Chap. 10.] Blue Coloilr of tie Sky. 289 



whole day, and fet with him in the fame manner as 

 they rife. Once or twice he faw a fourth parhelion 

 under the true fun, but this, he acids, is not com- 

 mon *. 



The caufe of thefe is apparently the reflection of the 

 fun's light and image from the thick and frozen clouds 

 in the northern atmofphere, accompanied alfo with 

 fome degree of refraction. To enter upon a mathe- 

 matical analyfis of thefe phenomena would be only te- 

 dious, and very foreign to our purpofe. From what 

 has been faid upon this fubject it is evident, that all 

 the phenomena of .colours depend upon two proper- 

 ties of light, the refrangibiiity and reflexibility of its 

 rays. 



The blue colour of the atmofphere has been beauti- 

 fully accounted for by Mr. Delaval, in the experiments 

 already detailed. The atmofphere he confiders as a 

 femi-pellucid medium, which abounds in volatile and 

 evaporable particles, difengaged from natural bodies 

 by feveral operations, as fermentation, effervefcence, 

 putrefaction, &c. Thefe particles differ greatly in 

 denfity, &c. from the air, and, as they reflect a white 

 light, may be confidered as fo many white particles 

 diffufed through the pellucid colourlefs air. In this 

 refpect the atmofphere is fimilar to the femi-pellucid 

 medium, which is formed by a mixture of arfenic with 

 glafs. In both thefe fubftances, whilft the white par- 

 tieles are rarely diffeminated through the tranfparent 

 medium, the lefs refrangible rays are tranfmitted 

 through the intervals which intercede the particles f , 



but 



* Prieftley's Hift. Opt. p. 617. , 



f On this account, diitant mountains covered with fnow (which 



it is well known reflect all the rays of the fun) appear, when the 



VOL. I. U air 



