30 On the Electrical Resolution and Broadening of Helium By Thomas R. Merton , D.Sc .
, Lecturer in Spectroscopy at the University of London , King 's College .
( Communicated by Prof. A. Fowler , F.R.S. Received April 20 , 1918 .
) The translatory motions of the radiating atoms in a luminous gas , which furnish a complete and satisfactory explanation of the widths of the spectrum lines which are emitted under certain specified conditions of excitation , are unable to account for the vastly greater broadening which occurs when the spectrum of a gas at a moderately high pressure is produced by the passage of a condensed spark discharge .
The suggestion of Stark* that the broadening which occurs under these conditions of excitation is intimately connected with the resolution of lines into components by the electric field , being in fact due to the electric field of neighbouring charged particles on the radiating atom , has been shown by the writerj* to be in qualitative agreement with the broadening which occurs under these conditions in the lines of the Balmer Series of hydrogen , and in a subsequent investigation with Prof. J. W. Nicholson , | it was found that a quantitative analysis of the structure of the line Ha was in complete harmony with Stark 's observations of the electrical resolution of this line , on the assumption that the broadening was due to this cause .
It was also found that certain effects in helium and lithium were in qualitative agreement with the hypothesis .
Without entering into a detailed discussion of the results obtained in the case of hydrogen , it may be stated that the most striking phenomenon observed was that the line Hp presented an appearance resembling that of a reversal , having a dark narrow space in the centre of the broadened line , which was symmetrical , as in the case of the lines Ha and Hy .
The existence of this dark space is in harmony with the fact that when the line is resolved by an electric field , there is no component ( or possibly a very weak component ) occupying the position of the unresolved line in planes polarised either parallel or at right angles to the electric field .
On the other hand , both Ha and Hy showed maxima in the position of the unresolved lines , which is in agreement with the presence of strong undisplaced components in the electrical resolution of these lines .
In the case of helium , as Stark has pointed out ( loc. tit .
) , both the electrical * ' Electrische Spectralanalyse Chemische Atom , ' Leipzig , Hirzel , 1914 .
t ' Roy .
Soc. Proc. , ' A , vol. 92 , p. 322 ( 1916 ) .
X 'Phil .
Trans. , ' A , vol. 216 , p. 459 ( 1916 ) .
i Electrical Resolution and Broadening of Helium Lines .
31 resolution and the broadening are in certain cases unsymmetrical .
The phenomena observed are thus concordant , but some recent results of Takamine and Yoshida* have provided the material for a closer and more made a detailed measurement of the electrical resolution of the more prominent lines in the spectrum of helium , and have confirmed and extended an important discovery of Brunetti , f which relates to the existence of isolated components in the resolution of certain lines .
They have shown that the effect is not merely a splitting of the lines into components , the separation of which increases with the electric field , but that in some cases an isolated component appears at a certain strength of field and is further resolved or displaced as the field is increased .
It is evident that in those cases where an isolated component appears at such a distance from the parent line as not to overlap to any great extent ( at the strength of field at which it appears ) what may be termed the normal components , it should be possible to directly observe the isolated component in the structure of the broadened line .
The method of experiment in the present observations has been identical with that used in previous investigations , and consisted of a prism spectrograph with a neutral-tinted glass wedge mounted in front of the slit , the spectrum thus consisting of lines which are bright at the end corresponding to the thin end of the wedge and which fade away in the direction of the dense end of the wedge .
The theory of this method and the apparatus used have already been fully described ( .
cit. ) , but the technique has been considerably improved , and it is to this that the observation of much of the detail recorded has been due .
Over the range of wave-lengths covered by a single broadened line , the density of the wedge may be taken as effectively constant , and since the observations under discussion refer to the structure of the individual lines , the precise constants relating to the wedge are irrelevant .
A vacuum tube of the conventional Pliicker form , containing very pure helium , at a pressure of about 1 mm. , was set at a distance of about 1 foot from the slit of the spectrograph , and the image of the capillary was brought to a focus on the slit by means of a convex lens , the wedge being in position in front of the slit .
Exposures were made for about six hours , the vacuum tube being excited by means of a fairly heavy discharge from an induction coil , which was capable of giving a 10-inch spark in air , with a condenser in parallel and a spark-gap * ' Mem. Coll. Sci. , Kyoto , ' vol. 2 , p. 6 ( 1917 ) .
t ' Rend .
d. Lined , ' vol. 23 , p. 719 ( 1914 ) .
rigorous enquiry into the connection between the two effects .
These investigators , using the method of Lo Surdo , have Dr. T. R. Merton .
On Electrical \#166 ; of about half a centimetre in series .
Under these conditions , the spark line at A , 4686 is well developed , and many of the " arc " lines show a marked broadening .
For the full elucidation of the structure of these broadened lines , it will be necessary to employ a much greater dispersion than I was able to use in these experiments , but it was nevertheless possible to observe a number of phenomena which may have an important bearing on the theory of the broadening .
It may be pointed out that on the photographs obtained in this way a line whose physical width is very small in comparison with the resolving power of the spectroscope nevertheless appears as a very acute isosceles triangle on the plate , owing to effect of irradiation , * but most of the lines were broadened to an extent which completely eliminates any errors which might arise from this cause .
The theory of this effect has been discussed in a previous communication ( loc. It has not been found possible to prepare a photographic enlargement suitable for reproduction .
I have , therefore , prepared a diagram showing the \#166 ; contours of the patterns obtained , by copying in Indian ink with a camera lucida the contours seen on an enlarged photograph .
Although the diagram , fig. A , is doubtless inaccurate as regards the finer details , it is believed to be a trustworthy representation of the principal features of the structure .
The structure of the individual lines may now be considered in relation to the results obtained by Takamine and Yoshida for the electrical resolution .
The lines peculiar to the series of helium ( as distinguished from parhelium ) are close doublets , but as their separation is too small for resolution with the spectrograph used , this circumstance has in itself no effect on the resulting structure .
W 5048 , 5015.\#151 ; For these two lines , the broadening is very small under the conditions described , the former being broadened to the red and the latter to the violet .
These two lines were not resolved by Takamine and Yoshida , but were found to be displaced to a very slight extent in the electric field , in the same directions respectively as the assymetry of the broadening .
A 4922.\#151 ; The structure of this line is complex .
There is a minimum resembling a very narrow reversal on the more refrangible side of the principal maximum .
This minimum is just resolved on the photographs.f Beyond the second maximum the line falls away rapidly in intensity , and then rises to a third rather indistinct maximum .
It is difficult to determine the precise distance of these maxima from the parent line , but the former is of the order of 2A .
, and the latter about 14A .
It will be seen that the general * Cf .
Mees , ' Boy .
Soc. Proc. , ' A , vol. 83 , p. 10 ( 1909 ) .
t The distance between these maxima in fig. is exaggerated .
i X=4472 Resolution and Broadening of Helium Lines .
VOL. XCV.\#151 ; A. D Dr. T. R. Merton .
On Electrical contour of the line and the third maximum at 14A .
separation are in close agreement with the results of Takamine and Yoshida , reproduced in fig.-B , but the second maximum of about 2A .
separation is not accounted for by their results .
This is the only exception to a complete agreement between their results and mine , but it will be shown later that this component at about 2A .
on the more refrangible side of the parent line may possibly be accounted for in a simple manner .
X 4713.\#151 ; This line is broadened towards the red , whilst it remains sharply defined on the more refrangible side .
It is not included in the lines measured by Takamine and Yoshida .
X 4686.\#151 ; This line is not greatly broadened under the conditions of experiment , but it appears to broaden symmetrically , in agreement with the result for the electrical resolution .
XX 4519 , 4046.\#151 ; These two lines were first observed by Koch* and belong to a combination series .
The former was observed by Takamine and Yoshida to be slightly displaced towards the red in the electric field , and for the latter no observations are recorded .
On my plates 4519 is broadened towards the red , and there are distinct indications of a similar effect for X 4046 which can just be seen on the plate .
A 4471.\#151 ; The structure of this line is remarkable .
There is a very narrow minimum somewhat resembling a reversal , but localised on the more refrangible side of the parent line .
The dispersion used was just sufficient for this minimum to be clearly shown.f It does not , of course , extend right up to the base of the wedge owing to irradiation .
This result is in striking agreement with the electrical resolution .
The line is broadened on the violet side of the minimum to a somewhat greater extent than on the less refrangible side .
X 4388.\#151 ; There is a sharp maximum , with diffuse wings which indicate unsymmetrical broadening to the violet .
For this line , Takamine and Yoshida find an isolated component on the more refrangible side .
I have not been able to establish the existence of any effect due to this component in the broadened line , but it will be seen ( fig. B ) that its effect in all probability would be masked to a great extent by the normal components .
X4121 broadens unsymmetrically to the red .
This line is displaced to the red in the electric field .
X 4026.\#151 ; In this line there is a sharp maximum with diffuse wings similar to X 4388 .
There are indications of asymmetry to the violet .
X 3965 shows a marked asymmetry to the violet , and is sharply defined on * ' Ann. d. Phys. , ' vol. 48 , p. 98 ( 1915 ) .
t The width of the minimum is exaggerated in the figure .
i Resolution and Broadening of Helium Lines .
the more refrangible side .
This line is not included in the published measurements of Takamine and Yoshida , but Prof. Takamine has very kindly sent me some of his photographs , which include this line , which shows a distinct displacement to the violet in the electric field .
X 3888.__This line shows no definite broadening under the conditions of my experiments .
The close resemblance between the contours of the broadened lines and the electrical resolution seems to establish , without any doubt , the intimate relation of the two phenomena .
Without in any way implying that , under appropriate conditions , there may not be other circumstances which give rise to a broadening of the spectrum lines , it seems extremely probable that , under the conditions of the experiments here recorded , the electrical resolution is by itself able to account satisfactorily and completely for the structure of the broadened lines .
It has been shown in a previous investigation ( loc.cit .
) that the potential gradient in the capillary does not rise tc sufficiently high values under these conditions to contribute appreciably to\gt ; the broadening , which must be referred to the electric field of neighbouring charged particles on the radiating atoms .
From the magnitude of the broadening , it at once follows that some of the radiating atoms are submitted to forces far in excess of any that have hitherto been applied to the study of the electrical resolution , either by the method of Lo Surdo or that of Stark .
This is consistent with the appearance of such lines as X 4686 and the series lines XX 4519 , 4047 , which make their appearance in a vacuum tube in those regions close to the cathode where the potential gradient is highest .
It seems probable that the condensed-spark discharge is still the most energetic method of excitation which we are able to apply in the laboratory .
Its weakness lies in the very heterogeneous distribution in the intensity of the force to which different atoms are subjected , with the result that lines which are strongly developed , in so far as their energy-content is concerned , way yet be so broadened and diffuse as to escape detection .
Fowler* has drawn attention to a very striking case of this kind in certain lines in the " spark " spectrum of magnesium , which were sharply defined in the vacuum arc , but which were almost invisible in the condensed spark , owing to their diffuse nature in this source .
At higher pressures , and with very intense excitation , X 4686 appears as the broadest line in the spectrum .
This might appear to be incompatible with the relatively small electrical resolution which has been found for this line .
The difficulty is only apparent , and this behaviour is , in fact , to be expected .
It has been pointed out that the magnitude of the broadening implies that , * ' Roy .
Soc. Proc. , ' A , vol. 83 , p. 10 ( 1909 ) .
Dr. T. R. Merton .
On Electrical in the heterogeneous excitation of the condensed-spark discharge , some of the atoms are submitted to electric fields far greater than any homogeneous field that has hitherto been applied in the study of the electrical resolution .
It seems probable that these fields are in many cases greater than the maximum possible field for a line of the " arc " type .
All the evidence goes to show that , since there is a transition with increasing intensity of excitation from the " arc " to the " spark " spectrum , the spark lines may be regarded as a kind of safety valve of the arc lines , that is to say , that the tendency of the atom to emit its characteristic " arc " radiation rapidly decreases when the degree of excitation exceeds a certain value .
At present we know of no helium lines corresponding to a higher degree of excitation than the lines of the X 4686 series ( and the \#163 ; Puppis series ) , and we have therefore no reason to suspect that any theoretical limit is imposed in this way on the extent to which this line can be broadened by the condensed-spark discharge , but the views expressed above do definitely postulate the existence of a practical limit to the breadth of lines of the " arc " series , in so far as the electrical resolution is the operative cause .
Although it has been stated above that the broadening observed is in close agreement with the electrical resolution , this does not in any way imply that the " isolated components " are , strictly speaking , to be regarded as a product of the electrical resolution .
A priori there is no reason to suppose that they are .
It is true that they appear in the neighbourhood of certain lines at a definite field strength , but this specification would apply with equal force to X4686 , or to the lines discovered by Koch , the former of which is well known as the principal " spark line , " whilst the latter belong to a combinatipn series .
It is therefore a matter of interest to consider whether these " isolated components " are not in reality an extension of the helium spectrum , and whether they can in particular be represented as lines belonging to combination series .
In some respects the series most likely to be found are akin to those of the Fundamental type .
Those which can in the case of helium appear in the visible spectrum can be represented by the formula 1/ X =VP1\#151 ; N/ m2 , since the separation of the doublets is very small , where 1 / X denotes the wave-number , and VPi the value of the variable part in the Rydberg or Hicks formula for the Principal series when the term number is unity .
Considering first the case of helium , we find that only one member of this series can be expected to appear .
The member corresponding to m = 3 , which does not seem to appear in series of this type would be at X 5870 , or about 5A .
on the more refrangible side of D3 .
The next member would be at X 4470-9 , and it seems possible that this corresponds to the " isolated component " at X 4470 observed by Takamine and Yoshida , I Resolution and Broadening of Helium Lines .
37 and also found in the broadened line .
The remaining members of this series would be too close to the succeeding lines of the Diffuse series at 4026 , etc. , to be separated from the latter on my photographs .
In the corresponding series in parhelium also only one member could be expected to appear on the photograph at X 4921 .
This does not correspond to the isolated component at X 4909 , but it agrees well with the component at about X 4920 observed in the broadened line .
An exact agreement is not to be expected in cases of this kind .
The suggestion that parhelium may also have series akin to Fundamental series is , of course , at variance with the Ritz specification of such series , which derives them only from spectra consisting of doublets or triplets .
In the case of doublets Ritz defines the Fundamental series as one in which the Rydberg phase y , and the subsidiary constants are the differences of those found in main series , but the actual agreement with observation of established Fundamental series , when calculated on this basis , has never been completely satisfactory , and it is a matter of considerable doubt whether this specification is correct .
If , however , as has been suggested , the Fundamental series are to be referred to a quite unrelated variable part ( where is very small or zero ) , which belongs to the series and is practically the same for all elements , the identification of the line at X 4920 may be justifiable .
With regard to the components found by Takamine and Yoshida at XX 4909 and 4384 , it has not been found possible to represent them in a wholly satisfactory manner as a combination series .
The results may be summarised as follows:\#151 ; 1 .
The broadening of helium lines by condensed spark discharges is in close agreement with the electrical resolution of the lines .
2 .
The " isolated components " in the electrical resolution , which have been recorded by Brunetti and by Takamine and Yoshida , have been found in the broadened lines .
3 .
An explanation is offered of the relative degree of broadening of lines of the " arc " and " spark " type , on the supposition that the latter act as a kind of safety valve to the former when the intensity of excitation becomes very great .
# 4 .
It is suggested that the " isolated components " are not a direct product of the electrical resolution , but are in reality an extension of the helium spectrum .
Two of these lines may perhaps be represented as lines of combination series .
[ Note added , July 4 th .
A line at about X 6624 has been found on a plate taken without the wedge under similar conditions of excitation to those recorded .
This line appears to fall into a series with the lines at 4909 Prof. T. H. Havelock .
and 4384 , the phase y , in the simple Rydberg formula being about 0-02 , and the limit being approximately that of the Diffuse Series of Parhelium , but the measurements of the wave-lengths of the lines are not sufficiently accurate to establish fully their series relations .
A number of other lines have been observed , but an investigation with higher dispersion will be required for a full elucidation of the spectrum .
] Periodic Irrotational Waves of Finite Height .
By T. H. Havelock , F.R.S. ( Received May 21 , 1918 .
) 1 .
The method of Stokes for waves of finite height on deep water consists in working upwards by successive steps from the infinitesimal wave towards the highest possible wave .
Although lacking formal proof of convergence , it is generally accepted that the method is valid , but that it does not include the highest possible wave when the crests form wedges of 120 ' .
For the highest wave itself we have Michell 's analysis by a distinct method , also involving an infinite series whose convergence has to be assumed.* The theoretical position of Stokes ' method has been stated concisely by Prof. Burnside in a recent criticismf:\#151 ; " The complete result would be to express the co-ordinates x and in terms of ( f\gt ; and yjr in the form x \#151 ; \#151 ; \lt ; f\gt ; + be*sin \lt ; \#163 ; -f- X PM sin n\lt ; f\gt ; , y \#151 ; \#151 ; yjr + be*cos \lt ; f\gt ; + 'Z ( 1 ) where Pn ( b ) , Qw ( b ) are power series in b. " These results have a meaning and can be used for actual approximate calculation only , if P " , Qn are convergent power series when b does not exceed some value , say b0 , while for suitable values of b and for real negative values of the series for x and y are convergent .
" Until the form of the power series P " and Q " have been determined , it is impossible to deal with their convergence .
Assuming that they are convergent , it is clear from physical considerations that there must be an * J. H. Michell , ' Phil. Mag. , ' Ser. 5 , vol. 36 , p. 430 ( 1893 ) .
t W. Burnside , ' Lond. Math. Soc. Proc. , ' Ser. 2 , vol. 15 , p. 26 ( 1916 ) .
i

