POPULAR SCIENCE 623 



The velocities in the Hne of sight indicated by the displacements 

 of the absorptions would become gradually less, when a new 

 pulse would seemingly eject a fresh jet of hydrogen, accom- 

 panied by other gases and vapours of metals. Certainly in 

 these valuable observations there is a suggestion of an explosive 

 outburst of hydrogen from moving bodies, or a moving body, 

 which carried the other elements. -"^ 



We have already adverted to the fact that the bright band 

 at wave-length 4640, which is partly due to helium, partly to 

 nitrogen, and partly to carbon, excited by an extra strong 

 electric field, becomes more pronounced as the star wanes in 

 brilliancy. Again the order of the various phases of the 

 spectrum, through Classes F, A, B, P, O, occur as the star loses 

 lustre, and presumably becomes cooler. But this order is that 

 of giant diffuse stars ascending in temperature. This anomaly 

 may be explained, and the behaviour of the bright band 4640 

 furnishes the clue, if we attribute the later spectra not to 

 temperature changes but to a successive strengthening of the 

 electric field. According to the experiments of Dr. G. A. 

 Hemsalech,' the principal conditions for the appearance of the 

 spark-produced radiations in metallic vapours are, first a low 

 degree of ionisation of the medium, and secondly the existence 

 of a strong electric field giving rise to an electric current 

 through the medium at a high potential gradient. To make 

 an application of these valuable experiments to the varying 

 spectra of nova : it would appear that in the early stages of 

 high temperature in the constituent gases, indicated by the 

 great brilliancy of the star and its extended spectrum far into 

 the ultra-violet, the ionisation would be very active. Hence 

 a strong electrical field could not be established. But with a 

 falling temperature, indicated by the star's waning in brilliancy, 

 the electric field would become correspondingly greater. Hence 

 the increased lustre of the band 4640, produced in the laboratory 

 by heavy electrical currents, and the successive changes of 

 the spectrum through the above-named classes. 



Several speculations have been broached as to the origin 

 of new stars. Prof. Bickerton has strenuously upheld a grazing 

 impact between two suns, resulting in the formation of a third 

 body. There is much to be said for this hypothesis. Or it 

 may be that the approach of a luminous to a non-luminous 

 massive body of low density would have caused the ejection 

 of a stream of gas towards the passing star. Others, again, 

 incline to the view, first proposed by Seeliger, that the pheno- 

 mena of new stars are due to the entrance of a star into a nebula, 

 on the analogy of the luminous appearances that accompany 



1 Journal British Astronomical Association, xxx. i8i. 



2 Report to the Department of Scientific and Industrial Research^ Oct. 8, 1920, 



