526 
Proceedings of Royal Society of Edinburgh . [sess. 
were not conspicuous. A few days afterwards these bright bands 
appeared strongly developed, and were then the most promi- 
nent feature of the spectrum. 
I have already mentioned that the rotating ring of luminous 
nebular matter modifies to a certain extent the appearance of the 
spectral lines. Its effect will be to produce two additional maxima 
of brightness, the one displaced towards the red, the other 
towards the violet. For in whatever direction we may view the 
star — unless the line of sight be at right angles to the plane of 
rotation — we will always have some substance of the ring moving 
towards us on the one side of the star, and matter moving from us 
on the other. Between these maxima a more or less hazy absorp- 
tion-line appears at approximately normal wave-length (leaving, of 
course, out of consideration the relative motion of the whole 
system : star + ring). This absorption-line is due to the gaseous 
particles of the ring travelling in front of the star, as seen from the 
standpoint of the observer. (Compare the sketch given in fig. 3.) 
The complete structure of the bands is determined by these 
