1904 - 5 .] On Prof. Seeligers Theory of Temporary Stars. 515 
If the results of the Glasgow and Mount Hamilton observations 
thus limit our field of search by excluding those explanations in 
which the internal properties of the vibrating atoms must be 
considered to play an important part, there remains, in my opinion, 
only one explanation which is in a priori agreement with the 
observed facts, viz., that based on Doppler’s principle. Indeed, if 
motions in the line of sight are the cause of the peculiar emission 
and absorption bands in the spectra of the Novae, the similarity 
of their structure, independent of the chemical nature of the 
elements, and the strict proportionality of all the displacements 
to the wave-length are necessary desiderata. The crucial point 
seems to be, therefore, this : On the one hand, observation has 
demonstrated that the structure of the bands is governed by no 
other conditions than that of wave-length ; on the other hand, of 
all explanations, only that based on Doppler’s principle accounts 
for this fact : hence motions of matter in the line of sight must be 
considered as the probable cause of the remarkable spectrum of 
temporary stars. 
This conclusion has increased my confidence in some theoretical 
views published two years ago, by which I attempted to explain 
the Nova spectrum. The new facts brought to light by subse- 
quent observations, especially by those referred to, make it now 
desirable to again publish these tentative speculations in a some- 
what modified form, and at the same time to compare the results 
of theory with our present empirical knowledge. 
Before entering upon the subject, I beg to pass a few general 
remarks of an historical character. In the earlier days of star 
spectroscopy explanations of the Nova spectrum were pre-eminently 
based on Doppler’s principle. Theoretical views focussed more or 
less round the one conception that motions of radiating and 
absorbing matter must be responsible for the observed displace- 
ments of the spectral lines. Later, however, doubts began to be 
felt as to the correctness of this view. Apart from the fact that the 
velocities of matter in the Novae exceeded by far the average 
motions in the line of sight commonly dealt with in stellar spectro- 
scopy, a serious objection to this view was thought to be found in 
the striking similarities between the spectra of all the Novae 
hitherto accessible to spectroscopic investigation. A universal 
