refractive and dispersive Powers of the Atmosphere, $c. 195 



And the most common observation on the rising and setting 

 moon will always confirm it on favourable nights. 



On other occasions, at the same time of night, I found an 

 unexpected difference in the appearance of the spectrum of 

 the very same planets ; the violet and in general the most re- 

 frangible colours being most conspicuous, and the spectrum 

 being more oblongated than ordinary. 



To reconcile this difference of effects with existing causes 

 apparently so similar, was now an interesting object; and at last 

 I found out that the difference was referable to a difference in 

 the quality of the falling dew, or diffused cloudiness. 



The greater prevalence of the red in the spectrum uniformly 

 accompanied that state of the atmosphere when the cirro- 

 stratus or wanecloud diffused itself after sunset ; while the more 

 oblongated spectrum with the violet and most refrangible co- 

 lours attended an atmosphere in which the descending dew was 

 stratus or fallcloud. I will not say positively that in the first 

 case the falling mist was itself cirrostratus, but it was a different 

 kind of mistiness and less purely white, when seen in the valleys, 

 than the common mist, and it happened when there were cir- 

 rostrati in the higher air. Another circumstance which con- 

 firms my explanation of the phenomenon is, that during the 

 falling of ordinary dew in clear and settled weather, particu- 

 larly with east wind, the horizontal haze shows beautiful tints 

 of the more refrangible colours, varying upwards, so that the 

 atmosphere becomes a sort of natural prism, of which I have 

 cited many examples in my "Researches about Atmospheric 

 Phenomena." Whereas when cirrostratus prevails, particu- 

 larly before wind, the red is the predominant colour of the haze, 

 and also of the clouds above it, which appear of the finest 

 crimson and vermilion. In some cases I believe the clouds 

 may themselves become prismatic, and the colour of which 

 they appear may be produced by their own structure ; but it 

 is more generally the effect of the haze through which reflected 

 rays pass in coming down. Clouds in this case may be com- 

 pared to planets viewed in a prismatic glass, their surfaces re- 

 flecting the borrowed light of the sun, which becomes dispersed 

 by the chromatopoietic properties of the atmosphere through 

 which it passes. 



Another circumstance which I witnessed several times in 

 November 1822, and which is of common occurrence, strik- 

 ingly corresponds with the above observations ; namely, that 

 clouds whose irregular surfaces and depending fringes were of a 

 golden or silvery colour, that is, reflected all the sun's rays 

 just as they received them, during a whole afternoon, were, on 

 the occurrence of the vapour point, suddenly turned red as if 



Bb 2 by 



