494 Scientific Proceedings^ Royal Dublin Society. 



magnitude do not owe their position in the stellar scale to the in- 

 tense illumination of the unit of surface. When a cold body is 

 gradually heated it gives forth at first dark heat, and then a red 

 light, after which the other hues in the solar spectrum gradually 

 spring up, and finally the ultra-violet chemical rays make their 

 appearance. At every stage in this heating process the illumina- 

 ting power of the unit of surface increases ; and the illuminating 

 power of the unit of surface can thus be determined within certain 

 limits by the nature of the stellar spectrum, being lowest in the 

 case of the red stars, and highest in that of the brilliant white or 

 blueish stars — at least, unless the latter owe their peculiar hue to 

 the loss of the red rays, which fact would always be revealed by 

 the spectrum. We are in this manner able to conclude that the 

 degree of illumination of several stars of the first magnitude is not 

 above, but below, the average. Such stars, for instance, are 

 Aldebaran, Antares and Arcturus. If these are really distant 

 stars, the extent of illuminated surface must be enormous. From 

 their lower temperature these stars are probably in a more con- 

 densed condition than the brilliant white or blueish stars already 

 alluded to, and the masses which we would have to assign to them, 

 if they were very remote stars, would be almost incredible. Before 

 leaving this subject, I may remark that our early astronomers, at 

 least in the northern hemisphere, appear to me to have framed 

 their nomenclature with a view to the colour of the stars. If we 

 examine the stars of the first magnitude visible in this country in 

 alphabetical order, we shall find ourselves passing at almost every 

 stage from a redder to a whiter or bluer star — the few departures 

 from which rule may perhaps be explained by a change of colour 

 since the names were imposed. Thus, Capella, which stands pretty 

 high in the alphabetical list, is described by many of the earlier 

 writers as a red or reddish star. 



However, it is said that, though much labour has been expended 

 in trying to ascertain the parallaxes of stars of the first magnitude, 

 very few of them present a sensible parallax, while several of the 

 fainter stars do so, and also that many faint stars exhibit a larger 

 proper motion than the stars of the first magnitude. This reason- 

 ing would be conclusive against a theory which supposed that the 

 distance of a star was invariably indicated by its magnitude ; but 

 I think it has little force against one which only supposes that the 



