PROFESSOR FORBES ON THE COLOURS OF THE ATMOSPHERE. 289 



These coronse, notwithstanding their apparent analogy to the colours of thin plates, 

 seem rather to be due to the effect of diffraction.* 



The non-appearance of the lines of the spectrum in my experiment, may be 

 plausibly explained in the following manner, which, however, I offer merely as a 

 conjecture. When steam of high pressure issues from an orifice, a horizontal 

 section of the expelled column will include vapour in every stage of condensation. 

 Its centre, up to a certain height, will be pure invisible steam ; at the exterior of 

 all, in contact with the cold air, there will manifestly be vesicular steam, and a 

 cylindrical space between the two will contain red steam. Now it is extremely 

 probable, that when the experiment is performed on the small scale, as I have 

 described it, by suffering light to pass through such a compound column, and 

 then analyzing it by the prism, enough of unabsorbed rays are reflected from the 

 highly luminous surface of the vesicular steam to prevent the fine lines from being 

 seen if they exist. And I am strongly confirmed in this conjectm-e by the fact, 

 that when the rush of steam is very violent, and always when much vesicular 

 vapour is present, the unabsorbed part of the spectrmn presents a washy and 

 impure tint (pai-ticularly mentioned in my former paper), which probably arises 

 from a lilending of the colours, produced by this cause. 



In conclusion, I have onl}!- a word or two to say respecting the application 

 of these facts to atmospheric appearances regarded as prognostics of weather. 

 The modified hues of the sky, and of the sun and moon near the horizon, have, 

 for so many ages, and in so many countries, been regarded as the surest indica- 

 tions of atmospheric changes, that we cannot doubt that it is to the variety of 

 conditions in which vapom- exists in the an-, more or less nearly condensed, that 

 these phenomena are due. Humboldt describes the colour and form of the sun's 

 disc at setting in tropical regions, as the most infallible prognostic,! and else- 

 where ascribes these variations " to a particular state of the vesicular vapour." f 

 Since the red steam occurs only dm-ing the critical stage of its pai-tial condensation 

 (and perhaps conversely during evaporation), it is evident that it must corre- 

 spond to a critical state of diffused vapour of the atmosphere. The applications 

 might be very extended ; I wiU only advert to one, the surest, most consistent, 

 and probably the most ancient of such prognostics. The red evening and grey 

 morning as the signs of fine weather, ai;e recorded in the verses of Aratus, || in 

 the New Testament,^ and in one of our most familiar proverbs. It is wholly in- 

 explicable on the theory of Bbandes, which considers the redness as due solely to 

 the purity of the atmosphere, since that is usually greater in the morning than 

 the evening. According to my view it occurs thus : Soon after the maximum 



■• See Young's article Chromatics, in Encyc. Brit., and Fraunhofee in Schumacher's Astrono- 

 mische Abhandlungen. Drittes Heft. 1825. 



t Relation Historique, 8vo, ii. 128. J New Spain (translation), ii. 326. 



11 Diosemeia, 93. Quoted by Kamtz. - § Matt. xvi. 2, 3. 



