]\gt ; Mr. R. C. Dearle .
Emission 2 .
Zinc vapour , when injected into Bunsen flames , is not ionised , and does not emit any light characteristic of the spectrum of zinc .
3 .
A Bunsen flame which is supplied with cadmium vapour emits light of wave-length .
when the intensity of the flame is weak , and when burning strongly it emits light of wave-length .
as well .
The cadminm vapour in such flames does not appear to be ionised .
4 .
Magnesium vapour which is fed into the flame of a Bunsen burner emits light of wave-length , and the vapour in the flame is ionised .
The ionising potential for atoms of magnesium vapour appears to be volts .
5 .
Thallium vapour , when it is fed into a Bunsen flame , becomes strongly ionised , and under these circumstances emits light of the wave-lengths .
and 6 .
The combined results of the investigatio neither conclusively support nor definitely tend to invalidate Bohr 's theory of atomic structure .
Emission Absorption in the Infra-Red Spectrum of By BAYMOND C. DEARLE , M.A. , University of Toronto .
( Communicated by Prof. J. C. McLennan , F.R.S. Received July 17 , 1916 .
) INTRODUCTION .
In a previous paper McLennan and Dearle , was pointed out that bands had been found in the absorption spectrum of non-luminous mercury vapour at , and , but that nothing had been done up to the present in the way of investigating the infra-red region of this spectrum for characteristic absorption bands .
In the same paper an account was given of some measurements made on the infra-red emission spectrum of the mercury arc for the purpose of est.ablishing the wave-lengths and intensities of the lines in that region .
The present paper deals first with some additional work on the relative intensities of these lines , and , secondly , with the absorption bands produced by ) assing white light thrqugh non-luminous mercury vapour .
* McLennan and Dearle , ' Phil. Mag , p. 683 ( 1916 ) .
Ab sorption in the Infra-Red Spectrum of Mercury .
SPECTRUM .
In the paper by Prof. McLennan and myself ( loc. cit. ) on the infra-red spectrum of the mercury arc , it was shown that there was a disagreement between the values of the relative intensities of the lines and , as given by Paschen and by us .
According to the former , the line is the strongest in the arc spectrum of mercury , while in our ation the line was found to be somewhat the .
It was pointed out by Paschen*that , with low vapour-pressures , the intensities of the two lines were about equal , while an increase in vapour-pressure caused a relatively greater increase in intensity of the line Kuch and Retschinsky , in their measurements on the temperature of the mercury arc , found that an increase in current caused an increase in temperature , but did not affect the vapour-pressure , whereas an increase in voltage caused an increase both in temperature and in vapour-pressure .
In Paschen 's paper no mention is made of any variation in the voltage , but only of variations in current , so that , from the information given , it is not clear whether there was actually any variationl in the vapour-pressure in his lamp , or merely variations in the temperature .
In a later paper by Kuch and Retschinsky , was stated that , with temperatures the mercury arc , the lines of shorter wave-length were relatively than with lower temperatures , but no measurements were made on the line .
As this is analogous to the temperature effect for black body radiation , it seemed possible that the distribution of intensity in the lines of the spectrum might follow the usual form of the radiation curve .
This point was investigated by and later by with the result that it was established that lines belonging to the same series have their relative intensities arranged along a curve which is convex to the wave-length axis , that is , a curve the short-wave portion of an energy curve .
It was also that an increase in tempera- ture caused a relatively reater increase in the iutenbities of the wave-lengths than in those of the longer wave-lengths .
These relations , however , did not hold within a single triplet , nor among the lines as a whole .
All of the wave-lengths investigated lay within the region of the visible spectrum or in the ultra-violet , and no observations were made on the * F. Paschen , ' Ann. .
Phys vol. 27 ( 13 ) , p. 559 ( 1908 ) .
R. Kiich and T. Retschinsky , ' Ann. .
Phys vol. 22 ( 3 ) , pp. 695 , 602 ( 1907 ) .
R Kuch and T. Retschinsky , ' Ann. .
Phys vol. 22 ( 5 ) , pp. 852 , 866 ( 1907 ) .
S A. Pfliiger , ' Ann. .
Phys vol. 26 ( 4 ) , pp. ( 1908 ) .
L. Grebe , ' Phys. Zeitschr vol. 11 , pp. 1121-1122 ( 1910 ) .
Mr. R. C. Dearle .
Emission infra-red , that is , in the region where the maximum of an energy curve would be expected to appear .
The present investigation was undertaken to find ( 1 ) whether , in the infrared region , the relative intensities of the lines in a given series are distributed along an energy curve ; and ( 2 ) the effect of increased temperature and vapour-pressure on the relative intensities of the two lines and , which do not belong to the same series .
EXPERIMENTAL ARRANGEMENT .
The aJ ; rangement of the apparatus was the same as that previously described in the paper by Prof. McLennan and myself , except that the light from the mercury arc , instead of being focussed on the spectrometer slit by a concave mirror , passed directly through a quartz lens having a focal length of 20 cm .
For the determination of the intensities of the various lines as distinguished from the continuous radiation due to the heated quartz , the method was followed of drawing a cooling curve for the lamp after exti uishi the arc .
Greater precision was given to these measurements by taking logarithms of the galvanometer deflections and plotting these against elapsed time .
This gave straight lines which could be continued back Absorption in the -Red ectrum of Mercury .
to zero time very accurately .
An additional advantage of this method is that it was only necessary to take two readings with the galvanometer in order to obtain a complete curve .
An example is given below , the curves being shown in fig. 1 .
From the curve we get , for zero time , deflection hence , deflection due to radiation from the heated quartz mm. RESULTS .
The lines chosen for the first part of the investigation were those to the main series of single lines and the first three members of the series were used .
These are given by The lamp was run with two different energy consumptions with the following results : * In the symbolic equation the frequencies are given by , where is Rydberg 's number , has a fixed value either integral or one of the numbers , etc. , and has successive integral values , each one giving the frequency of a member of the series .
that they are steeper on the short-wave side , and the increase in energy is quite clearly relatively greater in the shorter wave-lengths .
The lack of sufficient points on the curve prevented the necessary accuracy for determining whether the maximum intensity shifts to the shorter wave-lengths for increase in temperature .
If this were so then the analogy to the energy curve would be complete .
It is interesting to note that the distribution of energy is governed by the actual wave-lengths of the lines and not by the order in which they occur in the series ; i.e. , although the first member of the series is and the second is , the lines are distributed along the energy curve , not in this inverted order , but in the order of their wave-lengths .
For the second part of the investigation the current in the lamp was kept constant at amperes , and the voltage.was varied so as to give an energy consumption in the lamp which from 150 to 250 watts .
For each of the applied voltages the intensity was measured on the line and on by the method described above .
The results are given in the Table below , the first column gives the energy consumption in watts , in the second column represents the intensity of the line and in column ives the intensity of the line .
The fourth represents the ratio of the two intensities in each case .
Absorption in the Infra-Red Spectrum of Mercury .
It will be noted from the last column that the relative intensity of the line to that of the line increases regularly with increase in energy consumption , or in other words , with increase in vapour-pressure in the lamp .
In all measurements with this lamp the line always appeared to be the more intense .
II .
ABSORPTION BY MEBCUIY As far as the author has been able to ascertain there has been no work done by previous investigators on the selective absorption by llon-luminous mercury vapour in the infra-red region .
It was at first intended to pass the light from a mercury arc through a layer of the vapour and examine the emergent radiation to find whether any of the lines in the infra-red spectrum were extinguished .
This method was abandoned , however , and the radiation from a Nernst glower was used instead of that from a mercury arc .
The principal advantages in this were the greater steadiness in the radiation secured , and the fact that a continuous spectrum was produced in which absorption bands would take the form of depressions .
Apparatus .
The apparatus used for these measurements was similar to that used in the Aevious work .
The spectrometer , thermopile , and galvanometer were unchanged , and the mercury lamp was replaced by the Nernst glower and the absorption cell .
The latter is shown in section in fig. 3 .
It is made entirely of fused quartz , is about 6 cm .
long and cm .
iu diameter , with a small reservoir ( C ) for the mercury underneath .
The ends were formed of two circular plates of clear fused quartz which transmitted the visible light with very little diffusion .
A small tube ( D ) at the top served to connect the cell with an exhaust pump .
The reservoir was first filled with clean dry mercury , which was boiled to eliminate all traces of water-vapour , and the cell was then exhausted to a pressure of .
few millimetres of mercury and sealed off .
When in use the cell was heated by means of an electric current applied to the walls of the reservoir and tube , temperatures as high as 19 C. being Mr. R. C. Dearle .
Emission used .
In order to obtain uniform temperature and hence constant vapour density , the cell was placed in a box and packed in powdered magnesia .
The ends were kept clear for the transmission of the radiation by means of two brass tubes projecting it from the sides of the box , as shown at BB .
It was found that on account of the radiation through these tubes , and the consequent lowering of the temperature at the ends of the cell , the mercury vapour immediately condensed on the end plates , thus cutting off the radiation very considerably .
This difficulty was overcome by putting a single turn of wire against each plate from the outside ( AA ) , and passing a current through these wires of sufficient strength to maintain the plates at a slightly higher temperature than the rest of the cell .
The light from the Nernst glower was first focussed on the slit of the spectrometer by means of a quartz lens ; then the cell and its containing box was placed in such a position that the radiation from the Nernst glower had to pass through the cell before falling on the lens .
After all the readings had been taken , the absorption cell was replaced by a plate of clear fused quartz similar to those forming the ends of the cell .
The intensity o the radiation was cut down by means of a diaphragm of Bristol board having a small hole cut in the centre , in order to make the Absorption in the Infra-Red Spectrum of Mercury .
galvanometer deflections comparable with those obtained when using the cell .
Readings were taken over the spectrum with this arrangement so as to find out whether anyaf the observed absorption bands might be due to absorption by the quartz itself .
RESULTS .
The range of wave-lengths studied was from to , and this region was investigated very thoroughly .
A great number of readings were necessary on each wave-length to obtain a satisfactory curve , and each set of readings was repeated over a period of several months .
The necessity for this lies in the unsteadiness of the zero point in reading the galvanometer deflections , which was due to the combined effects of local magnetic disturbances on the galvanometer and stray air-currents on the thermopile .
Despite these difficulties it was found that certain well-defined depressions occurred in every set of readings , while others were obtained which were not so well defined , sometimes appearing quite strong and at other times being scarcely detectable .
In the following Table are given all the wave-lengths at which absorption bands were observed and the characteristics of each one .
The wave-lengths of the emission lines as recorded in our previous paper are given for purposes of comparison .
The temperatures used varied from 14 C. to 19 C. , and an attempt was made to show that the strength of the weaker bands was dependent on the vapour.pressure within the absorption cell .
But within the limits of these temperatures the results were entirely negative .
Readings were then taken with the cell at the temperature of the room , about C. , and the resultant curves gave all the absorption bands as doubtful except those at and , absorption being quite distinct on both of these The final readings were taken , as described above , with a plate of clear quartz in the path of the radiation in place of the absorption cell .
Readings were taken over the same range of wave-lengths and a smooth curve was obtained with no evidences of any characteristic absorption bands .
-Red Spectrum of Mercury .
spectrum , but it is difficult to see how its presence could be detected , as the intensity of the line .
is so small as to require an exposure of at least one hour in order to produce an impression on a photographic.plate .
It has recently been shown by Prof. McLennan and Mr. A. Thomson that this same line is given out by mercury vapour in a Bunsen flame , and an attempt was made to see if the radiation of wave-length could be detected as well from such a source .
The experiments , however , failed to reveal any indication of the emission of such radiation .
SUMMARY 1 .
Emission .
1 .
It has been established that the relative intensities of the lines in an individual series lie along a curve analogous to an energy curve , not only in the region of short waves , but also in the long-wave region beyond the maximum of the energy curve .
2 .
The intensities of the series lines become relatively greater in the shorter wave-lengths with increase in temperature .
3 .
For the two lines , and , which do not belong to the same series , the intensity of the former becomes relatively greater for higher temperatures and increased vapour-pressures , although it is the line of longer 2 .
Absorption .
1 .
Evidences of absorption have been found with each wave-length for which there is an emission line in the mercury spectrum between and 1 .
2 .
Strong absorption was obtained on wave-lengths , and 1 .
3 .
Absorption was obtained at wave-lengths and , with very low 4 .
It has been shown that these two wave-lengths correspond to the first members of the series , and , respectively .
In conclusion , the author wishes to acknowledge his indebtedness to Prof. J. C. McLennan for gesting the investigation , and for his kind direction and helpful suggestions through its progress .

