]\gt ; The Index Dispersion of Light in Argon Helium .
By W. , B.Sc. , Research Student , formerly Scholar , of Emmanuel College , Cambridge .
( Communicated by Professor J. J. Thomson , F.R.S. Received January 14 , \mdash ; Read January 30 , 1908 .
) The initial object of this research was to find the dispersion of light in the monatomic gases argon and helium , but as it was to know absolute value of the refractive index with considerable accuracy , determina .
tions of the active index have also been made .
Their relative powers ( air ) have been found by Rayleigh , amsay and Travers , I .
coloured interference eres , but no determinations of the absolut refractive index for light of given wave-length have hitherto been made .
The interferometer method due to Jamin was used .
A horizontal beam : parallel white light was incident on the first Jamin plate .
The two beams traversed two brass tubes of equal length closed by equal thickness of worked plane glass .
Each brass tube had a small side tube attache which led to apparatus for the density in the tube and measuring .
After reflection at the second Jamin plate , the recombined bean was focussed on the slit of a r$pectrometer , the spectrum obtained with a plane diffraction grating was examined through the telescope of spectrometer and was seen clossed by bands of maximum and minimum intensity .
At the same time , by means of a small reflecting prism , a portio1 of the slit was illuminated the light from a Plucker tube hydrogen , helium , and mercury vapour .
Thus one observed the interference bands and certain standard lines of known wave-length with any one of which the cross wire of the telescope could be brought coincidence .
\mdash ; The active index of a at C. and 760 mm of leno t is iven b .
C. is the temperature , is the coefficient of expansion , is nulilber of bands which move oyer the standard line of as th ' cm .
, and is the length of path in the tube .
The values of and are determined once for all , and those of ltayleigh , 'Boy .
Soc. Proc vol. 59 , p. 203 .
Ramsay and Travers , ' Roy .
Soc. , vol. 64 , , and vol. 6 p. 331 .
Index of 39 ] .
bserved in each experiment described below .
gas laws are assumed hold for argon and helium over the limited of pressure and emperature used , is taken as equal to Let and be the wave-lengths of two standard lines , and the corresponding refractive indices under pressure the corresponding refractive indices under pressure the diminution in number of fringes between the standard lines when the pressure changes from to a less pressure If be the nunaber of hands , then the number of bands rossing , so that the difference of path introduced is nd lso , from Gladstone and Dale 's law , Substituting for and , we } , from Cauchy 's , ve obtain Equating these values , we get The quantities to be determined are and The two former are known , and the latter are observed as described below .
If the absolute value of the refractive index be known for wavelength , then by for and , we determine in Cauchy 's equation .
in .
The straight filament of a Nernst lamp was used as a source of .
The lamp was mounted in a collimator tube so that the filament was vertical and at the focus of the collimating lens .
The portion of the filament nsed was limited by a small circular of about 2 mm. diameter placed close to it .
The tube was supported on an adjustable stand , so that one could thus a parallel horizontal beam of white light .
The brass tube in which the Nernst lamp was perforated in the neighbourhood of the filament with a ring of holes about 1 cm .
in diameter , so that the collimator tube should not get too hot : and radiation from it was prevented from reaching the optical tubes by screens of tin .
1908 .
] Dispersion of Light in and Heliu Intcrferometer.\mdash ; The dimensions of the Jamin plates were the reflecting surfaces being 6 cm .
After silvering the back reflecting surface of each plate , each was placed on ( an adjustable screw stand at a convenient distance apart , so that two horizontal parallel beams of white passed between them .
The brass tubes to contain the gases were mounted in shallow on an adjustable wooden support , and so placed that one of the reflected beams passed each tube .
The tubes were wrapped in cotton , and a hermometer reading to tenths of a lay between them .
The tubes Nere 1 cm .
long and about cm .
internal diameter .
They were closed each end by worked pieces of plane glass , 1 cm .
thick , attache by means cement .
These glass ends were held in place by perforated screw caps cemented in position , so that the tubes were capable of a coniderable internal pressul.e .
Into each tube a smaller side tube was screwed soldered , for attachment to the apparatus for altering the pressure .
The ubes and plane lass ends were obtained from Messrs. Adam Hilger and Co. recombined beam was focussed on the slit of a spectrometer by means of lens of short focal length .
The stancis supporting the above , and the source light , were all on a heavy slate slab built upon brick supports on the round floor of the laboratory .
The apparatus was thus free from vibration md consequent displacement of ) ' The \mdash ; The image of the filament produced by the -focus lens on the slit of the spectrometer covers only about 1 mm. of , he central portion of the length of the slit , so that the spectrum observed , asing a plane grating in the usual way , is narrow in a vertical direction and is crossed by bands .
By adjusting one of the Jamiu plates bhese bands can be obtained vertical and at a convenient distance apart .
This adjustment is most conveniently made in the first place with a sodium flame source .
By width of the slit , the minima can be made more less black , there is a cross wire in the eye-piece of the telescope which can be brought into coincidence with the middle of any one of these bands .
Standard liines in Spectrum .
These vere obtained from a small Plucker tube containing hydrogen , helium , and mercury } ) .
By the use of a small reflecting prism a portion of the slit was illuminated by the from the narrow part of the Plucker tube , when certain well-defined lines of known wave-length could be observed .
The lines used were five in nber , fairly distributed across the visible spectrum .
That there was no measurable error introduced by this means of illumination was shown by , at the same time another similar tube placed directly in front of the slit , when the Mr. W. Burton .
Refractire Index [ Jan. 14 , lines seen were continuous .
The standard lines used were the hydrogen red ( cm helium yellow cm mercury green ( cm helium green cm and hydrogen blue cm In determining the absolute refractive index , the sodium line was used , the slit being illuminated directly by a sodium flame , whilst the cross wire was fixed in coincidence with the line .
Pressure \mdash ; With the tubes used , a change of pressure of 1 atmosphere may cause the transit of about 500 bands in the case of argon , and of about 60 bands in the case of helium .
Hence if one wishes to get 300 bands by in the latter case there must be a pressure range of 5 atmospheres , whilst in the former case a range of of an atmosphere will suffice .
It was necessary therefore to adopt different methods for altering the pressure in the two tubes .
Argon.\mdash ; The apparatus used in the case of argon was essentially the same as that adopted by Rayleigh , comparing refractivities .
It is shown diagrammatically in fig. 2 , and the method of and the FIG. 2 .
FIG. 3 .
pressure is obvious .
The reservoir is of about 500 .
capacity , and the manometer limbs of 5 mm. diameter .
The reservoir is joined to the optical tubes by compo tubing of 4 mm. bore .
This tubing is soldered to the side * Rayleigh , ' Roy .
Soc. Proc vol. 59 , p. 203 .
1908 .
] Dispersion of Light in Argon Helium .
bube of the optical tube , and fastened by sealing wax to the glass tubing from the reservoir .
Except for the gauge tubin .
and the end into which the compo tubing is sealed , all the lass tubing is of capillary bore .
, of course , is used to the maximum pressure change from a given olume ( of gas .
The reservoir and optical tube could be xhausted from the side tube , and this side tube also the could be ntroduced .
After the gas is put in , the side tube is sealed off at a onstricted part .
The pressure was read on a cardboard millimetre scale laced behind the manometer tubes .
This scale was lued on to the wooden upport and was afterwards tested against a standard brass scale .
Over the ange of pressure that was to be used , the scale readings were accurate to vell within 1/ 10 mm. were taken on the scale ) eye to 1/ 10 mm. bfter a rise in both limbs of the manometer had been caused by he pressure tubing below the gauge .
Three at least were aken , and the mean value used .
Helium.\mdash ; The method of changing and measuring the pressule will be from an inspection of fig. 3 .
The longer reservoir , about 105 cm .
long and 1 sq .
cm .
cross section , is connected above with the pressure and indicator rauges , and through compo tubing with the optical tube .
Below it is onnected t a steel tap with a shorter cylindrical reservoir about 55 cm .
ong and .
cm .
in cross section .
As much mercury is introduced as will fill the longer reseryoir , and about 10 cm .
in of the shorter eservoir .
This leaves about 660 .
in the longer reservoir to be occupied the at atmospheric pressure .
The cylindrical reservoirs were made rom ordinary iron steam piping , and the smaller tubing with which nercury would come in contact was of steel ; all connections were made screw-in joints and sCaling wax , the latter bein melted into the hot 'hread before screwing the tubes into position .
Above , where the mercury lid not extend , brass -pieces and tubing were used , and the connections were made by screw-in joints and solder .
The glass for gauges of capillary bore , about 2 mm. diameter , and fastened in position with ealing wax .
The internal diameter of the brass or steel tubing into which he glass tubes were fitted was widened for the last 2 or inches , so that 'he glass tube could just fit in , and so that the end could press up against 'he ledge formed by the difference in internal diameter .
pressure was produced by forcing mercury from the shorter reservoir into the longer one .
This was done by pumping air into the former .
To do bhe valve from a motor tyre was soldered on the top of the shorter reservoir , and a foot pump used for motor tyres enabled a pressure up to 10 atmospheres to be obtained if necessary .
The steel tap was then closed VOL. LXXX .
396 Mr. W. Burton .
Refractive Index [ Jan. 14 and the valve opened .
By turning the steel tap now , and slowly reducin the pressure , it was quite easy to count the bands as they moved as as desired over the cross wire .
Whilst taking observations for the refractiv index , one limb of the manometer gauge was open .
The pressure rang then measured was about 1300 mm. Readings were taken by eye on millimetre cardboard scale placed behind the manometer .
It wa endeavoured to read to 1/ 10 mm. , but as the tubing was of capillary bor and the considerable , this was not always possible , especially at lower part of the scale .
This degree of accuracy is , however , not necessar !
for there was only a transit of 100 bands , and if these were read to twentieth of the distance between two bands , the error may be 1 in and , therefore , the pressure need only be taken to 12 mm. to obtai a corresponding accuracy .
When experiments for the dispersio were made over a wide pressure range , the open limb of the manometer sealed .
The gas laws were assumed to hold for helium , and the pressu1 change was not measured , the assumption being that for a range of pressu1 from 1 to 6 atmospheres the same laws held as from 1 to 3 atmosphere Before any into the apparatus , air was pumped in ti the pressure was about 10 atmospheres , and the whole was placed water , when no leaks could be discoyered .
, after putting mercul into the reservoirs and manometer , it was pumped up , and readings of th pressure taken on successive days .
By this method any leaks coul usually be detected , but very snlall leaks might be obscured by the due to temperature .
The sealing wax joints were satisfactory , and leaks were observed from them .
On one occasion a leak was found from solder joint , and on another from the compo tubing which led to the tube .
Refractive Index and of Argon.\mdash ; The gas was prepared by Wm. Ramsay and obtained from Messrs. Tyrer and Co. It was introduce directly into the apparatus , after the latter had stood exhausted overnig } with a phosphorus pentoxide drying bulb and a small Plucker tube the latter showing that there had been no leak .
The reservoir being seale off , the refractive index and dispersion were determined as below .
A set of readings taken is given in each case , and later the results of all th other readings made are tabulated and the mean is taken .
Refractive Index.\mdash ; Pressures were read to 1/ 10 mm. after causing a ris in both limbs of the manometer , the mean of three such readings at leas being taken .
The number of hands crossing the standard line was read one-twentieth of the distance between two bands when possible .
As pressure range used was about 450 mm. , and the number of bands abou smouem J ssqn$ am1Ji 'oe I ?
JOJ xapu ) SI eJadmaiL aqn$ ' ' ' ' ' ' aJnssaJd StItOUtm J ssqnr amIJ : san ; } } \mdash ; : xadxa j nssaJd 5sqns P UILURf aq9 50 SI SJnoo 50 oUBJ ssod SR SB ssod JOU SI ?
( suo i ) ) JO sy aua } ) Sl ?
At I xapur SB UITt UIBS pu St ?
At S0 aq$ 50 uo ; I O$ SUOI ) 50 oauuoo o OS A OJ UIomJaqB ; pun Al dmoo uasq Jadmat 9 OT paJSB [ aSJnOO a@u[0 sseo ou } SI SJUa ?
JOJJS [ 8061 Mr. W. Burton .
Refractive Index and [ Jan. 14 , Pressure .
Initial .
Final .
Difference .
Number of bands crossing red and direction to left .
Number of , Bands between Standard Lines and red .
Finally green .
blue .
Time , 4.30 .
femperature of tubes , C. Temperature of manometer , C. This gives for the red line at and 760 mm. ; also for the diminu tion in the number of fringes between the red and the standard lines above , ) bands cross the red , we get whence Results .
Refractive Index of Are at and 760 mm. for line .
1908 .
] Dispersion of Light in Argon Helium .
Befractive Index of ArgoIl at and 760 mm. for red .
Date . .
Dispersion in Argon , Date .
He yellow .
Hg green .
He green .
bluc .
Mean To find we reduce as below , where refers to the red line cm to the other standard lines .
Mean Mr. W. Burton .
Refractive Index and [ Jan. 14 , Taking , and for line , cm .
, we get ) ; therefore .
\mdash ; This was prepared from thorianite .
The thorianite was placed in one half of a porcelain tube , the other half containing copper oxide .
The tube was attached to a mercury pump , and through a tube containing solid caustic potash to a gas collector full of boiled caustic potash solution There was also a mercury gauge to indicate the pressure in the apparatus After exhausting and shutting off the pump , the porcelain tube heated in a combustion furnace until the pressure indicated by manometer was just above atmospheric pressure .
Connection with collecting reservoir was then made , and the gas slowly accumulated in it In this manner about llitre of helium was collected .
The thorianite heated at atmospheric pressure , the helium comes off slowly , but present should , under these conditions , be removed by the red hot copper oxide .
The gas was introduced about 200 .
at a time into exhausted reservoir attached to the optical tube .
Before entering reservoir , the gas was passed through a drying tube of calcium chloride a tube of charcoal surrounded by liquid air , each portion of the gas kept in contact with the charcoal for half an hour before being passed on the reservoir .
A Plucker tube attached to the mercury pump , and arrangec before the slit of the spectrometer , enabled one to test the gas from time to time .
No trace of foreign gases could be observed .
Afte ] sufficient gas had been introduced to give ) required pressure range , reservoir containing it was sealed off , and the experimental results givel ] below were obtained .
Index.\mdash ; The pressure range being 1300 mm. , it is not necessary , as stated before , to read to more than mm. to get an accuracy ol 1 in 2000 , though readings were attempted to 1/ 10 mm. The number of crossing the standard line used was about 100 , so that to get the same degree of accuracy it is necessary to read to one-twentieth of the distance between two bands .
As it was possible in the case of helium always to start and end with the cross wire coincident with a standard line and with the middle of a black band , I consider this accuracy was obtainable .
It may be remarked 1908 .
] Dispersion of Light Helium .
here that with ases of small refractive like helium the accuracy of the value for the refractive index depends largely on the accuracy with which the number of bands that cross the standard line be estinlated ; for with the same pressure in helium only ohth the number bands will cross the standard line as in the cRse of air or rrgon .
It will be observed belo that the value of the refi a index increase from day to day , hilsC the values well on any single Jay .
This slight increase was , I think , due to a small air leak .
If , in the ourse of the three days over which the experiments made , an air leak of 1 in had occurred , the variation observed would be accounted .
The ; am increase is observed in the value for the red line , and the saule agreenent on any one day .
That the inclease was less from Tuesday to ] fternoon is accounted for by the fact that the ooas was put in I the leyel of the in the leservoir so that after sealing off the tap connecting the reservoirs the gas should be un ressure shtly greater than attlospheric .
The barometer rose on Tuesday light and the temperature fell , so that the fall in pressure on Wednesday possibly have been due to these changes ; that this was } ' the case is shown by the results obtained on Wednesday afternoon .
The gas , however , left on Wednesday afternoon mder slightly less than atmospheric pressure , and subsequent taken night showed on reduction a decided but small increase in the refractive ndex .
on Thursday an increase observable .
The alues , herefore obtained on Tuesday , soon after the gas was put in , and on Wedlesday afternoon , are probably most correct .
It will be seen that the extreme ralues for the refractive index obtained the three days do not differ more than 1 in , so that the for the dispersion not neasurably altered by the change in the refractive index .
As in the case of rgon , measurements of the refractive index for the line and the red were made , and a summary of the is oive below .
Dispersion.\mdash ; In some of the experiments the opeu ) of the was sealed off , and the pressure rranged so that about bands rossed the red line .
As in the case of , the bands were estimated to one-twentieth when possible , and the used when the bands as far apart as possible over this pressure range .
There is no great in having a larger number of than .
the standard line , for the bands get so close ether in the blue and green that it is difficult to measure them accurately , and is ained in accuracy in the measul'ement of is lost in the measurement of Refractive Index at .
for \mdash ; The ations were Mr. W. Burton .
Refractive Index and [ Jan. 14 , taken and reduced in a similar manner to those of argon , except that the scale reading of the manometer used was the result of one observation only and not the mean of several .
Refractive Index of Helium at and 760 mm. for line .
As explained above , the value most probably correct is , the mean of the first three observations .
Befractive Index of Helium at and 760 mm. for red .
Dispersion in Helium , Where the number of bands crossing the red line was greater than 100 , the open limb of the manometer was sealed off and the pressure was not read .
1908 .
] Dispersion of Lighi in Argon Hetium .
To find b.\mdash ; Reducing these values as for referring to the red line cm we Mean Taking , and substituting for line , cm .
, we get ; therefore Mr. W. Burton .
Index [ Jan. 14 , Discussion of Argon.\mdash ; The value obtained for the refractive index for line is Ramsay and give the refractive power ( air ) as for the hGest part of the spectrum .
If air be taken for the mean line ( Kayser and ) , this makes to be .
Ramsay and Trayers used a pressure range of about 500 mm. , corresponding to a transit of between 30 and 40 bands .
The pressure used above is about the same , but corresponds to a transit of nearly 300 bands .
The results agree to 1 in 300 , and closer agreement can hardly be expected , for if the refractive index of air for the brightest part of the spectrum be taken as , the results are practically identical .
The dispersion of is htly less than that of air , if the coefficient for air be as found by Mascart .
Hetium.\mdash ; The value of the refractive index for the line is Ramsay and Travers give the refractive power ( air ) as , and if air be taken as , this gives a value for helium of which is considerably larger than that obtained above .
Assuming that Balnsay and Travers used the same apparatus as for argon , the same pressure range would correspond to a transit of between 4 and 5 bands only in the case of helium .
Hence the accuracy of the value of the refractive power will depend largely on the accuracy with which , after a change of pressure , the middle of the movable band can be brought back into continuation with the middle of the fixed band used as a fiducial line .
amsay and Travers considered that they could estimate to 1/ 25 of a band , so that the error of the may be 1 in 100 .
The accuracy of their pressure measurements was increased by comparing the refractive power of helium with that of hydrogen , and then that of the latter with air .
By this means the bands were kept in position in the field of view over a wider range of pressure than was possible when helium was compared directly with air .
The diffel'ence in the value obtained by Ramsay and Travers and that obtained above may perhaps be attributed to the fact that the pressure changes in the experiments of the former compensate for the transit of so small a number of bands , though a diff'erence of 3 per cent. in the refractive power is more than one might expect .
The dispersion is , as will be seen , much smaller than that of any other boas so far examined , the coefficient being almost exactly half that found by Mascart in the case of hydrogen .
* Ramsay and Travers , ' Roy .
Soc. Proc , p. 331 .
Kayser and Rung , 'Abh . .
Berl .
Akad .
' ( 1893 ) .
Mascart , ' Ann. de l'Ecole Normale ' ( 1877 ) .
1908 .
] Dispersion of in and Helium .
The results for argon and helium are tabulated below , and , for comparison , Mascart 's values for hydrogen are also given .
Refractive Index , reduced to C. and 760 mm. pressure for line .
Argon , Helium Hydrogen scart ) Dispersion : equation , or ?
A. B. Argou Helium 000003478 ( Mascart)f It may be noted that the values of for these substances are approximately in the ratio 3 : 1 : 2 .
I wish to thank Professor J. J. Thomson for suggesting this work to me and for the interest he has taken in it .
My thanks are also due to Mr. C. T. R. Wilson , to whom I am much indebted for advice in all that pertains to the optical part of this work .

