the exposures required in photographing the spec- 

 trum ranged from 2 to 16 hours. 



The most plausible hypothesis of temporary 

 stars accounts for their rapid brightening on the 

 supposition that a faint star suddenly plunges into 

 a gaseous nebula. After passing through a remark- 

 able series of changes, the spectra of Novae have 

 usually been supposed to correspond closely, in 

 the last visible stage of their existence, with the 

 spectra of nebulae. The Mount Wilson results, in 

 harmony with an observation by Hartmann, show 

 that after the lapse of years the characteristic lines 

 of the nebular spectrum disappear completely, 

 at least in some cases, as though the star had 

 finally passed out of the nebula which caused its 

 outburst of light. If the above hypothesis is cor- 

 rect, the temporary brightness of Novae resembles 

 that of meteorites, which are kindled into brilliancy 

 when passing through the earth's atmosphere. 



It also appears probable that at least a portion 

 of the remarkable Wolf-Rayet stars, most of which 

 lie in the Milky Way (where practically all Novae 

 occur), are temporary stars in the later stages of 

 their existence. 



VARIABLE STARS. 



An interesting result is derived from a study of 

 the color of a certain variable star, RR Draconis, 

 one of the revolving components of which com- 

 pletely eclipses the other at the time of minimum 

 brightness. It is found that the faint companion, 



