1904-5.] On Prof. Seeligers Theory of Temporary Stars. 527 
considerations. We obtain an at least approximate idea of the 
appearance of such a band, so far as it is due to the radiations of 
the photosphere and the expanding atmosphere, by combining the 
intensity-curves of (1) the continuous spectrum in the neighbour- 
hood of the line, (2) the absorption -band displaced towards the 
violet, and (3) the emission-band at normal wave-length. A com- 
bined band of this character is schematically represented in fig. 4, 
A A representing the normal position of the special line. 
A 
ofc<y- 4r , 5 OUYVti. 6 
On the other hand, the radiations contributed to the band by 
the luminous ring may be roughly represented by the intensity- 
curve in fig. 5. If, in combining the two curves 4 and 5, we apply 
the simple additive rule, we obtain the total intensity-curve of the 
band in fig. 6. Since our assumptions as to the relative shift and 
intensity of the various components must of necessity be vague, 
there are, of course, many ways of drawing these curves, and the 
diagrams therefore represent only one special case out of a great 
number of possible combinations. But in constructing the curves I 
have aimed at adapting their relative dimensions to the phenomena 
actually observed in one particular case of new stars, viz., in Nova 
