POPULAR SCIENCE 619 



metallic vapour is to broaden the radiations in the spectrum, 

 as shown by the laboratory researches of Dr. Duffield, and of 

 other investigators. The pressure required for the broadening 

 of the bands, such as are observed in new stars, would be very 

 great. Yet the possibility of pressure, as an agent in the 

 widening of the bands, when we consider the scale of the 

 catastrophic outburst which results in a new star, is not to be 

 neglected. Enormous pressure is competent to furnish the 

 continuous spectrum, recorded by several observers as the first 

 stage in the spectral phases of a new star. It may therefore 

 be a contributory cause for the widening of the bright bands. 

 But the chief cause would seem to be the motion of shells 

 or clouds of gas, moving towards and away from the observer. 

 For the breadth of the bands is a function of the wave-length 

 which, according to the well-known principle of Doppler, means 

 motion in the line of sight. These broad bright bands show 

 a considerable amount of structure, especially so in those due 

 to hydrogen. For the most part they are double bands, and 

 in the case of Nova Aquilse (191 8) the hydrogen bands were 

 triples. The central bright bands of the triples were practically 

 stationary, the other two members indicated line of sight 

 velocities of the order of 1,300 kilometres per second approach 

 and 1,200 recession. In all new stars these bands of hydrogen 

 remain permanently widened without change for a considerable 

 period, in some cases for several months. This would mean 

 a persistence of these enormous velocities of approach and 

 recession for such periods, which presents a very great difficulty 

 for any adequate explanation of the origin of new stars. Con- 

 sidering the scale of the initial catastrophe, it may not be found 

 to be altogether unimaginable or insuperable. 



On the analogy of a solar outburst we should expect that 

 there would be absorptions by the relatively cooler clouds of 

 hydrogen and other gases advancing towards us, and that these 

 absorptions would show indications of velocity of approach. 

 This expectation is realised, and in all new stars the absorption 

 spectrum which accompanies and reproduces the radiation 

 spectrum is, as we have noted, always on the violet side, and 

 greatly displaced, indicating motion of approach. In Nova 

 Aquilae (191 8) the hydrogen absorptions were doubled, and 

 even tripled, indicative of successive outbursts of masses of 

 this gas. The velocities of approach, corresponding to the 

 wave-length displacements, were of the order of 1,700 kilo- 

 metres per second for the first set of absorptions, and 2,400 

 kilometres per second for the second set, about twice and 

 thrice respectively the velocity of the quickest moving solar 

 prominences. The displacements are directly proportional 

 to wave-length, and therefore indicate line of sight velocities. 



