269 On the Constitution of the Mercury Green Line X = 5461 A.U. , and on the Magnetic Resolution its Satellites by an Echeloyi Grating .
By J. C. McLennan , Professor of Physics , University of Toronto .
( Communicated by Sir J. Larmor , Sec. R.S. Received May 14 , \#151 ; Read June 13 , 1912 .
) [ Plates 4 and 5 .
] In a previous communication on the series lines in the arc spectrum of mercury , attention was drawn by the writer to the excellence of the spectrograms obtained when a highly exhausted commercial Cooper Hewitt mercury arc-lamp was used as the source of the light .
With this lamp , provided with a side tube carrying a window of thin crystalline quartz , it was found that exceedingly good definition was obtained over the whole range of the spectrum from slightly beyond X \#151 ; 7000 down to X = 2150 .
When using the lamp at a later period in making some observations with an echelon grating of high resolving power on the structure of some of the finest lines in the visible portion of the spectrum , it was found that the components of a number of these lines also came out with exceedingly good definition .
A special study was made of the mercury green line X = 5461 and its satellites under a variety of conditions , and the present paper , which deals with some of the results obtained , contains a number of specimens of the photographs which were taken in following up points of interest which arose in the course of this work .
I. The Structure of the Mercury Green = 5461 .
The echelon grating used in the investigation was made by the Adam Hilger Company , and consisted of 30 plates P004 cm .
thick , with a width of step of 1 mm. The refractive indices of the glass of which the plates were made , taken from data supplied by the maker , were yac = 1*5712 , yap =1*5752 , yaP = 1*5852 , and yaG = 1*5935 .
With these a Hartmann dispersion formula ya = 1*553706 + 764860 X-2\#151 ; 4*8887 x 10I .
* * * * * 7 X-4 was established for the instrument .
Three reproductions of plates taken with the echelon of the green line X = 5461 and its satellites are shown in figs. 1 , 2 , and 3 ( Plate 4 ) .
Pig .
1 shows the resolution obtained with the light intensity thrown into a single order , 270 Prof. J. C. McLennan .
Constitution of the [ May 14 , and fig. 2 with the intensity concentrated in two successive orders .
In both these cases the source of the light was the Cooper Hewitt lamp mentioned above , and the green line in the spectrum in both cases was isolated by passing the light from the tube through a Hilger constant deviation spectroscope .
Fig. 3 shows the results obtained when the green line was isolated with an absorbing fluid .
The light , after leaving the tube , was passed through a glass cell containing a solution of neodymium ammonium nitrate , then through a second filled with a dilute solution of potassium bichromate , and finally sent into the echelon grating .
The various satellites in this case it will be seen are almost of the same intensity as those shown in figs. 1 and 2 .
A number of plates were made similar to those shown in figs. 1 , 2 , and 3 , and five of the best of them were accurately measured by Mr. J. K. Robertson , who kindly offered to do this for me .
The results are given in the following table , all measurements being made from the centre of the main line , which is given by some as a doublet , but which in my plates was not resolved:\#151 ; Table I.\#151 ; Resolution of Hg line \ \#151 ; 5460-95 .
Plate .
Sat. ] .
Sat. 2 .
Sat. 3 .
Sat. 4 .
Sat. 5 .
Sat. 6 .
A 0*210 0-137 0-090 -0 '070 -0 -105 -0-242 B \#151 ; 0-134 0-091 -0 -069 -0 *104 -0-243 C \#151 ; 0-137 0-091 -0 -070 -0-102 -0 -241 D 0*213 0-135 0-091 -0-071 -0 -104 -0 -243 E \#151 ; 0-134 0-089 -0-072 -0 -105 -0 -244 Average + 0-21 + 0 -135 + 0 -090 -0-070 -0 -104 -0 -243 As can be seen from figs. 1 , 2 , and 3 , all of the satellites given in Table II are clearly defined , excepting the one of greatest wave-length + 0'21 .
This satellite was too faint to be detected on most of the plates , but on two or three of them faint traces of it were seen .
For comparative purposes , the results obtained for the resolution of this line by a number of observers are taken from a paper by Luneland , and are given in Table II .
From this table it will be seen that the resolution obtained by the writer , both as to the number of satellites and as to their relative intensities , agrees best with that obtained by G-ale and Lemon with a Michelson grating , and with that of Stansfield obtained in his later work with an echelon .
The values obtained for the positions of the satellites + 0*090 and +0T35 are , however , somewhat greater than those obtained by other observers .
Table II.\#151 ; Resolution of Hg line X = 5461 by different Observers .
1912 .
] Mercury Green Line 5461 A.U. ft ^ \#163 ; r2 -*-'7 |1 Sh * *dg c3 ^ tdD \#169 ; * H fc'- ' rd d 13,2 cd ^ 0Q rP a \#169 ; QQ w g.S o III " ?
o 'S'i 1 " fi ?
* a js PQ rj g T5 fl f H cS \#169 ; \#169 ; / " -s JSflJ SM 5 * | hP ^ l ^ 0 k *d g^ .si-lg g \#169 ; rP -g 0 ^ .2 g 'SS W ) ^ 1 i | | r ?
1 ^ 1 e s 9 I I 9 o 05 1 9 1 to CO ^ 1 |S | | o 1 o 1 0 o 1 o o o tO | rl H H H 1w 1 on oo \#169 ; 1 \#174 ; 1 * 1 rH on | j 3 ?
* | to r\#151 ; 1 05 rH H 05 CO T* on r-t H rH \#169 ; 1 ^ 05 00 CO o to on o 19 9 9 9 CO 1 9 1 GO ^ | 00 to to O I I \#169 ; 1 o o o o ( III o 1 o o o o I 1 1 o o o o OH rH rH | rH rH HlOHbCOO on C5 on to on j tJ* CO 05 l\gt ; CO CO ( M H on on on Tf\lt ; tOCOiOrHCOa\gt ; X\gt ; 0 i\gt ; -^onrHOoOiOonQ I rHrHr-irH*\#151 ; HOOOO to oo on ^ co CO CO G5 rH CO O O O rH rH O CO H O on to | on on on \#169 ; \#169 ; \#169 ; 1 1 I oooooooooooooo 1 1 1 1 1 1 1 1 \#169 ; o o 1 to " Hh on | | 1 1 \#169 ; 1 ^ O CO o oo on o 19 199 ^ o 00 1 ^ 1 o | 9 | | o 1 o 1 o o o 1 1 o \#169 ; C5 on | I on 1 ?
So S 8 19 9 19 cq 1 ^ 2 i ?
1 Ml !
o 1 \#169 ; 1 o o o 1 1 \#169 ; o S-l 1 !
1 1 1/ 10 I\amp ; 111 IS- 1S-1 1 1 1 1 to ^ I on 1 1 1 1 1 on 1 0 05 loin to 1 1 9 iP i 1 1 1 1 \#166 ; o i o 1 o 1 o o on | 1 111 1 * j CO to rH | 1 1M 1 Mil CO on 1 | 1 1 1 1 ^ 00 05 TH O | ?
lSS CO 00 1 9 to on 1 ^ i Mil o 1 o 1 o o o o 1 1 1 \#169 ; o \#169 ; on 1 1 1 1 1 1 \#169 ; 1 * 1 II 1rH I w 1 1 II 1 on TfJ on 1 1 1 1 1 rH 1 ?
8 8 | 9 | | 9 JO 1 9 Sq 1 ^ 1 i c9 | , 1 w 1 o 1 o 1 \#169 ; \#169 ; 1 \#169 ; \#169 ; on i rH 1 1 1 1 1 j CO 1 M 1 1 1 | on 1 M 1 Mil CO r* on 1 1 1 1 1 |S \#169 ; 8 9 | | 9 o 1 9 1 ^ 1 S Mil o 1 o 1 \#169 ; \#169 ; 1 o o \#169 ; g P-I r\#151 ; i r-H* 0 r\#151 ; i *5 \#171 ; . !
-^ p , oo 00 o CP 05 JO A i jo m -2 Ph p , to~ r-H :i M -S n m A ~ ^ o " ^ " ... N or 11^ ill* S \#171 ; OQ ^ " S ^ 00 ^ ^ 05 ^ 51 b ?
\#174 ; ^ .
n .co \#163 ; ^ 2 g\lt ; S ^ ^ W . .
a ^ cT \#163 ; OQ \#169 ; o O .
EH JS fi S \#163 ; *3 ^ .
4^:-iiis j,4|S2fg9 g n " .
" * Mi ?
1\#187 ; 2^ rP 4i ^ I ^Ill 'll J 'd * 3 ^ ^ ^ S S \#163 ; M 3 \#169 ; T~^ 1* j\#187 ; g ^ is 3 S pq r* g *S 2-1 \#166 ; 3 -s S | I j 5 ?
S * H\#151 ; +.+ too = |jF * *4\#151 ; 272 Prof. J. C. McLennan .
Constitution of the [ May 14 ; IT .
Variations in the Structure of the H Line \#151 ; 5461 obtained by Varying the Source .
While investigating the structure of the green line with light from a number of other sources , an effect was noted which had been observed and described by Janicki , * Prince Galitzin and Wilip , f and by Stansfield.ij : The effect was observed with an Heraeus quartz glass arc-lamp , and with a lamp of special construction , described in a previous communication , both with end on and with side exposure .
With these lamps , immediately after the arc was struck , the structure of the green line was much the same as that given in figs. 1 , 2 , and 3 .
As the lamp became heated , however , the appearance changed to that shown in fig. 4 , which was that obtained with the Heraeus lamp , and to that shown in fig. 5 , which was obtained with an end-on exposure with the special lamp referred to above .
The change consisted in the central bright band splitting up into two , while the lines corresponding to the satellites widened and became more diffuse .
In the final appearance the two branches of the main line , and the bands corresponding to the satellites \#151 ; 0'070 , \#151 ; OT04 and +(H)90 , 4- 0T35 , were all about the same width and of the same intensity .
These six bands can be seen in both figs. 4 and 5 , where they appear between two successive orders of the satellite \#151 ; 0'243 .
This structure is somewhat different from what Janicki observed .
In the paper in which he described this effect he states that he observed in place of the original line and its satellites a peculiar system of five equidistant bands when the original Fig. 6 .
* Janicki , 4 Ann. der Phys. , ' 1906 , vol. 19 , p. 35 .
t Prince Galitzin , ' Bull , de l'Acad .
Science de St. Petersbourg , ' 1907 , p. 159 .
J Stansfield , ' Phil. Mag. , ' September , 1909 .
L912 .
] Mercury Green Line X = 5461 A.U. components of the lines were lost in a continuous spectrum .
The appearance which he obtained is shown in fig. 6 , which is taken from his paper .
Prince Galitzin and Wilip suggest that the effect is probably due to a reversal of the lines or to some property of the echelon , while Stansfield rather favours the view that the bands are due to secondary spectra produced by the echelon .
To the writer , the phenomenon had the appearance of a reversal of the main line due to absorption , together with a widening and an intensification of the satellites arising from an increase in the temperature and the pressure of the mercury vapour in the tube .
III .
On the Magnetic Resolution of the Satellites of the Green of Mercury X = 5461 .
In attempting to obtain photographs of the Zeeman effect with the satellites of the Hg line X = 5461 it was found necessary to use light of strong intensity and to give exposures of long duration .
To meet these experimental requirements the arrangement shown in fig. 7 was adopted .
The magnet was provided with pole pieces each of which was divided into two H " Fig. 7 .
parts that could be bolted together .
Along the break each of these parts had a groove in it , and into this groove the Cooper Hewitt lamp was fastened as shown in the figure with its axis parallel to the direction of the magnetic field .
This tube emitted light of strong intensity and when the field was on the discharge was concentrated along its axis .
This protected the glass from the discharge and therefore permitted exposures of long duration to be made .
It was found possible with this arrangement to obtain fields up to about 3500 gauss .
With the apparatus arranged in this way photographs were taken of the magnetically resolved satellites \#151 ; 0 243 and +0'090 .
Two reproductions are 274 Prof. J. C. McLennan .
Constitution of the [ May 14^ shown in figs. 8 and 9 , which illustrate the magnetic resolution of the satellite \#151 ; 0243 .
In taking the plates from which these reproductions were made a double image prism was inserted in the telescope of the echelon .
The eyepiece was removed from this telescope and a microscope objective provided with a camera bellows and plate holder was focussed on the two real images formed in the focal plane of the object glass of the echelon ( see Plate 5 ) .
The echelon itself was so disposed that the light was thrown into two successive orders ; and the satellite in question , it will be seen , appears in the centre of each of the figures .
The upper part of the figure gives vibrations perpendicular to the lines of force and the lower part those parallel to the field .
In the upper plate of fig. 8 the satellite appears as a wide doublet and in the lower plate as a narrower doublet .
Fig. 9 represents the same result , but in this reproduction while the outer pair of the quartet is rather faint the inner pair is quite clear and well defined .
This photograph also shows the main line resolved into three diffuse components , the outer ones in the upper part of the figure being perpendicular to the field and the inner one in the lower part being parallel to it .
This photograph permitted the separation of the outer pair of the satellite 's quartet to be compared with that of the outer pair of the components of the main line .
Since the main line is known to be resolvable into a nonet under a high magnetic field with its lines displaced 0 , + fa , +a , + f + 2 where a , represents the displacement of each of the outer components of a normal triplet , it follows that the mean displacement for the outer pair of the main line shown in the upper part of fig. 9 corresponds to 3a .
A number of settings on plates similar to that shown in fig. 9 showed the displacement of the outer pair of the quartet of the satellite to be double that of the inner pair and equal to the mean distance between the outer pair of the main triplet .
It follows then that the components of the quartet into which the satellite \#151 ; 0*243 was resolved by the field used are represented by the + f and +-| a. The reproduction shown in fig. 10 was made with the echelon adjusted so as to throw the light well into the \#151 ; 0*243 satellite of one order and into the +0*090 satellite of the neighbouring order .
In the upper part of the figure each of these satellites is shown divided into two bands , one of the outer bands of satellite + 0*090 being displaced so as to obscure the satellite + 0*135 .
In the lower part of the figure neither of the satellites is resolved , as the field used was not sufficiently intense .
In fig. 11 a stronger magnetic field was used , and it will be seen that the components of the satellite +0*090 parallel to the field were resolved into a doublet , as well as those of the satellite \#151 ; 0*243 .
In the upper part of the figure it will be seen the separation of the outer components of both 1912 .
] Mercury Green Line A. \#151 ; 5461 A.U. satellites was so great as to cause overlapping with the components of the neighbouring satellites .
Comparative measurements made on the relative displacements of the different members of the quartets into which each of these satellites was divided showed the distance between the outer pair of the satellite + 0*090 to be the same as that between the outer components of the satellite \#151 ; 0*243 , while the distance between the inner pair of the former satellite proved to be only about four-fifths of that of the inner pair of the latter satellite .
The lower part of fig. 8 brings this point out very clearly , for in that reproduction while the field used was sufficient to resolve the inner pair of the satellite \#151 ; 0*243 , the inner pair of the satellite +0*090 does not show any separation .
The displacements of the four members of the quartet into which the satellite 0*090 was resolved are represented therefore by + fa and + fa , where as before a represents the displacement of the outer component of a normal triplet .
Lohmann* in his paper on the Zeeman effect states that he observed that the satellites of the Hg line \ = 5461 underwent a complex resolution in a magnetic field , but he apparently made no measurements on the displacements of the components .
This , however , was done by G-ehrcke and von Baeyer by means of crossed Lummer plates .
In their experiments they only succeeded in resolving the satellites mentioned above into triplets , and from the displacements of the outer components they deduced the mean value ejm \#151 ; 2*63 x 10~7 for the satellite of shorter wave-length and e/ m = 2*77 xlO-7 for the satellite of longer wave-length . .
These results it will be seen correspond closely to the value f-a obtained for the separation of the outer component of both satellites on my plates .
Lunelandf in his recent paper gives a quartet for the resolution with moderate fields of each of these satellites , and states that with high fields the lines forming the quartets appeared to be capable of still higher resolution .
As regards resolution of the two satellites into quartets there is therefore an agreement between his results and those given in the present paper .
In one particular , however , they differ .
In his resolution the separation of the outer components of both satellites was the same ; and the separation of the components of both satellites vibrating parallel to the magnetic field was very closely one-half that of the components , vibrating perpendicularly to the field .
According to my measurements , on the other hand , while this * Lohmann , ' Inaug .
Biss .
, ' 1907 .
t Luneland , ' Ann. der Phys. , ' 1911 , vol. 34 , p. 505 .
276 Constitution of the Mercury Green Line \ \#151 ; 5461 A.U. latter characteristic was true of the separation of the components of the satellite \#151 ; 0243 , it was not true of the separation of the components of the satellite + 0*090 where the displacement of the inner pair was only four-fifths of that of the outer pair .
The reproduction shown in fig. 8 makes this point of difference quite clear .
The magnitude of the separation obtained by Luneland for the outer components of the two satellites is in practical agreement with that obtained by Gehrcke and von Baeyer and with that shown by my own plates .
1Y .
Summary of .
( 1 ) The line Hg X = 5461 from a Cooper Hewitt mercury arc lamp has been resolved by a 30-plate echelon into a main line which is probably a doublet , together with three satellites of greater wave-length and three of shorter wave-length .
( 2 ) When the mercury arc was formed in a fused-quartz lamp , with a large consumption of energy , the main component of the Hg line X = 5461 broke up into two members , probably the result of absorption , and all the satellites widened and became approximately of the same intensity .
( 3 ) The use of a Cooper Hewitt mercury arc lamp with its axis along the magnetic lines of force provided a source of light of strong intensity and permitted exposures of any desired length of time to- be made .
( 4 ) The satellite \#151 ; 0*243 of the Hg line X = 5461 has been resolved by a magnetic field into four components whose displacements are represented approximately by + f a and +fa , and the satellite +0*090 of the same line into a quartet for the components of which the displacements are approximately +f a and + -| a , where a represents the displacement of the outer component of a normal triplet .
For much preliminary work in adjusting the apparatus used in this investigation I am indebted , and my thanks are due , to my students Messrs. E. N. Macallum , S. Smith , and S. Kennedy .
MLennan .
Ro* ' So'- P , ,oc- A ' vo1 ' 87\gt ; plFig .
3 yicfennan .
Roy .
Soc. Proc. A , vol. 87 , PI .
5 .

