428 The Absorption Spectrum and Fluorescence of Mercury Vapour .
By W. N. Hartley , D.Sc .
, F.E.S. ( Received June 16 , 1905 .
) Having undertaken the investigation of the absorption spectra of metals hi a state of vapour , the first substance examined was mercury , and as the results are interesting I have deemed it advisable to make them a separate communication to the Society .
F. P. le Roux describes the vapour of mercury as having a bluish colour , * and according to R. J. Strutt , it transmits a feeble steel-blue colour , but the absorption coefficient is small.f Experimental.\#151 ; The substance to be volatilised was contained in a flask of Heraeus ' quartz-glass , with a side tube to the neck from which the metal may be distilled and condensed .
To the side tube a water-jacket is fitted through which a constant stream of water may be passed if necessary .
The rays from the condensed spark of a pair of lead-cadmium and tin-cadmium electrodes were passed through the flask and on to a cylindrical condensing lens of quartz which focussed the rays on to the slit of a quartz spectrograph .
The mercury to be used was first purified by distillation .
The photographic plates used were various , such as " Rainbow Fast " Warwick plates , Lumiere isocliromatic , yellow-green sensitive , and Cadett and Neall 's " Lightning Spectrum " plates .
The mercury vapour in the flask was at a pressure of 847 mm. , the barometer standing at 763 mm. , but the vapour was under a pressure of a column of 84 mm. of mercury above that of the atmosphere .
The temperature was about 360 ' C. , the b.p. at 760 mm. being 357 ' .
The volume of the vapour was 31 c.c. , and its weight was calculated to be 0T33 gramme .
The thickness of the layer of vapour was 37 mm. Several photographs were taken , and particular care was exercised so as to have both ends of the spectrum , as well as the central part , in accurate focus .
The developer used was " imogen sulphite .
" The Absorption Spectrum.\#151 ; The whole rays were transmitted from the red to a point in the ultra-violet where there is a tin line at \ 2571*67 .
From there to X 2526'8 there is a very sharply defined and intense absorption band , somewhat degraded on the side towards the red ; beyond that the rays are transmitted with full intensity to a wave-length about 2000 .
The Fluorescence.\#151 ; When the mercury was boiling briskly the whole side of the flask nearest to the spark was lighted up with a green fluorescence , * 'Comptes Rendus , ' vol. 51 , p. 171 , 1860 .
t ' Phil. Mag. ' ( 6 ) , vol. 4 , p. 596 , 1902 , and vol. 6 , p. 76 , 1903 .
Absorption Spectrum and Fluorescence of Mercury Vapour .
429 this penetrated about one-third of the space within the flask , and lighted up the interior .
The quartz-glass itself was not fluorescent in the slightest degree .
When all the liquid mercury had become converted into vapour , the temperature no doubt rose above that of the boiling mercury , the vapour was quiescent , and the fluorescence ceased .
The interior of the flask was then quite dark .
By shaking some of the condensed cold mercury down into the flask the fluorescence was resumed directly the liquid boiled again , but the dropping of cold mercury into the heated vapour caused condensation , and only after the flask had again become filled with the mercury vapour was the fluorescence fully displayed .
When the vapour was rising from the boiling globule of mercury after the cold metal had condensed all within the flask , the vapour could be seen by its fluorescence to undergo condensation in the upper part of the vessel and descend to the hotter space below .
It occurred to me that the actual fluorescence might be associated with oxidatiou of the vapour , and that it appeared only when such chemical action was taking place , but subsequent observations showed that this could not be the case , because the temperature was above that when oxidation could occur at the time when the fluorescence was most brilliant , and when it most completely tilled the vessel , and also when the mercury vapour had expelled all the air .
When the temperature rose above the boiling point of mercury and excess of liquid mercury and vapour had been expelled from the flask the fluorescence ceased .
This fact leads to the inference that the fluorescence occurs only between a lower and a higher limit of temperature .
What these small differences really are I had no means of determining .
Having established the fact that the property of selective absorption is possessed by small quantities of mercury vapour , it was resolved to ascertain whether the band showed itself in solutions of mercury compounds .
As a rule the absorption spectra of compounds differ from those of the elements entering into their composition entirely , as in the case of the halogen compounds of the alkali metals ; sometimes it is a question of degree , as in the case of the compounds of the rare earth metals , in which similar bands are observed in different salts of the same metal , but in different positions , which vary with the molecular weight of the salts ; and there are , still further , other instances where the absorption bands of the solutions are distinctly the properties of the salts , as in the case of the chlorides , bromides , and iodides of cobalt .
The salt chosen for examination , because it is the most definite and most soluble , was mercuric chloride .
It was examined in cells of 40 mm. thick , diminishing to 1 mm. in thickness .
430 Absorption Spectrum and Fluorescence of Mercury Vapour The solution contained in the same volume ten times as much mercury as the vapour which filled the fiask , or T8 gramme of mercuric chloride in 31 c.c. of water ; more dilute solutions were examined containing 0T8 gramme and 0018 gramme .
No absorption band was visible on any of the spectra photographed , but there was a continuous absorption at the more refrangible end of the spectrum , which regularly diminished as the quantity of mercuric chloride in the solution decreased .
Further details are as follows :\#151 ; 1'8 grammes of mercuric chloride in a cell 40 mm. thick transmitted all rays to X 2702 , in 1 mm. to X 2572 ; 0*18 gramme in a cell 2 mm. thick transmitted all rays to X 2265 ; and 0'018 gramme in similar circumstances transmitted very feebly to X 2145 .
The absorption band in the vapour of mercury belongs to the vapour , and is accompanied by strong fluorescence between a certain maximum and minimum of temperature lying very near to the boiling point .
In studying the fluorescence of solutions of organic compounds I have shown that it is necessary to use the ultra-violet rays and quartz apparatus , * as it was found that fluorescence was associated very generally with a powerful absorption of rays in the ultra-violet .
It is a question still undecided whether the rays absorbed by mercury vapour as shown by the band I have measured , reappear with a lowered refrangibility as yellowish-green light in accordance with the law of Stokes .
The spectra were photographed with all due care by my assistant , Mr. Douglass Melon , A.E.C.Sc .
I. * " Observations on the Origin of Colour and Fluorescence , " ' Chem. Soc. Trans. , ' vol. 63 , pp. 245\#151 ; 256 , 1893 .

