]\gt ; [ R05 ] VIII .
Effect of Pressure upon Arc Spectra .
No. 3450 to 5500 , including an Account of the Rate of Displacement with Wave-length , of the Relation the Pressure the Displacement , of the Influence of the Density of the Material and of the Intensity of the Lines the Displacement , of the Resolution- of the Nickel Spectrum into Groups of Lines .
W. GEOFFREY DUFFIELD , D.Sc .
, Professor of Physics , and Dean of the of Science in University College , Reading .
Commun , by Prof. A. ScHUSTI ] , Ab [ PLATES 1-5 .
] Received May Read June , 1914 .
CONTENTS .
ge 1 .
liminary . . . . . . . .
206 2 .
Behaviour of the nickel arc under pressure . . . .
206 3 .
The photographs . . . . . . . .
206 ( 1 ) Region investigated . . . . . . . . .
206 ( 2 ) Description of the plates . . . . . . . . .
207 4 .
The broadening of the lines . . . . . . . . .
208 1 ) features . . . . .
208 ( 2 ) Continuous spectrum . . . . .
209 5 .
The reversal of the lines . . . . .
209 6 .
Changes in the elative intensities of the lines . . . . . . . .
210 7 .
The displacement of the lines . . . . .
211 The measurement of the photographs . . . . .
211 ( 2 ) Table of displacement . . . . . . . . .
213 ( 3 ) The spectrum of the nickel arc under pressures of and atmospheres .
218 ( 4 ) Displacements towards the violet . . . . . . . . .
219 ( 5 ) Displacement diagrams . . . . . . . .
220 ( 6 ) Comparison with previous observations . . . . . . . . . .
220 ( 7 ) Relation between the pressure and the displacement . . . . .
224 ( 8 ) Relation between the displacement and the wave-length . . . . .
228 8 .
Resolution of the nickel spectrum into groups . . . . . .
236 9 .
Relation between the intensity of a line and its displacement . . . . .
244 10 .
Foreign metals in the spectrum of nickel . . . . .
246 ( i ) The influence of the density oi the material . . . . .
250 VOL. CCXV .
[ Pubhshed , 1916 .
1 .
Preliminary.\mdash ; The apparatus and method of taking photographs of the spectra of metallic arcs under pressure have been described in previous papers .
* The arc was formed between two poles of the metal ( not quite pure ) S-inch diameter and about 6 inches long , which were enclosed in a steel cylinder , the design of Prof. F.R.S. , capable of resisting high internal pressure .
The light from the arc passed through a window in the side of the cylinder , and was reflected by a system of mirrors upon the slit of the -foot Rowland grating spectrograph in the Physical Laboratory of the University of Manchester .
As in previous experiments , the spectrum of the .
second order was used , the dispersion being ngstrom Unit per 1 mm. An increase in pressure was obtained by the admission of air into the cylinder from a gasholder , suitable valves and gauges being interposed .
The arc was fed by continuous current from the Corporation mains , which supplied 100 volts ; this was reduced to about 50 volts at the terminals of the arc .
2 .
Behaviour of the Nickel Arc under Pressure.\mdash ; As in the case of the gold arc the ease with which the arc burned depended both upon the coolness of the poles and the freshness of the air supply .
The arc was maintained for short intervals without difficulty .
Of electrodes previously used the nickel arc behaved more like that between copper poles .
The intensity increased very markedly with the pressure , but photometric measurements were again out of question , owing to the unsteadiness of the arc .
At low pressures the nickel arc was distinctly mauve in tint , but it became whiter as the pressure of the surrounding air was increased .
3 .
The Photographs : ( 1 ) Region .\mdash ; The investigation of the spectrum extended from to : as this range of wave-lengths extends over about 162 , it was necessary to move the 50-cm .
camera into four different positions .
This involved a large amount of labour , as it practically quadrupled the amount of work which would have been necessary had it been possible to photograph the whole of the spectrum at once .
The photographs were taken partly during October , 1908 , and the remainder during April , May , and June of 1910 , the series being interrupted by the writer 's absence in Australia .
The following table shows the pressures at which photographs have been taken in different regions of the spectrum:\mdash ; W. G. DUFFIELD , 'Pbil .
Trans , 1908 ( Iron Arc ) ; vol. 209 , p. 206 , 1908 ( Copper Arc ) ; vol. 211 , p. 33 , 1910 ( Silver Arc ) ; vol. , p. 61 , 1910 ( Gold Arc ) .
EFFECT OF PRESSURE UPON ABC NICKEL .
The plates used were Imperial Flashlight and the developer Imperial Pyro-Metol Standard .
The exposure varied from five minutes to one hour according to the region examined , and the difficulty with which the arc burned .
It should be noted that one hour the total time expended upon the exposure , and includes the time when the arc was being re-struck .
The total time during which the arc burned is a small fraction of this as a rule .
The effect upon the photographic plate is that of the sum of a very large number of short-lived arcs .
( 2 ) iption of the Plates.\mdash ; Plates 1 to 5 illustrate the behaviour of the Nickel Arc under different pressures .
Plate 1 includes the region to Plate 2 the region to , Plate 3 the region to Plate 4 the region to , and Plate 5 the region to The photographs are full-size positive reproductions of the originals , and are arranged in order of increasing pressure from the top at one atmosphere to the bottom at or atmospheres .
The arbitrary numbers enumerated in Table III .
have been affixed to facilitate reference to them .
The central strip in each photograph is the spectrum at atmospheric pressure , and corresponds to the spectrum at the head of each Plate .
Above and below this strip are the lines as they appear when the arc is subjected to pressure .
The central strip as taken partly before and partly after the pressure exposure , and this provides a check upon the value of each photographic plate .
The shutter which made this possible is a modification of that originally used by HUMPHREYS , it has already been described .
4 .
Broadening of the Lines : ( 1 ) General .\mdash ; The general phenomenon of the broadening oflines under pressure has been described elsewhere .
In the Nickel Spectrum :\mdash ; Some lines broaden nearly symmetrically ; but ( b ) Most lines broaden unsymmetrically .
Of the latter class by far the larger number are more extended towards the red end of the spectrum , but a few are unmistakably broadened more on the violet side .
It is usually possible to distinguish between two classes of lines at atmospheric pressure : namely those that are sharp and those that are soft or nebulous .
Under pressure some few of the former retain some of their characteristic hardness of outline at moderate pressures ( 10 to 20 atmospheres ) , while others lose their sharp arance altogether and quickly become nebulous .
Many that start by being nebulous become more so and disappear .
Measurements of the broadening of the lines have not been made , but some indication of their relative behaviour in this respect is given in the first row of the eighth column of Table IX .
, which classifies them in the following order of increasing width : slight , ; moderate , ; considerable , ; great , ; very great , ; very very great , G. The other columns give some account of the nature of the broadened line , whether sharp or nebulous , symmetrical or unsymmetrical .
The abbreviation indicates that the broadening is nearly symmetrical ( very few are quite symmetrical , there is usually a slightly larger wing on the red side ) , indicates that the broadening is greater towards the red , and that it is greater on the violet side .
It is impossible to examine the broadening of lines in any detail without being .
impressed by the inadequacy of the nomenclature of the spectroscopist .
Under pressure a " " line\ldquo ; is but a courtesy title for the extended patch of luminosity into which its original sleek proportions have degenerated , and it is valueless as a description .
Nor is the term band quite appropriate , since a banded spectrum is by common usage something rather different in appearance from that of a spectrum ordinarily produced by pressure .
With reluctance I have retained the word line .
The terms sharp and nebulous are usually employed to distinguish between the two well-known types of spectrum lines , but the former does not seem quite satisfactory when applied to a line whose energy has been diffused over several ngstrom Units .
Nevertheless a distinction is to be discerned even under pressure in the appearance of the lines , which it is valuable to make , and I have adopted the usual terms .
The origin of the difference seems to lie in the shape of the intensity curves\mdash ; these are gradual at the boundaries of those lines which are usually called * HUMPHREYS , ' Astrophysical Journa[ ' vol. VI .
, p. 169 , 1897 .
DUFFIELD , ' Phil. Trams , vol. 208 , p. 117 , 1908 .
EFFECT OF PRESSURE UPON ARC SPECTRA.\mdash ; NICKEL .
209 nebulous , and steeper in the case of those which are called sharp .
The nebulous lines are not so clean-cut as those described as sharp , indeed the former frequently stand out against what looks very much like a fogged background .
Many of the strong unsymmetrically broadened lines from 159 onwards are accompanied by a haziness towards the red .
It is doubtful whether it is continuous with the intensity curve of the chief part of the line or whether the luminous patches are superposed .
It is possible that the different portions are due to light emitted by different parts of the arc .
A critical comparison between the broadening of the lines of various metals will be given in a subsequent paper .
( 2 ) Continuous Spectrum.\mdash ; It has been shown that under great pressure the silver arc spectrum becomes continuous , and that it is due , at any rate in the region of the spectrum investigated , to the broadenin of the lines of the first subordinate series .
* This phenomenon has been looked for in the nickel spectrum , but though there is a certain amount of continuous spectrum upon certain photographs ( e.g. Plates 3 and 4 at 155 atmospheres ) , it is generally caused by the hot metallic poles rather than by the spreading of a line ; as evidence of this we note that the continuous background is less pronounced at the higher pressure of 200 atmospheres .
There is , however , a very great broadening of some of the lines upon Plate 1 , the spreading of lines 19 and 31 being particularly noticeable ( Plate 1 ) .
Under 100 atmospheres the wings of 31 extend beyond lin on one side and 26 on the other , and may be responsible for some of the continuous spectrum which extends towards the region of longer wave-lengths .
It is thus in some ways analogous to the silver lines which belong to the 1st sub-series , and it may be a member of this sub-series in the nickel spectrum .
5 .
The Reversal of Lines.\mdash ; With increase of pressure there is at first a greater tendency for lines to reverse , and these reversals are indicated in Table III .
, those in italic thus representing a strong reversal , and those in clarendo .
type thus a weak one .
It will be observed that more reversed lines are found in the region of small wave-length than in other parts of the spectrum .
In addition to the information conveyed in Table III .
, further details concerning the reversal of lines are given in Table IX .
, where the width of the absorption line is indicated , and also the symmetry or dissymmetry of the position of the fine line upon the broadened emission line .
As in the case of broadening , the widths of the reversals are classified by slight , moderate , considerable , great , very great , very very great .
The term indicates that the absorption line is nearly symmetrically disposed upon the emission line ; indicates that the absorption line is on the violet side of the centre of the emission line .
*W .
G. DUFFIELD , ' Phil. Trans , vol. 211 , p. 33 , 1910 .
Where it is stated that a line is nearly symmetrical , but that the absorption line is slightly to the violet of the centre of the broadened line , it is to be understood that it is the geometric centre that is indicated ; it is possible that the positions of maximum emission and maximum absorption are coincident , though this is not necessarily the case , as was demonstrated for certain iron Lines .
We note from the photographs that when aline which is self-reversed is encroached upon by the wing of an adjacent line , it gives rise to an absorption lin upon the bright wing .
( Cf .
32 on 31 .
) But that when a line which is not self-reversed at a lower pressure is similarly encroached upon it does not reverse .
( Cf .
35 and 36 in the wings of 31 .
) It seems probable that this is due to differences in the distribution of the vibratin centres responsible for the different lines in the arc itself , which give rise to self-reversed lines having a different density or temperature gradient from those which do not produce absorption so readily .
6 .
Changes in the Relative Intensities of Nickel Lines.\mdash ; With increase pressure the spectrum undergoes a change which involve some lines becoming relatively more prominent than they were before .
It has previously been pointed out that it is very difficult to assign a value to the intensity of a broadened line , because though the area it covers is greatly increased by pressure the intensity per unit area is reduced .
The energy due to each line is the important quantity , but unfortunately it could only be determined by an integration which it would be extremely difficult to carry out .
In the sixth and seventh columns of Table IX .
is given an account of the changes which have been observed in the intensities of nickel lines .
The following are the enhanced lines of nickel given by 406730 419 4 Only one of these lines has been observed to show a marked change in i.ntensity under pressure , namely , ( line 87 upon Plate 2 ) , which is classified as weakened with increase of pressure ; this is the result of its great broadenin , and does not necessarily denote any reduction in the total energy emitted by the vibrating centre responsible for it .
One other enhanced line , 406730 , appears upon the photograph at 10 atmospheres pressure , which has had a prolonged exposure , and it is there strengthened relatively to some of the faint lines near it ; it does not persist at higher pressures ; as none of the other enhanced lines have been observed under pressure , it is the only exception to the general conclusion that pressure does not favour the appearance of enhanced lines .
This is in agreement with previous work upon other spectrae , ' Report , ' Solar Physics Committee .
REECE , in a less conservativs list of enhanced lines includes 4368 45 , here dassifled as strengthened under pressure , ' Astrophysical Journal , ' vol. XIX .
, , 1904 ; ako , here weakened .
EFFECT OF PRESSURE UPON ARC SPECTRA.\mdash ; NICKEL .
211 MITCHELL* attributes the occurrence of enhanced lines in the chromospheric spectrum partly to the reduction of pressure consequent upon the greater altitude and partly to the presence of hydrogen there .
This will be further discussed in a subsequent paper .
7 .
The Displacement of the Lines : ( 1 ) Measurement of the Plates.\mdash ; The bulk of the measurements were made by Mr. F. E. PEARSE , for whose assistance I am indebted to the Government Grant Committee .
The photographs were placed in the fixed carrier of a modified Hilger photo-measuring machine in which the movable part was the microscope which was controlled by a screw whose drumhead reading could be estimated to the thousandth part of a millimetre .
In order that lines of various breadths could be measured the microscope could be converted into a telescope and a considerable range of magnification achieved .
There were two pairs of parallel wires ofdifferent intervals in the eye-piece , either of which could be set perpendicular to the direction of travel of the slide and parallel to the spectrum lines .
This latter operation had been found difficult in previous work , so the later photographs had been taken with a shutter , in which short slits had been cut to allow the top and bottom of the comparison lines to affect the photographic plate above and below the central strip .
In each 50 cm .
there were three such extra pairs of slits , each about 2 cm .
long , e.g. , lines 15 , 16 , 17 , 18 , Plate 1 .
It was always possible to find one at least of these in the range of spectrum upon the measuring machine , and the parallelism of the cross wires was consequently attained with ease and considerable accuracy .
Two readings were taken with the plate placed with the red on the right-hand side , the setting being first on the upper and then on the lower half of the line under pressure ; the plate was then reversed and the readings repeated .
Four readings were thus invariably taken .
In many instances others were made .
The readings were checked by the writer , who made a point of measuring each line at some one pressure .
As a rule these readings were not included in the mean results , because there was some personal equation in the measurements , and it seemed best to have a homogeneous set of readings made by individual because these are then more strictly comparable with one another .
It is interesting to note that even though different observers may obtain different absolute values for the displacements upon a single photographic plate , there is usually agreement between the relative values of their measurements of the displacements of different lines .
For instance , the groupings are usually the same and also the ratios of the mean displacements of the groups .
The order of accuracy obtained is shown by the following , Table II .
, in which a few readings taken at random are reproduced .
They illustrate the agreement of Mr. PEARSE'S readings amongst themselves and with those of the vriter .
*MITCIIELL , ' Astrophysical Journal , vol. XXXVIII .
, p. 407 , 1913 .
' .
EFFECT or PRESSURE UPON ARC SPECTRA.\mdash ; NICKEL .
213 Except in a few instances the agreement is good .
When there were differences Mr. PEARSE made additional readings ; from these and from the measurements of the displacements at other pressures it was usually possible to decide upon the more probable value .
Experience in the measurement of displacements under pressure clearly shows that the personal equation of a computor is not a fixed quantity ; there is a tendency for a novice to record values that are too high , and it is found by experience that an interval may make a considerable difference in one 's judgment of a set of displacements .
instance , the writer measured some lines before and after a voyage of some months ' duration , and found marked differences in the readings , the second set being only about 70 per cent. of the first .
There was , however , excellent agreement between the rate of displacement with wave-length for lines with the same type of intensity curves .
It is on this account that the writer does not wish to lay too great a stress upon the absolute values of the displacements for any one metal .
The accuracy of the relative values for different metals depends also upon the shapes of the intensity curves of the lines ; if these are similar they are more likely to be reliable .
In measuring displacements it is very important that the photographs shall be illuminated by a source of constant brilliance , and for this purpose measurements made in artificial light are more constant than those made in daylight of variable intensity .
( 2 ) Description of of Displacements ( 1 to 110 Atmospheres).\mdash ; Table III .
gives in thousandths of an ngstrom Unit the value of the displacement of each line at the pressure stated at the top of each column .
The first column contains a list of the arbitrary numbers assigned to the lines , the second the wave-lengths of the lines according to HASSELBERG .
The displacements measured for various pressures follow in successive columns .
Reversed lines are indicated in the manner stated in Section 5 , p. 209 .
That the displacement increases with the pressure is at once evident .
The second half of the table contains the displacements per atmosphere in thousandths of an Angstrom Unit , the readings being obtained by dividing those in the first part of the table by the excess pressure above that of one atmosphere .
A column is devoted to the Mean Displacement per atmosphere in thousandths of an ngstrom Unit , to which reference will be made later ; and the final columns which contain the quotient obtained by dividing the Mean Displacement per atmosphere respectively by the first power , square and cube of the wave-length of the line , will also be the subject of subsequent discussion .
VOL. CCXV .
has been questioned , but in the writer 's opinion they are real and not due merely to unsymmetrical broadening ; it is true that the negatively displaced nickel lines are more broadened towards the violet than the red , but the broadening takes place about a negatively displaced position .
Several theories can explain qualitatively how displacements and unsymmetrical broadening towards the red may be accounted for ; the displacement of a line towards the violet should not be more difficult to explain than an unsymmetrical broadening in that direction ; the latter phenomenon is unquestionably true .
The following are the displacements of lines towards the violet measured by Mr. PEARSE and by myself The photographs were taken when the pressure of the air was 10 atmospheres:\mdash ; Line 243 is apparently reversed and displaced towards the violet , but this is found to be due to the strengthening pressure of a faint lin on its violet edge .
( 5 ) Displacement .\mdash ; In Diagram 1 lines are drawn connecting the different of the displacements at different pressures of a few of the spectrum lines dealt with , and each line represents the behaviour of one spectrum line .
The proportionality between the displacement and the pressure is apparent from the diagram , and is approximately linear .
In Section 7 , p. 224 , this point is further examined .
The diagram further illustrates the fact that the lines are capable of resolution into two groups according to their rates of displacement , a feature which is more fully treated later .
The diagram includes both reversed and bright lines , but does not distinguish between them .
Without exception the former fall into the group with the smaller displacement .
Lines displaced towards the violet are not included in this diagram .
( 6 ) Comparison with Previous Observations.\mdash ; The displacements of certain nickel lines have been observed by HUMPHREYS and MOHLER* at pressures , and * HUMPHREYS and MOHLER , ' Astrophysical Journal , ' vol. III .
, p. 114 , 1896 .
pressures from 1 to 200 atmospheres , where it was shown that the pressure-displacement curves were slightly concave to the axis of pressures and where the curves representing the mean displacements per atmosphere had a general downward tendency as the pressure increased\mdash ; as in this research .
Dealing with the spectrum produced by the copper arc when subjected to the highest pressure , the same features appeared , so that the evidence favours the general conclusion : " " That though the relationship between the pressure and the displacement is approximately linear , the displacement does not increase quite as rapidly as the pressure ( 8 ) The Relation between Displacement and length.\mdash ; In Diagrams 4 and 5 each black circle represents the mean displacement per atmosphere of the nickel line whose wave-length is given by the horizontal scale , the data being derived from Table III .
Inasmuch as the displacements at 10 atmospheres pressure are dispro- portionately large they are treated separately in Diagram 5 , whereas they are excluded froJn the calculations of the mean displacements which are plotted in Diagram 4 .
The prominent feature of these diagrams is the increase of the ments as the wave-lengths increase , but the division of the lines into two groups is also indicated .
Treating the diagrams critically it is scarcely open to doubt that the displacement is dependent upon the wave-length , though Diagram 4 alone is perhaps not conclusive in this respect as there are not many lines in the region 3900 to 4200 .
Many more lines have been measured at 10 atmospheres pressure than at higher pressures , so Diagram 5 is able to provide more information about this region of the spectrum , though on the whole the values are not so reliable since they are derived from readings at only one pressure .
Granted then that the occurrence of larger displacements in the less refrangible parts of the spectrum is not fortuitous , it remains to discuss the actual relationship between these two variables .
The diagram at 10 atmospheres points to a steep descent which might be regarded as approximately linear if it were not that it would involve the displacements becoming zero and subsequently negative in the more refrangible regions of the spectrum .
Though negative values for displacements due to pressure have been recorded , the crossing of the axis has not been observed in any spectrum , and the asymptotic trend of the black dots in Diagram 4 is contrary to this occurring in the case of nickel .
This conclusion is supported by HUMPHREYS ' measurements of lines of smaller wavelength down to which are included as open circles in the diagram .
Though the majority of these readings are based upon observations at only one high pressure ( either 42 or 101 atmospheres ) they are in such good agreement with the results of the present research as to leave no doubt that the axis is not crossed in this region of the spectrum .
If we make the assumption that the origin is on the curve relating to displacement and wave-length , we can at least say that it is not contradicted by the results of this research .
On the assumption that the graph is of the form the values of have been calculated ( excluding measurements at 10 atmospheres ) ; they will be found in the final columns of Table III .
In the discussion upon the copper arc , a test of the rate of displacement with wavelength was made by finding the value of which enabled the quantities to be most distinctly separated into groups .
This did not appear so likely to yield promising results in the case of nickel , because in Diagram 1 two groups are remarkably distinct without reference to the wave-length , a point which is further illustrated in Diagram 6 , fig. 1 , in which each number .
represents the presence in the spectrum of a line with a mean displacement corresponding to its position upon the horizontal scale .
The distribution resolves itself into a well-defined group with a mean displacement per atmosphere of about thousandths of an Unit and another very diffuse and ill-defined group with a flat-topped maximum extending from about 9 to 12 thousandths of an ngstrom Unit .
It might be argued that the probable errors of the measurements of the different types of line are in accord with this grouping , because those lines whose displacements are small have a smaller probable error than those.with larger values , the latter being in general ( but not invariably ) associated with a greater width of the lines .
But though this is a consideration which must be given due weight , to regard it as satisfactorlly accounting for the distribution in fig. 1 would be to disregard the significance of the tendency for the displacement to increase with wave-length which has been demonstrated in Diagrams 4 and 5 .
It is further important to note that lines as 208 , 212 , 217 , 228 and 238 do not clearly fall within either group , also that the line 247 is assigned to the group with large displacement in spite of the fact that it is reversed unsymme , and would not be expected , in accordance with previous experience of similarly reversed lines in the iron spectrum , to belong to a group with the highest rate of displacement .
Plotting against wave-length , fig. 2 is obtained , and the lines 208 , 213 , 217 , 228 and 247 occupy anomalous positions .
There is still less reason for regarding this distribution as satisfactory , Fig. 3 shows the distributions of the values of when the lines 247 , 219 , and 217 form a group by themselves , though 208 has attached itself to the outskirts of the first group .
In fig. 4 , in which the values of are given , these lines have been absorbed by Group 1 , and the resolution into two groups is clearer .
We note also that the second group is more compact ; this is obviously the most satisfactory diagram , and we may therefore conclude that as far as values above 10 atmospheres are concerned , the rate of dispJacement is not far removed from that of the cube of the wave-length .
It may appear remarkable that any order can be arrived at considering the apparently chaotic distribution of dots in Diagram 4 in the region of ave-lengths 4200 to 4500 , and it is necessary to emphasize the fact that the method consists in the elimination of the dots with displacements varying from 5 to 8 thousandths of an ngstrom Unit from adequate correction could be applied to the measurement of the lines in the neighbourhood of to 4400 , which appear abnormally high in the diagram , and there is little doubt that the readings are in fact rather too large .
It has been stated that for the sake of uniformity Mr. pIBARSE'S readings have been used almost exclusively in preparing the tables , my own readings of the lines serving as a check upon them .
The check readings in this region indicate that the measurements of the lines in region 4800 to 4900 are to those in region 4400 as 5 is to 4 .
This would improve Diagram 4 , and also make the second group in Diagram 6 , fig. 4 , more compact by reducing the displacements of lines 160 , 167 , 172 , 174 and The above argument is based upon the resolution of the nickel spectrum into only two groups , but reasons will be given in the next section for distinguishing between three groups .
If this is the case we must consider whether the isolated lines between the two main groups could form a group by themselves .
If the displacement varies with the square of the wave-length , Diagram 6 , fig. 3 , indicates that the following lines would constitute ; and 247 , The last-named is not of the same nature as the first two , and there is a considerable separation between 213 , 217 , and the similar lines 87 and 165 which should be associated with them according to their behaviour and general appearance .
Five more lines , 10 , 19 , 21 , 26 , 31 , which resemble them by being nebulous at atmospheric pressure , may also be associated with them .
It will be seen from what follows that they form a more compact group when classified according to the cube of the wavelength .
Though the determination is by no means conclusive , the balance of evidence favours this rate of variation of the displacement .
It is important to see to what extent the readings at 10 atmospheres confirm the above conclusion .
EFFECT OF PRESSURE UPON ARC SPECTRA.\mdash ; NICKEL .
In Table VIII .
the values of the mean displacements at 10 atmospheres have been divided by the square and cube of the wave-lengths , and in Diagram 7 , figs. 1 and 2 , these are shown in statistical form .
Many more values are now included , and each diagram is now very irregular .
They are not as trustworthy as the previous diagrams for the purpose of ascertaining the rate of displacement with wave-length , but they serve as a guide to the classification of lines which cannot bs measured at pressures higher than 10 atmospheres .
The separation into groups is clearer in the diagram , which consequently favours this ratio .
If , however , the lines be sorted out diagram in accordance with their grouping at pressures above 10 atmospheres , it becomes much clearer , as in fig. 2 , and we see that the confusion between values 8 and 10 on the horizontal scale is due to the overlapping of two frequency curves with rather large variations .
The distributions for now appear about equally probable for the readings made above 10 atmospheres .
The rate of increase of the displacement with the wave-length has been a matter of previous investigation by the writer ; the copper arc spectrum*yielded the result that the displacement was proportional to a power of the wave-length " " at least as great as the power and possibly as hi gh as the sixth The experiments upon the gold arc under pressure favoured a dependence upon third power of the wavelength .
Subsequent important experiments by GALE and upon an extended region of the iron spectrum under a pressure of 9 atmospheres support this conclusion , and so do the results of the present research .
The bearing of this upon the spectra of novae has been discussed by the writer in a paper entitled " " The Spectra of Novae and the Pressure Effect.\ldquo ; S It is there shown how the Doppler and pressure effects may be distinguished .
An interesting feature of the nickel lines is the gregarious tendency of lines with large displacements , almost all of which occur within a region of the spectrum between and .
Reference to Table X. which gives lists of lines of similar appearance\mdash ; reversed lines , nebulous lines , \amp ; c. , indicates that they occupy only limited regions of the spectrum .
A similar tendency for the iron lines of the same group to congregate was recorded .
8 .
Resolution of the Nickel Spectrum into Groups.\mdash ; The lines may be grouped in different ways\mdash ; according to their displacement , broadening , intensity , reversal , \amp ; c. , but of these the most important is the first-named .
It has already been seen in the previous section that it is possible to resolve the nickel arc spectrum into two main groups according to their displacement , and that the first of these may be capable of further sub-division .
The following table has been compiled from the available data to indicate the general nature of the lines belonging to the two groups .
The line and its wave-length are given in the first two columns , in the third column ' ' \ldquo ; denotes that the line was nebulous at atmospheric pressure , in the fourth column indicates that it became nebulous or diffuse under increasing pressure ; " " \ldquo ; in the next column indicates that the line was classified by HASSELBERG as " " sharp\ldquo ; at atmospheric pressure .
The changes in relative intensity are given by " " \ldquo ; or " " \ldquo ; in the next two columns .
The broadening and reversal are treated in the two following columns , and the last two indicate the group into which the line falls in the diagrams , Diagram 6 , fig. 4 , and Diagram 7 , fig. 2 .
The first of these is the more reliable as it includes a larger number of readings , but the second is useful in supplementing the information supplied by the preceding column ; it is derived from observations made at 10 atmospheres only .
* DUFFIELD , ' Phil. Trans. Roy .
Soc vol. 209 , p. 205 , 1908 .
DUFFIELD , ' PhiL Trans. Roy .
Soc vol. 211 , p. 33 , 1910 .
GALE and ADAMS , 'Astrophysical Journal , ' XXXV .
, p. 10 , 1912 .
Through a misapprehe1lsion , these authors quote experiments upon vanadium by ROSSI as supporting this conclusion .
S DUFFIELD , ' Monthly Notices , R.A.S. , ' 73 , 1 ) .
631 , 1913 .
EFFECT OF PRESSURE UPON ARC SPECTRA.\mdash ; NICKEL .
For the purpose of comparing the characteristics of lines longing to these two groups Diagram 8 has been prepared .
It closely resembles Diagram 6 , fig. 4 , but instead of denoting the es by means of numbers , their characteristics are quoted in accordance with the abbreviations given below the diagram .
It deals only with lines whose displacements above 10 atmosphere have been determined .
It is clear that Group I. is by no means homogeneous , and that it may be divided into two Subgroups .
and IB .
The characteristics of the lines falling into the first part of Group I. ( Group IA .
) 1 .
Slight broadening .
Nearly al. 2 .
Sharp at atmospheric pressure .
3 .
Reversed lines are very nearly symmetrical .
4 .
Two lines are strengthened , none are weakened .
The characteristics of lines falling into the second part of Group I. ( Group IB .
) .
Greater broadening .
[ Many are described as moderately broadened , some as greatly broadened , though several are described as slightly broadened .
] 2 .
Nebulous at one atmosphere or nebulous at 10 atmospheres .
3 .
The reversed lines are less symmetrical .
4 .
Tendency to weaken under pressure .
The characteristics of the lines falling into Group II .
1 .
Great broadening .
2 .
Their unsymmetrical appearance under pressure .
They are all oadened greatly to the red side .
3 .
None are reversed under pressure .
If we include in our survey the readings made at 10 atmospheres more lines come under observation ; among them are several that are nebulous , of these most are weakened , e.g. , 82 , 135 , 140 and 145 , \amp ; c. , but one , 133 , has been classed as strengthened .
We also note that certain other lines which are nebulous under pressure are included in Table IX .
, namely , those which are broadened towards the violet , 232 , 233 , \amp ; c. These nebulous lines all occur in Group I. Group II .
is similarly extended by this process , and it now appears to include some lines which become nebulous under pressure ; these differ from those assigned to Group I. in appearing to be broadened to the red side , but there is a certain amount of doubt about the accuracy of their measuremen , and therefore of their position in Group II .
Adopting the law that the displacement varies as the cube of the wave-length for lines of the same group , we see from Diagram 6 , fig. 4 , that the maxima in the distribution curves occur at and , which gives a ratio of 1 to for the maximum displacements of the two groups , a result very different from that found for the groups in the iron spectrum .
If , however , Group I. be sub-divided into Group IA .
and Group IB .
, we may take the similar nebulous lines 19 , 21 , 26 , 31 , 87 , 165 , 213 , 217 , as forming the more displaced group for which the value of is about 5 The corresponding value for the remainder is not very different from This would give a ratio of displacement for the three groups , which is close to that found for the iron spectrum , namely , 1 : , and not far from 1 : 2 : 4 , but it would only be justified if the division into three groups were an established fact .
It may be pointed out that the diagram ( fig. 2 ) , favours this relationship ( the ratios would be ) , but , as has already been discussed , the diagram labours under the disadvantage of including 247 in the intermediate group , dissociating it from lines which it more strongly resembles , and it separates the similar lines 87 , 165 , 213 , and 217 .
These difficulties are obviated in the curves .
We are thus faced with the possibility of the following representations of the groups : Group I. Group II .
in which case the ratio is 1 : , or if there be three groUps:\mdash ; Group IA .
, ' or IB .
, II .
, In either case the ratio is approximately 1 : 2 : 4 .
The writer inclines to the view that this is the more probable ratio between the different groups , but the determination is open to doubt .
inteltsity is calculated , these are show1l by the circles .
Since the high-valued lines 10 , 247 , 213 , 217 , and 87 possibly form a separate sub-group , more cogent evidence is perhaps afforded by the absence of low values for the displacement of lines of great intensity .
It relnains to be proved whether the phenomenon is subjective or objective , whether it may be explained by a tendency for the computor to read high values when the density of the silver deposit is great , or whether the amount of energy involved in a line does actually influence the displacement .
One reason which may be urged for regarding the phenomenon as independent of the observer is that intense lines are usually broad and consequently more closely resemble lines under higher pressure ; these have , however , been shown to be relatively less displaced than lines at lower pressure , there would not seem to be any satisfactory reason for regarding the measurements of the displacements of intense lines as being too large .
If the energy is responsible for the magnitude of the displacement one further enquires whether the increased intensity is due to a larger number of vibrating particles , or to their possessing greater amplitudes or to a combination of the two ; thus if the phenomenon is objective it involves the dependence of the displacement either upon ( 1 ) the density of the particular atom or modification of the atom responsible for each line , or ( 2 ) upon the amplitude of its vibration .
In any case the dependence of the displacement upon the energy involved in a line is not the only factor concerned , since some faint lines have very large displacements .
The former would seem the more probable , but attempts to show that the displacement of the line of an element is due to partial pressure have hitherto not been successful .
This will be discussed elsewhere ; the present discussion suggests that the partial pressure effect due to the density of similar atoms is superposed upon an effect due to the total pressure .
10 .
Foreign in pectrum of Nickel.\mdash ; In addition .
to the nickel lines whose displacements have been measured , a large number of lines due to impurities appear in the spectrum , especially between and .
They have been measured , and their displacements will be found.in the following table .
As they have been examined at several pressures , it should be possible to ascertain whether the " " density\ldquo ; of the element has any effect upon the displacement by comparing the displacements of the lines produced respectively by pure poles and by small traces of the element .
( 1 ) The Influence of the Density of the Material in the Arc.\mdash ; In Table XII .
the measurements made during the present research upon the lines due to traces of iron in the nickel poles are compared with the displacements of the same lines produced in an arc between solid iron rods .
At the top of each column will be found the pressure at which the examination was made , and opposite each spectrum line the ratio of the displacement of the diluted to that of the undiluted material .
It will be seen that if the readings at 10 atmospheres be left out of account ( they are unduly high in nickel as we have already shown ) there is a smaller reading when the material is diluted .
TABLE XII .
The measurements of the displacements of the lines due to pure iron were made some years ago both by the writer and by assistants , while those due to a small trace of iron in the nickel poles were made Mr. PEARSE .
In order to see if a comparison between these readings was legitimate , Mr. PEARSE measured one of the original pure iron arc photographs ( at 70 atmospheres pressure ) , and his readings agreed so excellently with the earlier ones that there was no hesitation in regarding the two sets as strictly comparab]e .
The ratio of the displacements with diluted to those with pure iron is , with the exception noted above , less than unity , suggesting that the density of similar atoms influences the displacement ; if the displacement depends upon the proximity of similar vibrating centres it is to be expected that the displacements would be more EFFECT OF PRESSURE UPON ARC CTRA .
251 .
marked at high pressures , since increasing the pressure reduces the mean free path between similar molecules just as increasing the total number of molecules does .
The evidence thus favours the amount of the displacement being dependent in part , at least , upon the amount of material present , a conclusion which is in keeping with that arrived at from the consideration of the variation of the displacement with the intensity of the line .
The chief sources of contamination of the nickel poles are iron , cobalt , and manganese .
There is some doubt in one or two instances as to the origin of the lines ; lines 136 and 137 for instance are very much alike under pressure , both in amount of displacement and in their intensities , but one is ascribed to nickel and the other to iron .
The feature of the lines due to impurities is that they remain fairly compact even at very high pressures , and do not spread out to the same extent or become so foggy as the nickel lines .
It is interesting to note that all the lines of the highest intensity due to iron do not appear in the nickel spectrum .
Some lines seem to characterize the spectrum due to only a small quantity of material .
For example , of the two iron lines ( 10 ) and , only one , the latter , appears in the nickel spectrum .
Of the two iron lines ( 156 ) at and , both classed as of intensity 10 , only the latter appears , similarly , only the last-named of the two iron lines ( 10 ) , is visible upon the plate .
It is important to note that it is the line which is self-reversed which most readily shows itself , it is the less refrangible line in the pairs quoted .
It has always been surprising to the writer that no dependence upon the density of the material manifested itself in previous experiments .
On one view one might expect the general magnetic field of the surrounding atoms to influence the frequency of any particular atom , and since , presumably , this general field depends upon the nature of the atom the amount of material present or the nature of the surrounding gas should have some effect , but nothing definite has hitherto been observed .
On another view , the specific inductive capacity of the medium in which the atom under consideration is immersed should similarly affect the frequency , but this has been tested without positive result .
The explanation seems to be that in the arc the isolation of a molecule is never actually accomplished , that the vaporization of the metal , even if only a trace of it be present , involves the liberation of an immense number of atoms of that element all in close proximity , so that this incandescent mass behaves very much as though it were isolated from the other materials in the arc , rendering it very difficult to influence the immediate environment of an atom , since only the few atoms on the outskirts of this mass are affected by the inductive properties of the surrounding gas ( or other metallic vapour produced by the vaporization of the poles ) , consequently the predominant frequency is that of the atom surrounded by similar atoms .
DR. W. EOFFREY DUFFIELD ON ARC SPECTRA.\mdash ; NICKEL .
Decreasing the amount of impurity present would make the outer portions of its incandescent vapour relatively more important , and increase of pressure would achieve the same end by decreasing the average distance of each vibrating centre from the molecules of the surrounding gas which might then exert an influence which is negligible when the pressure is low .
We thus expect to find a reduced displacement when the material is greatly diluted and a relative reduction with increased pressure .
The former is supported , and the latter suggested , by Table XIL We have assumed that the surrounding atmosphere is less effective in producing displacement than an atmosphere of similar molecules , but it should not be impossible for the opposite to be the case , according to the nature of the substances employed , unless it can be proved that similar molecules alone have influence upon a radiating molecule .
Approaching the problem of the structure of the arc from a different direction , the writer has been impressed by the importance of the part played by the surrounding gas in maintaining the arc ; it would seem require that each metallic atom is , for a brief interval at least , associated with one of the atoms of the surrounding gas so that something akin to chemical action takes place between them ( or at least involves the influence of what may be called chemical affinity ) .
If this is the case it would appear contrary to the view of the density effect just put forward , since that does not contemplate the commingling of th - individual atoms of the metal with those of the surrounding gas .
But it is further possible that the spectrum line is due , not to single atoms but to a system , such as one consisting of a metallic atom combined or interacting with one atom of the surrounding gas ( the function of the latter being , by its interactions with the atom , to excite it to emit its characteristic radiation ) , and that such systems form the aggregate already alluded to , and that the frequency of the resultant spectrum line is characteristic of this particular atom system in its environment of similar systems .
These systems would not be interfered with by foreign systems until either the reduction of the amount of material below a certain minimum amount or the increase of pressure made the proximity of foreign atomic systems relatively more important .
Possibly this consideration is responsible for the decrease in the rate of increase of displacement with pressure when the latter is high .
The writer is indebted to Dr. SCHUSTER , at whose suggestion the series of experiments upon the effect of pressure upon spectra were begun some years ago , and to ERNEST RUTHERFORD in whose laboratory the photographs were taken , and expresses his thanks to them for having placed the necessary apparatus at his disposal .
The photographs were measured by Mr. PEARSE in a careful and thorough manner .
Part of the expense of this research was defrayed by a grant from the Government Grant Committee .
\mdash ;

