19 1 Dr. Forster on the Variations of the reflective, 



ticular cloud, that universal tables of mean refraction have 

 been made out, as if they would apply to all places. This is, 

 however, an essential error: and in developing it I find the 

 subject naturally divides itself into three distinct considera- 

 tions, as follows : 



I. 

 On the Variation in the refractive Power of the Atmosphere at 

 different Times of the Night and Day and on different Oc- 

 casions and Seasons. 



There is one question in the history of refraction which for 

 obvious reasons ought, if possible, to be cleared up ; namely, 

 Wherein consists the dispersive power of the atmosphere, which 

 is found to vary at different times and in different places? 

 To me it appears that this variation is owing principally to 

 the quantity and nature of the aqueous particles suspended in 

 the air. Astronomers have hitherto confined their considera- 

 tions too exclusively to the refractive property of pure air, 

 and have overlooked the circumstance that the atmosphere is 

 seldom pure for any length of time; they have consequently not 

 taken into account the varied effects of different sorts of dif- 

 fused vapours, which, though unseen by the common observer, 

 exist in the air in different proportions at sundry times and 

 places, and which prevail much more at some places than at 

 others. I was led to discover this by observing the vast dis- 

 proportion between the result of my observations on the stars 

 made at different times and seasons. In observing the planets 

 and brightest of the stars through prismatic glasses, I found 

 that the spectrum was less oblongated while the red colour 

 was more distinctly apparent at the time of the first falling of 

 the dew than at almost any other time of the same nights. 

 Venus and Jupiter afforded a fine opportunity of ascertaining 

 this fact last spring, as these planets could be both seen 

 of an evening as early as the period of the vapour point*. 



* The vapour point is that precise period which occurs in the progress 

 of evening, when, from the declining temperature, the air can no longer 

 hold in solution the same quantity of aqueous gas that it maintained du- 

 ring day ; and when consequently the aquosity is first deposited in the form 

 of visible vapour or cloud, and gradually descends to the earth ; on arriving 

 at the surface of which it eventually forms a stratus or fog, called for this 

 reason the fallcloud. In the morning the counterpart of this process 

 takes place, and the gasified water being taken into the composition of the 

 air again, ascends, till rising into a colder region it is again condensed into 

 visible clouds which float on the vapour plane, or that precise elevation 

 whereat the air begins to be too cold to hold the vapour in solution. At 

 this elevation it forms clouds dissimilar in their modifications, according to 

 electrical causes which exist in the air, but generally speaking cumuli, the 

 lighter modifications of cloud occurring higher up from local electrical and 

 other causes which disturb the equilibrium of the temperature of the at- 

 mosphere. 



And 



