KN()\\Li:i)r,i:. 



Makcii. 1912. 



THK SPKCTROSCOPK ASII i I ol Mil 



IMPACT THi:<iin. 



Til the Editors of " Knowi.i;|)(;i;. " 



Sirs, — I l).iv<>m)t h.-id thc»pp(irtiinity of cxainiiiiiiK .-md com- 

 pariiiH in ilrlail llic several spoclroscopic observations of Novae 

 as n-Kards the correspondence between them and the deductions 

 of the theory, but it appiars to nie that the main phetiomena 

 arc siifRcicntly striking, and that the general agreement, 

 providing as it does the fundamental working hypothesis, 

 endows the details with an increasing interest. Novae have 

 presented to astrophysicists some of the most bewildering 

 problems in the whole of astronomy. Their sudden appear- 

 ance, followed by declining luminosity and the extraordinary 

 characteristics of their spectra, defied explanation. Yet iii 

 their spectra (considered in conjunction with telescopic 

 observations), was written the secret of their origin. It 

 behoves us to remember that Professor BicUerton's beautiful 

 generalisation was conceived and worked out at a time when 

 observational confirmation was meagre, and this should 

 increase our respect for it in the light of subsequent discovery. 

 I doubt if Professor Bickerton himself ever imagined to 

 what extent his early conception would be enriched and 

 strengthened by continued observation and discovery. 



The spectra of Novae are of supreme interest ; they tell of 

 matter in the crisis of extreme stress, matter under conditions 

 which we cannot reproduce in the laboratory; and every small 

 detail in them is worthy of close study. All these details may 

 not yet be capable of explanation, but the main phenomena 

 arc met by the theory of the third body, and it is surely most 

 logical to adopt as a working basis the one theory w^hich 

 explains the most characteristic features of such spectra. 



Without attempting to discuss the various other aspects of 

 the theory from the standpoint of general astronomy, we ma.\- 

 consider briefly these spectroscopic phenomena on their own 

 merits. The first circumstance which strikes the attention is 

 the emergence from the general blaze of the Balmer series of 

 hydrogen lines, but not as we are accustomed to see them 

 either in the laboratory or in normal stellar spectra. They 

 are wide bands with diffused edges. This appearance has 

 been sufficiently emphasised in Professor Bickerton's own 

 writings, and his explanation of it as a necessary deduction 

 from his theory appears to me to be the only one possible. 



In an examination of the spectrograms, we notice that the 

 width of the bands increases as we proceed along the series 

 towards the violet. This is a necessary consequence of 

 prismatic dispersion, as will be evident by plotting the 

 dispersion curve of any spectrogram giving the normal 

 hydrogen series in a stellar comparison spectrum, e.g., the 

 Stonyhurst spectrogram of Nova Persei with that of 

 o (ieminorum as a reference spectrum. Therefore, to 

 measure the amount of widening, it is necessary first to 

 ascertain accurately the form of the dispersion curve repre- 

 senting the varying degree of dispersion under which the 

 several lines are presented. 



Although great pressure will widen lines, rendering diffuse 

 lines which are normally sharp, it is clearly quite incapable of 

 explaining the widened lines of Novae for the reasons given so 

 clearly by Professor Bickerton in" Kxowi.iiDGl^ '" (September 

 pp. 365. .}66). namely, that the bands remain sensibly constant 

 and bordered on the more refrangible edges by dark absorption 

 lines. In passing it may be interesting to note that in accordance 

 with the Doppler principle, spectrum lines should be widened 

 by reason of molecular motions in a source in a state of high 

 thermal disturbance, but this effect is very small even at high 

 temperatures. 



It is, I think, very desirable that all available spectroscopic 

 observations of Novae should be collected and discussed on 

 the basis of the theory of the third body, especially as regards 

 the careful comp.irison and reduction of all spectrograms. 

 According to the theory, the extreme edges of the widened 

 lines represent the maximum velocities in the line of sight 

 towards and away from the earth of the gaseous shells of the 

 elements to which such lines are spectroscopically attributable. 

 It is important to ascertain whether these velocities are what 

 the respective atomic weights would lead one to expect ; also 



whetlKT lines of any one element all indicate by their 

 measured displacements the same velocity. These and 

 several other points re<|uire investigation, but there are many 

 factors which have constantly to be kept in mind. One of 

 these is the very important ijiieslion of possible elemental 

 dissociation. If this has been claimed for elements present 

 in normal stars, subject to normal temperature conditions 

 consequent on their locality, how much more must the 

 possibility be considered in the case of the violent forces 

 brought into play in stellar impacts. And if this dissociation 

 occurs, then the velocity in the line of sight calculated from 

 the displacement of one line of an element would not 

 necessarily be the same as that deduced from another line 

 normally attributable to the same element. If the third body 

 as a whole be practically stationary in space, then the 

 centres of the widened lines would coincide with the corres- 

 ponding lines of the comparison spectrum ; but if it should 

 happen that there was considerable motion in the line of sight, 

 a complicating factor would be introduced, for in addition to 

 being widened, the lines would be bodily shifted. It will be 

 seen that the detailed interpretation of such spectra is by no 

 means an easy matter. The measurement of the spectrograms 

 is in itself beset with difficulties on account of the diffused 

 edges of the bands merging into a background of more or less 

 continuous spectrum. Combined with this there is the 

 question of blends, differing intensities of the lines, and the 

 varying sensitivity of the photographic plate to light of 

 different wave-lengths. 



With reference to the displacement of lines due to the 

 motion of the source in the line of sight, it should be pointed 

 out that the magnitude of the effect varies with the wave- 

 length of the line in question, the displacement corresponding 

 to any given velocity being greater in the red than in the 

 violet. If, then, we are considering a practically uniform dis- 

 persion like that given by a grating, the linear shift will be 

 more apparent in the region of longer wave-lengths. The 

 general formula for calculating the velocity in the line of 

 sight from the observed displacement mav be given as : — 

 v = v'A^^' 



where V --= velocity of light 



\\ = normal wave-length of the line in question 

 Xo = the wave-length of the displaced line. 

 Since Xi — X2 represents the change of wave-length or dis- 

 placement, the formula becomes: — 



v=vd_ 

 X 

 The displacement which would be produced in any given line 

 bv a given velocity in the line of sight is thus : — 



The linear shift corresponding to a given velocity is thus 

 twice as great at X 6000A° as at X 3000A° when represented 

 on a wave-length scale of equ.al parts, which is approximately 

 the condition with grating spectrograms. In the case of 

 prismatic spectra, we have the two effects of shift and dis- 

 persion acting in opposition, but the increasing dispersion in 

 the region of shorter wave-lengths very umch more than 

 compensates for the decreasing line shift, hence the displace- 

 ments appear more obvious with decreasing wave-length. 



These are some of the circumstances which render the 

 study of the spectra of novae a matter of considerable 

 complexity, but of the general truth of the interpretation 

 supplied by the theory of the third body I personally feel no 

 doubt. It is probable that difficulties, apparent or real, may 

 be encountered in matters of detail, but the theory which 

 Professor Bickerton has formulated supplies a fertile, and I 

 believe, a true basis for the future investigation of these 

 remarkable celestial objects. Small discrepancies, or apparent 

 inconsistencies when carefully investigated may throw un- 

 expected light on the processes involved, for there nmst 

 follow many problems involving conceptions of great interest 

 to students of atomic physics, such, for example, as the 

 stabilit\' of the elements referred to above. 



Kast Crovpon. CHAKLI.S W. RAFFETY. 



