decomposed in contact with the air .
The hydrogen had already passed through two towers containing acid chromate solution , and it is possible that hydrogen , like oxygen , becomes more ctive when it takes part in a chemical reaction .
DR. W. TRAyERS , MR. G. NTER , AND .
A. JAQUEROD Manipulation of Liquid Hydrogen .
In each experiment we employed about 400 centims. of liquid hydrogen , which , when sufficient liquid litres ) had been tuccumulated to cool the apparatus , could be obtained in less than 1 hour the time commencing the operations .
Though perhaps it is little to be expected , it is a fact that it is much easier to measnrements of the \amp ; hyphen ; point and vapour ) ressures of liquid hydrogen than of liquid oxygen .
As we have already stated ( p. 140 ) , liquid air tends to become supelheated , and only with difficulty can it be made to boil steadily .
Liquid hydrogen , on the other hand , can be made to boil quite steadily at any temperature between its boiling\amp ; hyphen ; point and melting\amp ; hyphen ; point .
Possibly this may be due to the resence of finely divided particles of solid air suspended in it , for it does not appear that the ) ] leuomenon of superheating is ) endent upon any intrinsic property of a liquid .
the of late1lt heat of tion is very high , 'Roy .
Soc. ' June , vol. 68 , ) is also in vonr of the experimenter ; for as the boiling\amp ; hyphen ; point of ogen is sixty degrees lower than the tempel.ature of liquid , the operation of cooling thermometer bulb , previously cooled in liquid , to the boiling\amp ; hyphen ; point of liquid hydrogen , necessitates the absorption of a considerable quantity of heat .
When the vessel which eceived the liquid the liquefier was filled with the liquid it was at once emoved from the apparatus , plugged with animal wool , enclosed in an olltel vessel containing liquid air .
The mouth of a vessel containing liquid hydrogen cannot } ) left open , in that case air enters the vessel , solidifies , and rapidly evapol.ates the liquid .
A plug of natural wool is preferable to a cork or rubber stopper , for the intel.stices of wool become filled with cold vapour , and the plug acts as au excellent insulatol , whereas the heat radiated from the stoppel helps to evaporate the liquid .
4 .
of the of The thermometers employed in these researches were identical with those used in determining the essures of liquid oxygen , and in nearly every case in which Jneasurelllent was made with liquid hydroge ] , a ) oint on the vapour\amp ; hyphen ; pressure curve of oxygen was also determined as a check upon the results .
The apparatus as arranged for these experiments is shown in section in fig. 1 .
The thermometer has already ) fully described in Part .
II .
of this vestigation , so that to simplify the description only the bulb of the main thel.mometer , and the bulb of the auxiliary thermometer , by means of which the temperature of the stem of the main thermometer was determined , are shown in the present diagram .
The stem of the main C'Jnd auxiliary thermometers passed through two holes in a rubber stopper , .
Through one of two holes , pierced at right angles to the first , the stem of the bulb which contained pure hydrogen ( pp. 138 , 139 ) and commu\amp ; hyphen ; nicated with a manometer .
The capillary tube , , which was sealed at its lower end , communicated with a compression apparatus containing helium , passed through the other hole .
This tube was employed in an attempt to liquefy helium , which will be dealt with later .
In order to make the diagram clearer , the glass walls and stem of the thermometer are ] acked in , the compression tube is drawn in outline , and the tube and bulb for the pure liquid hydrogen is represented by the dotted line .
The thermolneter bull ) , \amp ; c. , was enclosed within a wide glass ) , 75 millims. in diameter and 400 millinls .
long .
This tube was open at the bottom for the intro\amp ; hyphen ; duction of the vessel containing the liquid hydrogen , and at the top was formed into a ]lPck to fit th( rubber ) .
Surrounding this tube for a considerable part of its length was an annular zinc vessel , open at the top , which could be filled with liquid .
This vessel was in surrounded with natural wool enclosed within ajacket of linoleum .
The diameter of the linoleum jacket was such that it exactly filled the space between the vertical wooden supports to which the auxiliary thermometers fixed ( p. 141 ) .
The liquid\amp ; hyphen ; air jacket was not essential to the working of the apparatus , even when the liquid hydrogen in the vacuum vesse ] was boiling under reduced at the extremely low temperature of abs .
It was thought that by using it an economy in the liquid hydrogen would be cted , but it is impossible to determine whether this was actually the case .
Wool was packed into the of the tube , , and between the tubes which projected downwards through .
the rubbel stopper , to shield off radiation from above .
The vacuum vessel containing the liquid inserted from the lower and open end of the tube , a to surround the thermometer bulb , and was kept in place by the arrangement shown in the figure .
A stopper , , made of boxwood , was turned on the lathe to fit the open end of the tube , , and air\amp ; hyphen ; tight junction between the two was made by lleans of a rubber sleeve , .
The sleeve was fixed permanently to the , and the part of it could be olled back , and then slipped over the tube , , when the ) was in position .
A brass tube , , passing through the stopl ) , could be connected with the exhaust pump by means of a rubber connection and lead tube ; as the brass tube was soldered to a heavy brass plate , which was screwed to the outer surface of the stopper , it never became cold , and a good iunction with the rubber tubs was maintained during the experiment .
It may be remarked here that in carrying out experiments at low temperatures , it is impossible to make air\amp ; hyphen ; tight junctions between frozen rubber and metal on account of the greater contraction of the latter .
The vessel was supported on four wires , , let into the wooden stopper and covered at the top with pads of cloth .
Wire rings were soldered to them at ON THE MEASUREMENT OF TEMPERATURE .
intervals , and round the cage so formed baize , , was wound to form a plug a little smaller than the inner diameter of the tube , .
A glass tube , , served as a continuation of the tube , , and the space between and was packed with animal wool .
By means of this plug the vacuum vessel was effectually shielded from radiation from below .
The method of filling the thermometer bulb has already been dealt with in a previous section of the work ( p. 117 ) .
The ice point of the thermometer was of course determined before the tube , , was placed in position .
Before commencing an experiment with liquid hydrogen a vacuum vessel containing liquid air was first introduced into the tube , , and temporarily held in position by means of the plug ; in this manner the bulb of the thermometer was cooled down to \amp ; mdash ; 185o C. When the vessel containing the liquid air was again lowered , the wool was removed the mouth of the vacuum vessel containing the hydrogen , which was brought below the apparatus , and slowly raised into position by means of the plug .
The rubber sleeve , , was then turned upwards over the tube , , and secured in position by means of a piece of wire .
The actual measurement of the vapour pressure was made in the same manner as has been described in the case of oxygen ( p. 138 ) .
One observer read the pressures indicated by the manometer connected with the tube , , in which pure hydrogen was liquefied , while the other took simultaneous readings of the main thermometer , of the auxiliary thermometer , and of the temperatul.e of the dead space .
So steadily did the hydrogen boil under normal pressure that this operation could be carried out without the least difficulty .
It is noticeable that the vapour pressure of the pure hydrogen in the bulb , , was always slightly higher than the barometric pressure .
This is to be attributed to the presence of small quantities of impurities dissolved in the main quantity of hydrogen in the vacuum vessel , and confirms our reasons for adopting this method of measuring the vapour pressures .
In determining the vapour pressure of liquid hydrogen at lower employed the following method : The tube , , was connected by a short rubber tube to a wide lead tube which communicated the exhaust pump , with a large glass \amp ; hyphen ; globe 300 millims. in diameter , with a mercury manometer , and with a fine adjust\amp ; hyphen ; meant stopcock , through which air could be admitted to the apparatus and the pressure so regulated .
The exhaust was maintained by means of a large two\amp ; hyphen ; cylinder 'Fleuss ' pump , worked by a one horse\amp ; hyphen ; power electric motor .
This pump , which was also used in connection with the apparatus for producing liquid hydrogen , was arranged ) that the two cylinders could be worked in parallel to vacuum of 50 millims. of mercury , or in series to give a vacuum ofless than 1 millim. of mercury .
For reasons which will be entered into presently we have confined our measurements to above millims. of mercury .
By opening and closing the fine adjustment cock , so as to admit more or less air into the apparatus , the pressure could be rnnintained VOL. CC.\amp ; mdash ; A. steady for a length of time amply ient for the equalisation of temperature throughout that part of the apparatus which was enclosed within the vacuum vessel .
5 .
Calculateon of the Results .
The method of calculating the temperatures of liquid hydrogen was the same as has been described in Part IL of this work .
It may be pointed out here that , at very low temperatures , the mass of gas in the stem of the thermometer becomes con\amp ; hyphen ; siderable , while the mass of gas in the dead space becomes nearly negligible .
Taking a specific case : Constant for thermometer Pressure of gas in thermometer .
Volume of bulb .
stem , , , , dead\amp ; hyphen ; space , , Temperature of stem , , dead\amp ; hyphen ; space 44.769 55 .
millims. 0.035 abs .
The equatio11 for determining the ture T now becomes:\amp ; mdash ; 44 .
where the second term within the bracket is equal to , and the third term is less than .
In both cases the correction is exceedingly small .
In calculating the volume of the bulb , the coefficient of expansion of glass between and 25 C. was taken as for the following reasons : coefficient of expansion between and C. is , between and it is ; on extrapolation one obtains the value which we have taken for the coefficient .
The following values for the volume of the bulb at are calculated for different coefficients:\amp ; mdash ; This difference o.f volume corresponds to a difference of in the boiling\amp ; hyphen ; point of hydrogen .
\amp ; hyphen ; ON .
THE fEASU .
OF TEMPERATURE .
Vapour Pressures of Liquid Hydrogen .
June 12th , 1902 .
For the measurement of the vapour pressures of liquid oxygen with thermometer , see Part II .
, 11 ( IV .
) .
8 .
of the Results.\amp ; mdash ; Their Accnra As we shall presently find , the melting point of hydrogen is on the helium scale , and consequently the whole of the observations recorded in the egoing tables refer to liquid hydrogen only .
As in the case of liquid oxygen Part , the experimental results were smoothed by the method of RAMSAY and YOUNG , and the vapour\amp ; hyphen ; pressure curves for liquid hydrogen on the scales of the hydrogen and helium thermometers were plotted on a diagram as in ] .
The last columns of the preceding tables contain the temperatures read off the at points corresponding to the observed pressures .
It will be observed that at pressures near 760 millims. the points representing actual observation lie either on , or very close to , the curve .
Of these points there are four on the curve representing the vapour pressures on the helium scale , obtained by means of three separate thermometers ; there are two similar points on the curve represefiting the vapour pressures on the hydrogen scale , the results of observations with two different thermometers .
We can therefore consider that the boiling\amp ; hyphen ; point ofliquid hydrogen , which is on the hydrogen scale and on the helium scale , has been determined with a high degree of accuracy .
*The mean of two nearly coincident valions .
In these observations the vapour pressures of hydrogen were measured on two manometers , connected with two bulbs containing pure liquid hydrogeIl , and immersed in the vacuum vessel surrounding the thermometer buib to the depth of the top and the bottom of latter respectively .
The two sets of readings , of which the mean is given above , did not diRer by more than 1 millim. , indicating that the temperature of the thermometer bulb was uniform .
The following table shows the order of th of temperatures near the boiling\amp ; hyphen ; point .
With regard to the measurements at lower pressures , the greatest difference between the observed and calculated temperatures in the first series of measurements with the helium thermometer exceeds .
These errors can be accounted for by supposing that when the liquid hydrogen surrounding the thermometer bulb was made to boil under reduced pressure , the observations were made before the tempera\amp ; hyphen ; ture of the thermometer bulb , and of the bulb containing the pure liquefied hydrogen , had become steady .
At pressures below 200 millims. it appeared to be much more easy to maintain a steady temperature than at intermediate pressures .
In the last series of measure\amp ; hyphen ; ments , as has already been stated , bulbs containing pure liquid hydrogen and connected with two manometers .
were immersed at different levels in the vacuum vessel containing the liquid hydrogen surrounding the thermometer bulb , to the depth of the top and of the bottom of the thermometer bulb respectively .
The pressures indicated by the manometers were read simultaneously and were never found to differ by more than 1 millim. , proving that the temperature of the thermometer bulb was practically uniform .
In concluding these remarks on the accuracy of the results , it may be well to poin out that the expression of differences oftemperatures in degrees , particularly at low tem\amp ; hyphen ; peratures , in degrees , is somewhat apt to lead to a false conchision .
In dealing with the thermodynamic cycle , on which our idea of an absolute scale of temperature is APPENDIX I.\amp ; mdash ; On the Melting\amp ; hyphen ; Point of Bydrogen .
DEWAP .
( ' Roy .
Soc. Proc 1891 , vol. 68 , p. states that the melting\amp ; hyphen ; point of hydrogen on the helium scale is .
The method by which the measurement was made has not been recorded ; he had previously given the melting pressure as 55 millims. ( see p. 106 ) .
Our experiments of the vapour pressures of hydrogen appeared to indicate that the melting\amp ; hyphen ; point lay considerably below the temperature given by DEWAR , for not only did the numerical results bear evidence that at temperatures close to we were still dealing with liquid , but when after main\amp ; hyphen ; taining a pressure of 55 millims. in the apparatus for some time the vacuum vessel was found to contain liquid only .
We therefore decided to carry out a separate experiment with a view to determining the exact melting\amp ; hyphen ; point .
The apparatus employed in this experiment is shown in fig. 2 .
The tube , which was 6 lims .
in diameter , and 150 millims. below the side tube , passed through the rubber stopper of the apparatus shown in fig. 1 , so that the lower end of it lay at the level of the middle of the thermometer bulb .
The tube contained a very thin glass rod , , bent at its lower end into a hook , and joined at the top to a piece of soft iron wire wound into a solid spiral , .
Round the tube above the side tube a coil of insulated wire was wound forming a solenoid ; by passing a current through the coil , the rod with its iron head could be given a vertical motion of about 3 millims. The side tube communicated with a mercury reservoir con\amp ; hyphen ; taining pure hydrogen from palladium , with a manometer of the type described in Part II .
( fig. 2 ) , and through a stop\amp ; hyphen ; cock with the mercury pump .
In carrying out the experiment the lower part of the tube and the thermometer were cooled in liquid hydrogen boiling under reduced pressure in the manner described on p. 160 , and about 100 cub. centims. of gaseous hydrogen was allowed to enter and liquefy in it .
At all temperatures down to , which corresponded to the lowest pressure which we were able to obtain under the conditions of the experiment , the hydrogen appeared to remain liquid , for , on passing the current through the solenoid , the piece of iron was seen to move upwards .
Maintaining the temperature outside at from to , the stop\amp ; hyphen ; cock connecting the side tube with the mercury pump was opened for a moment .
The pressure in the apparatus immediately fell to 49 millims. , and remained steady there ON THE MEASUREMENT while the stop\amp ; hyphen ; cock was open ; on passing the current through the solenoid the piece of iron did not move , indicating that its lower end was embedded in solid hydrogen .
The stop\amp ; hyphen ; cock leading to the pump was then closed , and when the hydrogen in the tube had melted it was again opened and the melting pressure was again determined .
This operation was repeated several times ; the melting pressure was invariably found to be 49 or 50 millims. This pressure corresponds to a temperature of on the helium scale , which is the melting\amp ; hyphen ; point of hydrogen .
APPENDIX II.\amp ; mdash ; Note on the Prep of Pnre It is obvious that the simplest method for the preparation of pure dry vould be to liquefy some of the gas in a bulb immersed in liquid hydrogen and allow the gas to evaporate directly into the appal.atus which it is intended to fil ] .
At the temperature of liquid hydrogen all possible impurities would be practically non\amp ; hyphen ; volatile , and the gas would not require further purification .
The hydrogen employed in these esearches was obtained by the action of dilute sulphuric acid on pure platinised zinc .
The gas was passed through a solution of potassium permanganate , and through a tube containing pentoxide of phosphorus into a tube fitted with a stop\amp ; hyphen ; cock containing pieces of palladium sponge .
When the palladium was saturated , the stop\amp ; hyphen ; cock was closed and the tube connected with it was sealed to the apparatus for fillin the thermometer .
The gas from the palladium tube passed through a tube , about 40 centims. long , containing pentoxide of phosphorus , through a large bulb filled with glass beads and immersed in liquid air , to the inlet tube of the thermometer .
The apparatus also communicated through a stop\amp ; hyphen ; cock with the mercury pump and with a mercury manometer .
Before filling the thermometer the whole apparatus was exhausted and a considerable quantity of hydrogen allowed to escape from the palladium .
The actual operations involved in filling the thermometer have already been described ( p. 117 ) ; it is only necessary to state here that every precaution was taken to avoid contamination of the gas by moisture or by other impurities .
APPENDIX III.\amp ; mdash ; Note on the\amp ; hyphen ; Preparation of Purc , Helium .
As has been point out by one ( TRAVERS , The Experimental Study of Gases the constant\amp ; hyphen ; volume helium thermometer furnishes the only reliable means of measuring low temperatures .
For the purpose of measurin the temperature of liquid hydrogen boiling under reduced , it is of particular importance that the helium should be free not only from argon , but even from neon ; for as that element\amp ; hyphen ; has a vapour pressure of millims. of mel.cury at the boiling\amp ; hyphen ; point of liquid drogen , and millims. at , its vapoul at ) .
would probably be less than 1 millim. of mercury , and consequently a very small quantity of this impurity would be sufficient to vitiate the results .
As it is obviously impossible to separate neon completely from helium by cooling the gas to the temperature of liquid hydrogen , it is necessary to employ helium which is initially as free from that gas as is possible .
DEWAR ( loc. cit. , p. 155 ) has employed the gas from the Bath springs as a source of helium , but , as his own ations show that the resulting helium contains per cent. of neon , it is unsuitable thermometric purposes , particularly for the measurement of temperature , which can be obtained by boiling ogen under reduced pressure .
Some years ago it was shown TRAVERS , ' Roy .
Soc. Proc vol. p. 206 ) that by subjecting the gas from the mineral cl\amp ; egrave ; yeite to the process of frac\amp ; hyphen ; tional diflusion , it was possible to sepal'atefi.om it small quantity of argon and to leduce the density of the lightel fraction to .
Later it was found by one of us , though the experiment has not been that when the gas from cl\amp ; egrave ; veite was passed ough liquid oxygen , cooled to C. in a bulb immersed in liquid boiling under reduced , a small quantity of rgon was removed fi.ulll the , but appal.elltly no neon .
this reason cl\amp ; egrave ; veite gas ) eared to be the most suitable source for obtainin pure heliuul .
The light helium by the diffusion ineIlts 1 in these ches ; it was purified by ssing it at the nessure through a coil imnel.sed in liquid hydrogen .
The employed will now be described .
The , of which we had bout 1 cub. centims. , was oduced tough the syphon ( figr .
3 ) , into the gas\amp ; hyphen ; holder , which contained mel.cul.y .
The gas\amp ; hyphen ; holder was connected by a ubber tube with a mercury reservoir not shown in the figure , and in order to eliminate any risk of air entering the gas\amp ; hyphen ; holder through leakage of the rubber tube , the colmection waoe made through a trap , D. The gas\amp ; hyphen ; holdel conmlunicated through the two\amp ; hyphen ; way stop\amp ; hyphen ; cock , , with the syphon for introducing the gas , and with the ating coil , which led through a second stop\amp ; hyphen ; cock , , to a second holdel .
similar to the .
The coil communicated yith the mercury pump )\amp ; hyphen ; cock After oughly e the coil , the stop\amp ; hyphen ; cock closed , and the gas admitted to it .
A vacuum vessel containing about 300 cub. centims. of liquid hydrogen , surrounded by a larger vessel containing liquid air , was brought below the coil and raised till the latter was thoroughly immersed in the liquid .
The stop\amp ; hyphen ; cocks , and , were then carefully opened , and by adjusting the level of the mercury reservoir connected with the gas\amp ; hyphen ; holders , the gas was passed three times backward and forward through the coil .
When the whole of the gas had returned to the first gas\amp ; hyphen ; holder , the stop\amp ; hyphen ; cocks were closed .
The gas in the gas\amp ; hyphen ; holder we called fraction 1 .
About 100 cub. centims. of the original gas , measured under normal conditions , still remained in the coil , at the boiling\amp ; hyphen ; point of liquid hydrogen a gas occupies only about 1/ 15 of the volume which it fills at the temperature .
On the cock E. a moment , about 70 cub. centims. of the gas passed into the pump .
This fraction ( 2 ) was probably as pure as fraction 1 , for the additional cooling produced by its sudden expansion would compensate for the tendency of any light impurity to porate ; it was , however , kept separate ( fraction 2 ) .
When fraction 2 had been pumped off and collected , the ck was again ened , the remain volatile gas was pumped off and collected .
It formed action 3 , and consisted of about 25 cub. centims. of gas .
Finally the liquid remoyed , the esidue in coil , which volatilised and passed into , was collected as fraction 4 .
action 4 was found to consist largely of nitrogen , which had entered the tube in which the gas had been stol.ed since 1898 at the time it was last ndled .
was with oxygen over potash some hours , and after removal of the excess of oxygen about cub. centim. of gas was ieft ; it was found on slJectroscopic examination to consist of th.gon and krypton only , showing the of the latter with brilliancy .
fact that cl\amp ; egrave ; veite helium yields ) indicates that that gas must be in the original mineral .
As krypton is present in .
to a extent greater than 1 pal't in 1,000,000 , the presence of that gas in the helium could not possibly be due to leakage of air into the apparatus .
The quantity of krypton , like mercury vapour , which is necessary to give a brilliant spectrum in presence of argon , is exceedingly small , and possibly the total quantity of ypton in the gas might not exceed cub. centim. , or per cent. on the original helium .
It is interesting to note that when WSKI attempted to liquefy helium by compressing it at C. , and allowing it to expand , a small quantity of a white substance separated .
It is intended on another occasion to investigate the heayy fraction of gas obtained in 1898 by the diffusion of helium .
The first fi.action of gas , hich we called fraction 1 , was used in the first series of measurements of the boiling\amp ; hyphen ; point and vapour pressures of liquid hydrogen down to .
Fractions 1 and 2 were subsequently mixed and passed a second time ough a coil liquid hydrogen directly into the large constant\amp ; hyphen ; volume thermometel .
the measurement of the pressure coefficient ( p. 129 ) , which was found to be For the second series of ements of the of liquid hydrogen the helium was further pul.ified by passing it through a coil immersed in liquid hydrogen boiling under reduced pressure .
For this pose the coil , through which the helium was passed ( fig. 3 ) , was enclosed within an apparatus similar to that employed in determining the pressure of hydrogen .
The arl.angement which is shown in fig. equires little tion .
The vacuum vessel containing the liquid hydrogen } introduced from below into the ( p. 160 ) tube so as to surround the coil itlld the bulb which cined the pure ogen , was colmected onlete by which the ya on the } ) from this ON THE OF TEMPERATURE .
the temperature of the coil could be determined .
By means of the exhaust pump the temperature of the liquid hydrogen was reduced to 110 millilns .
of mercury , corre\amp ; hyphen ; sponding to a temperature of Helium was then allowed to fill the coil at normal pressure , and by opening the stop\amp ; hyphen ; cock the gas was slowly admitted into the there \amp ; hyphen ; mometers , which had previously been ex\amp ; hyphen ; hausted .
APPENDIX IV.\amp ; mdash ; Note on the , of Solid The following experiments are of interest both with regard to their application to the preparation of pure helium and to the light they throw on the homogeneity of neon .
In the first experiment the neon was con\amp ; hyphen ; densed in a fractionating bulb ( RAMSAY and TRAVERS , 'Phil .
Trans 1901 , vol. 197 , p. 51 ) which was connected with a manometer reading pressures below 100 millims. ofmercury and , through a stop\amp ; hyphen ; cock , with a mercury pump .
About 20 cub. centims. of neon were introduced into the bulb and condensed by cooling it by means of liqnid hydrogen , boiling under the normal pressure on the helium scale ) .
The vapour pressure observed by means of the manometer was millims. of mercury .
On opening the stop\amp ; hyphen ; cock , and allowing some of the neon to evaporate into the pump , the pressure fell momentarily , but on closing it the again rose to millims. The operation was repeated several times with the same result , proving that neon is a homogeneous substance .
On remoying the vacuum vessel containing the liquid hydrogen , a small quantity of solid neon was seen in the fract\amp ; hyphen ; ionating bulb .
* NOTE , th , 1902 .
\amp ; mdash ; After filling the thermometer , the portion of the gttS cined in the coil , which was maintained , was allowed to escape into the mercury pump , by oponing the stop\amp ; hyphen ; cock communicating with it for a few scconds .
When this gas had been mped off , the stop\amp ; hyphen ; cock as agaiu opened , and the remaining gas collected separately .
The last fraction , which consisted of about 10 cub. centims. , examined spectroscopically , but it appeared to contain no trace of ncon .
DR. M. W. TRAyERS , MR. G. SENTER , AND DR. A. JAQUEROD In the second experiment , the neon was introduced into the bulb of the apparatus employed in determining the yapour pressure of liquid hydrogen ( fig. 1 ) .
At a temperature of ( helium scale ) , the vapour pressure of the neon was millims. of mercury .
NDIX V. the Probable Values of the and Boiling\amp ; hyphen ; Points of , and on Attempt to that As the foregoing results show , helium behaves at low temperature as a more perfect gas than hydrogen , and even if no further data were forthcomin , one would expect to find that its critical and boiling\amp ; hyphen ; points were lower than those of the latter gas .
This , as we shall presently show , receives further confirmation from theoretical considerations and from experimental results .
In 1892 , OLSZF , WSK1 ( loc. cit. , p. 174 ) showed that when helium , under a pressure of 80 atmospheres , was cooled to C. , and suddenly allowed to expand till the pressure fell to that of the atmosphere , no mist was seen in the compression tube .
More recently ( ' Roy .
Soc. Proc 190 ] , vol. 68 , p. 360 ) DEWAR performed a similar cooling the compression tube in hydrogen reduced to its freezing\amp ; hyphen ; point .
He did not , however , succeed in liquefying the gas , which he believed to have been cooled to or absolute , the temperature which he assigns to the upper limiting value of its critical point .
Though by compressing any gas , and subsequently allowing it to expand adiabatically , the relationship between the initial and final pressures and the initial and final temperatures can be found by the equation when is the atio of the specific heats for gas , this equation may not be applicable to the case in which a gas is compressed into a capillary tube and then allowed to expand , the thermal capacity of the gas being then very small compared with that of the alls of the containing tube .
It occurred to us that if we could cool a few csntimetres of a compression\amp ; hyphen ; tube containing helium down to a temperature below the critical point of the gas , then , on compressing the gas , its volume would decrease and the pressure would rise till the vapour pressure of helium corresponding to that particular temperature was reached .
It should then be possible , supposing the gas to be homogeneous , to decrease the volume of the helium without increasing the pressure , as a change in the volume of gas would imply liquefaction in the cooled portion of the tube .
Before describing our experiments we will first deal with the probable value of the critical and points of helium .
ON THE UREMENP OF Of the five gases , helium , neon , , krypton , and xenon , the vapour pressures and critical constants of the last three only have been determined .
They are as follows RAMSAY Phil. Trans. .
It will be that the ratios of the absolute boiling\amp ; hyphen ; points to the absolute critical points has a nearly constant value , the mean of the actual ratios being This relationship is common among similar substances , for which the boiling and critical temperatures may be considered as ponding states .
The boiling and critical points of neon are not known .
We are c\amp ; eacute ; rtain , however , that the critical point lies below abs .
, and , as the vapour pressure is millims. at abs .
the boiling\amp ; hyphen ; point must lie above .
Tf the atio for the critical and boiling\amp ; hyphen ; points is the same as for other gases , , we may calculate the temperature of the boiling and critical points which will \amp ; hyphen ; point Critical point In the paper in the ' Philosophical Transactions ' already referred to , it pointed out that argon , krypton , and xenon resemble one another , though krypton bears a stronger resemblance to neon and helium , , in fact , a link between the two sub\amp ; hyphen ; series of this group of elements .
If now we plot \amp ; obar ; ) on a the critical and boiling\amp ; hyphen ; points of the elements as ordinates and tneir atomic weights as abscissae , the points nting argon , krypton , and xenon lie on a straight line .
Taking 30 abs .
as the boiling\amp ; hyphen ; point and 53 .
as the critical point of neon , and plotting these points on the diagram the atomic weight of neon , we find that the joining these points to the corresponding points for krypton , wh.en prolonged , pass through the origin of the axes .
If the points corresponding to helium lie also on these lines , the critical and boiling\amp ; hyphen ; points of this gas will be and , Even if we assign to the critical point of neon its upper limiting value , and suppose the points on the representing krypton , neon , and helium on a curve , the critical point of helium cannot be far above In our experiments the helium was compressed in the compression tube of an apparatus similar to that employed by CAILLETET or , only the pressure was transmitted to the surface of the mercury by means of compressed air , instead of V0L .
ON THE ASUREhIENT OF PERATURE .
sealed to the tub , which internal diameter of milli ] ) .
When the lower portion of the tube was cooled to the lowest tempel.atm.e , it required a pressure of 60 atmospheres to compress the wholc of the helium into the capillary tube .
The pressure on the helium was determined by means of two gauges of the Amagat type , filled with nitrogen , and indicating pressures from 2 to 12 atmospheres , and from 8 to 60 atmospheres respectively .
The compression tube was filled with pure helium pre\amp ; hyphen ; pared in the manner described on p. 174 ( fraction 1 ) .
In our first experimeent the capillary compression tube was enclosed together with the thermometer bulb , \amp ; c. , in the apparatus shown fig. 1 ( see also p. 160 ) .
At temperatures between and abs .
( helium scale ) , the pressure on the gas was slowly increased to CO atmospheres , and thetl slowly reduced .
Under all con\amp ; hyphen ; ditions the smallest change in the position of the mercul.y meniscus in the compression tube was always accompanied by a corresponding in the pressure .
This experi\amp ; hyphen ; meant was repeated on three occasions .
In another set of experiments , the capillary portion of the compression tube passed ough a rubber stoppcr ( fig. 6 ) into a small silvered vacuum vessel , with a contracted mouth , enclosed within a wider glass tube sealed at the bottom to a tube , 7 millims. in diameter , and fitting closely to the rubber stopper E. The upper part of the tube contained a plug of natural wool wrapped in gauze , to shield the mouth of the vacuum vessel from radiation from above .
A tube passed through the stopper , and was comlected with a mercury manometer which seryed to measure the pressure in D. The vacuum vessel was filled with liquid hydrogen and placed in the tube , which was then rapidly fitted to the rubber stopper ; the compression tube had previously been cooled with liquid air .
A vacuum vessel containing liquid air was then brought outside the tnbe as in fig. 6 .
The tube was connected with the double ' Fleuss ' pump , arranged with the cylinders in series ( p. 161 ) , and the pressure on the liquid hydrogen was reduced to 5 millims. of mercury , and maintained at that pressure for 20 minutes .
at this temperature , which is probably not far below abs .
, no evidence could be obtained that helium had efied .
The extremely permanent character of helium confirms our view that gas is by 180 OTHERS ON TH1 IEASUltEMENT OF TEMPERATURE .
the most perfect nometric substance known to us .
Possibly , even at the lowest temperature that we have arrived at , the temperatures measured on the constant\amp ; hyphen ; volume helium scale are not far from absolute measurements .
Turning to the question of liquefying helium , it would of course be possible to compress large ities of the gas , and allowing it to expand adiabatically to at least produce a mist of helium .
As the compression could only be effected in vessels with walls of considerable thickness , it would probably be impossible to produce any quantity of liquid in this way .
With regard to the application of the regenerative process , which can so easily be employed in the case of air or of hydrogen , to the liquefaction of helium , we can say nothing ; for apart from the experimental difficulties involve in the problem , we do not know whether helium becomes heated or cooled allowed to expand freely at high or at low temperatur this point ( 11 view , and for the correction of gas thermometers , the Joule\amp ; hyphen ; Thomson should ) determined over a wide range of temperature .
In conclusion , we wish to express our thanks to the overnment -rant Committee of the Royal Society for the assistance they have given us in carrying out this investigation .

