198 Dr. Forster on the Variation of the reflective, 



air is encumbered with the cirrostrativeness of the wanecloud, 

 then all the celestial bodies appear redder than at other times, 

 from the greater disposition of the air at those times to se- 

 parate and transmit the red colour, which must considerably 

 influence the foregoing observations. 



The last observation I have here to make relates to a phse- 

 nomenon hitherto entirely unexplained. I allude to rapid 

 alternations in the colour and brilliancy of the light of certain 

 stars viewed near the horizon. 



Some years ago, on looking towards the constellation of 

 the Scorpion, I observed a remarkable changing of colour in 

 the fluctuation of Airfares : for a second or less of time it ap- 

 peared of a deep crimson colour, then of a whitish colour; then 

 the crimson was resumed ; and so on at alternating periods. 

 Sometimes every other twinkle showed the red colour, while the 

 alternating twinkle appeared of the ordinary colour of starlight. 



What is commonly called the twinkling of a star seems to 

 be an apparent fit of dilatation and increased bi'illiancy rapidly 

 succeeded by the opposite state of apparent contraction of 

 surface and dulness. I have observed, also, that the twinkles 

 are of longer or shorter duration, at different times: now, in 

 general, the intense red light I allude to occurs in every alternate 

 dilatation ; but sometimes only in every third, and at other 

 times quite irregularly: moreover, it lasts longer sometimes 

 than at others, and scarcely ever exceeds two seconds of time 

 at once. This phaenomenon as viewed in Antares was par- 

 ticularly conspicuous at Clapton, in autumn 1809, at Tun- 

 bridge Wells on June 19, 1817, and during the whole summers 

 of 1822 and 1823 at Hartwell. I. saw this phaenomenon also 

 in other stars, in France, in September 1823; and I have seen 

 it in the summer of 1822, in the atmospheres of Switzerland, 

 Alsace, and all along the Rhine : it is evidently not confined 

 to any local cause. 



I have formerly published accounts of this phaenomenon in 

 the Journals, and have ascribed it to some sort of change in the 

 star itself, or to a revolution round an axis, whereby different 

 coloured portions of the luminous sphere might be presented 

 to us : but this explanation vanishes on a moment's reflection ; 

 and I am inclined to ascribe it to some atmospherical cause. 

 I have sometimes thought that the upper portions of the atmo- 

 sphere might have some undulatory motion, and that the al- 

 ternating colour might be produced by its refractive powers : 

 for the atmosphere, in this case, acting as an imperfect prism, 

 might present different colours, according to the varying in- 

 clinations of its wavy surface*. I have thought, too, that por- 



* See Phil. Mag. vol. xlix. p. 452 ; and the Perennial Calendar, Jan. 16, p. 17. 



tions 



