J ULV II, I 90 I ] 



NA TURE 



figure, we have a motion of the incandescent nebulous matter 

 towards us. We must therefore expect a bright line displaced 

 towards the violet. The maximum intensity of this line will be 

 near the edge farthest from the normal position, and it will 

 gradually dim oft' in brightness as the normal wave-length is 

 approached. This dimming off, which is chiefly the consequence 

 of decreasing incandescence with decreasing rotatory motion, is 

 enhanced by the absorbing action of the cool substance in the 

 outer rings of the vortex. For this matter, having no motion 

 wiih a direction towards us, can absorb only those wave-lengths 

 which coincide with its own, viz. those emanating from the 

 incandescent matter at and near the arc Bb"b', while it leaves 

 the light of higher wave-lengths emanating from A unaffected. 



If we now consider the conditions prevailing in the annular 

 section bcc"c'b', we see at once that the motion of the particles 

 is here in the opposite direction, viz. away from the earth. 

 The light emanating from these particles will, therefore, be 

 displaced towards the red, and consequently a bright line 

 must appear on the less refrangible side. But the velocity of 

 motion in the line of sight in this annular section is smaller than 

 at \, and the incandescence of the particles much inferior ; 

 hence the maximum intensity of this new line will not be so far 

 from the normal wave-length, and the line will also be fainter 

 than that on the violet side described before. Taking into con- 

 sideration these circumstances, we may then assume that the 

 intensity curve resulting from the whole radiation in the segment 

 c'c'cA on the left hand side of Fig. 2 is approximately repre- 

 sented by the curve B in Fig. 4 The radiation of the corre- 

 sponding segment on the right hand side of Fig. 2 must obviously 

 be the image of B in the normal 00', and is thus represented by 

 c of Fig. 4. 



As all the light emanating from the star must be supposed to 

 pass through the slit of the spectroscope, the line seen in the 

 spectrum will be the resultant of all the component curves. 

 Clearly the character of this compound line remains unaltered, 

 whatever position the line of sight may have with reference to 

 the motion of the star or to the axis of gyration, except the one 

 case when the line of sight is parallel to this axis. Obviously 

 the band would then be reduced to a single line at normal 

 wave-length. The probability of such an occurrence is, how- 

 ever, excessively small. Hence the theory here advanced would 

 lead us to accept the peculiar character of the bright lines 

 exhibited in the curves B and c of Fig. 4 as a feature character- 

 istic of the whole class of temporary stars. 



So far we have traced the structure of the line emitted by a 

 substance of the nebulous matter on the assumption of circular 

 rotatory movements. We have still, however, to take into 

 account the influence of the flow of this matter to and from the 

 centre of the vorlex, as indicated in Fig. 3. 



The cool matter flowing in at the poles of the vortex must be 

 supposed to be in a non-luminous condition. It can neither 

 radiate nor absorb selectively in the way as Kirchhoff's law 

 would require, and hence it has no elTect on the structure 

 and position of the bright and dark bands. The hot 

 and incandescent matter flowing out at the equator, however, 

 has an important influence in this respect. Reverting to Fig. 2, 

 it may be readily seen that the radial component of the motion 

 of the gaseous particles within the space .^ A D D included be- 

 tween the two tangents to the surface of the body is directed 

 towards the sun. Consequently the absorption bands caused 

 by these particles must all appear displaced towards the violet. 

 Thus, instead of curve A in Fig. 4, which represents the intensity 

 of these bands when there is no radial motion, we obtain curve 

 ij of the same figure as the actual representation of the intensity 

 of these bands in the spectruin. The striking difference between 

 .\ and D is therefore the considerable displacement towards the 

 more refrangible side of the absorption band in D. No matter 

 what direction the line of sight has with regard to the motion of 

 body or nebula, or w^hether we consider a section through the 

 vortex at right or oblique angle to the axis of gyration, in all 

 cases, except the one mentioned above, the displacement of the 

 absorption bands must be towards the more refrangible side. 

 The effect of the radial components of the gyratory motion on 

 the position of the bright bands is easily seen from a considera- 

 tion of the conditions prevailing in the segments cc"c'Ain 

 Fig. 2. Apparently there are as many motions towards as there 

 are away from the sun. Hence the effect will consist in a 

 general broadening of the four maxima represented in B and c of 

 Fig. 4, without, however, affecting their position relatively to the 

 normal wave-length. 



NO. 1654, VOL. 64] 



The combination of B, c and D in Fig. 4 vi'iW therefore ap- 

 proximately represent the complete intensity-curve of a line 

 emitted by a substance in the gaseous envelope of the Nova. 

 The combination of B and c alone represents the structure of the 

 bright bands ; it agrees perfectly with the drawing given by Sir 

 Norman Lockyer for the case of H3. Often enough the two inner 

 maxima may overlap each other and thus produce the impression 

 of a single strong maximum at normal wave-length. This would 

 explain the curves with only three maxima exhibited in the 

 diagrams of Sir Norman's paper. 



There can be no doubt that the absorption band in i> will 

 partly interfere with the maximum of the emission band on 

 the violet side. As a rule the former may be assumed to be the 

 more refrangible, since its displacement must be enhanced by 

 the expansion of body and vortex as a consequence of increased 

 production of heat. In the case of Nova Persei, the difference 

 in the displacement owing to this latter effect must have been 

 very considerable ; the two bands were here placed beside each 

 other with comparatively little encroachment of the one upon 

 the other. The conditions in Nova Aurigdj appear to have 

 been somewhat different. Here the displacements of the emis- 

 sion and absorption bands seem to have been fairly eqiial, the 

 latter obliterating the former almost completely. I consider the 

 bright lines noticed in almost all the absorption bands of this 

 Nova to be the remnants of the more refrangible maxima of the 

 bright bands. 



In any case the effect of a partial encroachment of the absorp- 

 tion band upon the emission band must be a displacement of the 

 centre of the bright band towards the red. We therefore derive 

 two most important results from the theoretical considerations 

 here given : — 



(1) In all the temporary stars the absorption bands must be 

 displaced towards the more refrangible side. 



(2) It! all the temporary stars the centres of the emission bands 

 must show displacements towards the less refrar.gible side. 



These conclusions are in entire accordance with the facts. As 

 already mentioned, an exception may happen when the line of 

 sight is approximately parallel to the axis of gyration. In this 

 case both the emission and absorption lines would appear in 

 their normal positions, since all the vortex-motions are then 

 more or less perpendicular to the line of vision. Hence the 

 two lines would overlap each other almost completely, and the 

 result would be a purely continuous spectrum with little or no 

 traces of selective absorption or emission. Such an exceptional 

 case many perhaps have presented itself to our eyes in Nova 

 Andromedre. 



The new star in Perseus, thanks to its discovery by Dr. 

 Anderson almost immediately after its appearance in the 

 heavens, offered to astronomical science the unique opportunity 

 of recording the initial stages of its development. None of the 

 theories hitherto propounded have so far succeeded in accounting 

 for the spectral changes so markedly exhibited in the star's 

 light during the first days of its existence as a j^adiating 

 celestial body. But just these quite unexpected and at 

 first sight perplexing changes find a marvellously sin- pie 

 explanation by the modification of Prof. Seeliger's nebular 

 theory here offered. The first effect of the collision 

 between the dark body and a cosmical cloud must be an 

 enormous heating of the body's surface and the generation of 

 an incandescent atmosphere around it. The depth of this 

 "cloak" of incandescent matter will at first be small, so that 

 the star at that time presents the aspect of a luminous nucleus 

 surrounded by a comparatively shallow atmosphere of incan- 

 descent gases. The spectrum yielded by such a star must be 

 continuous, showing dark lines generated by the absorbing 

 faculty of the glowing gases between the nucleus and space. 

 At the moment of the outburst these dark lines will be exceed- 

 ingly faint, and they will show only such a displacement as is 

 necessary from the amount and direction of the translatory motion 

 of the body. As time passes, however, and the gyration of the 

 atmosphere becomes stronger, the outward flow of the hot par- 

 ticles must rapidly increase, and thus, in accordance with the 

 developments given above, the dark lines, while becoming 

 broader and more distinct, must gradually shift towards the 

 more refrangible side. This increase in the displacement of 

 the dark bands during the first days has been actually ob- 

 served by Prof. Vogel and many other spectroscopists. To 

 yield a bright line spectrum the incandescent atmosphere must 

 have attained a considerable depth, otherwise the bright lines 

 emanating from particles in the space cc"c'.\ would make no 



