516 Proceedings of Royal Society of Edinburgh. [sess. 
feature of all these spectra is the appearance of broad bright lines, 
bordered on their more refrangible edges by diffuse absorption- 
bands. Now, assuming that these composite bands were caused 
by two bodies moving in different directions, why should the bright 
bands invariably appear on the less refrangible and the dark 
bands on the more refrangible side 1 ? In other words, why 
should the body or bodies giving the bright line spectrum 
always move from us, while those showing the absorption- 
lines should be invariably directed towards us? It was the 
seeming inexplicability, on the ground of Doppler’s principle, of 
this universal phenomenon which led astrophysicists to search 
for other explanations, such as high pressure and abnormal 
dispersion. 
But if we look more closely into the question, the reasons for 
abandoning Doppler’s principle seem by no means so convincing 
as they were thought to be. The position was abandoned before 
the field was thoroughly reconnoitred. Indeed, I shall endeavour 
to show in the subsequent remarks that an explanation of the 
extremely complicated spectrum of new stars based on Doppler’s 
principle is not only possible, but also sufficiently probable, on 
account of the simplicity of the underlying hypotheses on the one 
hand, and the satisfactory agreement between theory and observa- 
tion on the other. 
The leading idea upon which these considerations are based is 
well known to astronomers through Professor Seeliger’s ingenious 
investigations : my present contribution is indeed merely an 
extension of the celebrated theory which we owe to this dis- 
tinguished astronomer. Seeliger’s hypothesis, which ascribes the 
outburst of a new star to the collision between a dark solid body 
and matter of a nebular constituency, has so far not been worked 
out in detail, so that no definite conclusions have been formed as 
to the motions of the matter involved in the catastrophe. In a 
general way, however, Professor Seeliger draws attention to the 
important r6le performed by the star’s gravitational attraction on 
the approaching nebulous matter, a consideration we often find 
seriously neglected in subsequent investigations. He remarks 
that, as the star approaches the nebulous cloud, the latter, through 
the action of gravitation, will extend out to meet it. The attracted 
