Tlce Theory of Temporary Stars. 11 



— and quite unexpected — phenomenon, its interest pales in com- 

 parison with that of the changes subsequently found to be 

 apparently going on in it ; for a photograph taken seven weeks 

 later^ shows that the nebula was, to all appearance, rapidly 

 expanding in all directions, and that one portion of it, apparently 

 in contact with the star, had greatly increased in luminosity 

 during the the interval, while the rest of the nebula had lost 

 much of its light.'- 



7. — The Theopy in the light of the later observations. 



In this section there are many points for discussion. I have, 

 as above, distinguished them by numbers. 



1. Variations in brightness per se are readily explicable on the 

 collision theory by attributing them to successive encounters, as 

 Lockyer and others have pointed out ; they are equally easy to 

 explain on the eruption theory, by ascribing them to alternate 

 expansions and contractions of the gaseous matter, consequent 

 upon the checking of the outru.sh. Of the two hypotheses, the 

 second accounts satisfactorily for the quasi-periodicity of the 

 light-variations, which are pretty much what we should expect if 

 it were true^ ; while unless we are to assume a quasi-periodicity 

 in space of the structure of the invaded system, similar to the 

 quasi-periodicity of the light-curve, it is difficult to see how the 

 theory of successive collisions can be made to fit the facts ; a 

 theory which necessitates any such subsidiary hypothesis is, to 

 my mind, self-condemned. 



2. Neither theory can be regarded as satisfactorily accounting 

 for the change of spectrum from a stellar to a nebular type.* 



1 Ap. J., xiv., p. 293. 



2 The question obviously suggests itself — Is the nebular spectrum due to the star or to 

 Ritchey's nebula ? The answer, I take it, must be that it is due to the star, otherwise we 

 should find on the plates a narrow stellar spectrum crossed by long bright lines, just as in 

 Huggin's photograph of the combined spectrum of the Orioti nebula and the trapezium 

 stars ; but nothing of the kind appears to have been observed. Besides, the nebula is too 

 faint an object to give any spectrum in the time allotted to the exposures by the 

 spectrographers. 



3 The light curve closely resembles the displacement curve of a damped vibrating 

 system. 



< Lockyer's meteoritic h3pothesis might have done so, had the spectrum passed throug-h 

 some or all of the stages characterised by him as Polarian, Aldebarian, and Antarian 

 (M.N., Ixi., App., p. 19) before assuming the nebular type ; but this, apparently, it failed 

 to do. 



