MARY W. WHITNEY. 251 
differs from IIJ,, not in the order of stellar existence, 
Vogel thinks, but in the constituency of atmosphere. 
What substance or substances produce the absorption 
bands of III, neither Vogel nor other investigators are 
ready to conjecture, though they incline to believe that 
they are attributable to one substance rather than to 
several. The three bands of the II], type, suggest 
strongly carbon compounds. Similar bands have been 
seen in the spectra of comets, notibly the large one of 
1882. Vogel has published (1883), a carefully prepared 
catalogue of star spectra. His catalogue includes four 
thousand fifty stars extending to the seventh magni- 
tude. Duner, of Lund, Sweden, has published (1884) 
still another catalogue of spectra of the class III, and 
II], containing about three hundred fifty stars, forty- 
five coming under IIJ,. It is a significant fact that no 
one of the III, type yet discovered is larger than the 
fifth magnitude. It may be that they are in themselves 
neither smaller nor less brilliant than other stars, but 
that their heavy atmosphere cut off the larger part of 
the ight which would otherwise radiate to us. 
Nearly all the variable stars (excepting those of the 
Algol order), fall into the third class. 
Now, what does the spectroscope tell us in regard to 
the Nova Orionis? It confirms our conclusion that itis 
an ordinary variable. Duner says, ‘‘Its spectrum isa 
fine III, with broad and dark absorption bands” ; Vogel, 
of Potsdam, ‘‘ It is a‘beautiful, clearly-defined spectrum 
of class III, with many dark bands”’; and Wolf, of 
Heidelberg, ‘‘Itis a pure III, spectrum and the star is 
without question a variable of the Mira Ceti class.” . 
We must now ask, what are the spectra of temporary 
stars, to which division the Vovua Andromede must be- 
long if we attempt to classify it at all? The new star 
of 1866 was the first one examined by the spectroscope. 
Huggins obtained its spectrum about four days after its 
185 
