On the Chemistry of the Hottest Stars. 



183 



The probable Physical Causes of the Differences. 



T submit that the differences found are precisely those we should 

 expect on the meteoritic hypothesis. I deal with the differences 

 indicated seriatim. 



1. The Inequality of the Continuous Absorption. — I have already 

 pointed out that in the case of a swarm there must always be cooler 

 vapours mixed with the hotter ones in the most valid absorbing 

 regions. It is these cooler vapours which produce the absorption in 

 the violet and ultra-violet. 



In a condensed mass of vapour they can only exist at the limit of 

 the atmosphere where their absorption is reduced in consequence of 

 the low pressure. 



2. The smaller thickness of the Hydrogen Lines. — The difference in 

 thickness of the hydrogen lines is also sufficiently accounted for by 

 the difference of absorbing conditions. In the stars of increasing 

 temperature, consisting of uncondensed meteoritic swarms, the inter- 

 spaces will be largely occupied by hydrogen at a high temperature, 

 and the radiation of this gas will tend to mask the absorption pro- 

 duced by tbat in the immediate neighbourhood of the incandescent 

 meteorites which merely graze. In the condensed stars with photo- 

 spheres, any masking effect of this kind must be very much less pro- 

 nounced, so that the hydrogen lines will be broader than in the 

 spectra of uncondensed swarms. I have shown in a previous part of 

 this paper that the hydrogen lines are more constant, in regard to 

 their intensities, than the lines of other substances in the stars of 

 increasing temperature, and this greater life of the hydrogen mole- 

 cule seems to explain the fact that the hydrogen lines are specialty 

 picked out after passing to the downward side of the temperature 

 curve when a state of quiescence is reached. 



It should be remarked that the great distention of the hydrogen 

 lines in the hotter stars which have begun to cool does not necessarily 

 indicate a great thickness of absorbing hydrogen, since, in the case 

 of the sun, the very broad H and K lines are produced by an absorb- 

 ing region of small thickness in comparison with the sun's diameter. 



3. The greater thickness of the Metallic Lines. — In the case of a 

 swarm, the thickness of the effective absorbing gases and vapours in 

 the line of sight will be very much greater than the effective thick- 

 ness in the case of an atmosphere surrounding a photosphere, even if 

 the masses and average temperatures be the same in both cases. The 

 light proceeding from the central parts of the swarm must pass 

 through the whole depth of vapours filling the interspaces, and on 

 this account the absorption lines would be more intense in the case of 

 a swarm. The greater intensities, and to some extent the breadth of 

 the metallic lines, are thus explained, for since the metallic vapours 



