refractive and dispersive Powers of the Atmosphere, $c. 199 



tions of the aqueous atmosphere, possessing different prismatic 

 powers, might be transmitted downwards in dew, or that there 

 might be some other unknown motion in the real air, which 

 might cause the appearance. Antares, Betalgeus, and some 

 other red stars, show this change of colours very strongly, par- 

 ticularly the former : while Sirius and the brilliant white stars 

 show this alternation of colour in a less degree. It is not ob- 

 served in Procyon, is weak in Capella, but is considerable in 

 Vega and in Arcturus, the latter being a red star. Antares 

 shows this phenomenon more brilliantly than any of the others. 

 Differences in the composition of the light of stars may explain 

 these varieties of effect. 



Nobody can, I think, reflect on all the above details, hasty 

 and imperfect as they may be, without at once seeing the 

 importance of extending our observations on the phenomena 

 to which they relate, in constructing tables of refraction ; and 

 I shall still be excused, I trust, in the absence of more ma- 

 tured and extended details, for this imperfect attempt to ex- 

 cite the attention of philosophers to facts which seem calcu- 

 lated to produce an important influence on many of our most 

 useful astronomical calculations. 



II. 



Of the Variation of the mean refractive Power of the Atmosphere 

 in different Places. 



Although the refractive and dispersing power of the atmo- 

 sphere varies at different times in the same place of observation, 

 yet by dividing the sum of a vast number of observations by 

 their number, we shall get at a mean or average of the refractive 

 power of the atmosphere of any particular place. This I call 

 the mean refractive power of the atmosphere, and this mean 

 differs very much in different situations, so that tables of 

 mean refraction will not apply universally, but must be made 

 out separately for each place of observation. Thus a cor- 

 rection which would be applicable to observations made at 

 Greenwich, would not equally apply to Dublin or to Paris, 

 much less to Palermo or the Cape of Good Hope: and they 

 would be still less applicable in Peru, where the dispersive 

 power of the atmosphere is prodigious. I was led to the adop- 

 tion of this opinion, in the first instance, by analogy, it being 

 strictly conformable to a thousand other corresponding in- 

 stances of local variation in atmospheric phenomena. Ex- 

 periment seemed to confirm it. And I have lately heard of a 

 further corroboration which this opinion has received from 

 the able observations of Dr. Brinkley, communicated to the 

 Royal Society. But as I have only heard casually of that 



paper, 



