f2So ^^ THE nORIZOS'X.'SL MOON', 



placed. Wliat is this ? It is certainly fomethlng real and ffia- 

 terial, or elfe it would not appear coloured. For bodies, to 

 appear coloured, muft have parts of fome determinate mag- 

 nitude. 



Where doth it exift ? Not in thofe immenfely diftant parts 

 of fpace, where the heavenly bodies revolve. For if thole 

 fpaces contained any bodies of a determinate magnitude, and 

 confequently of a determinate deniily, the heavenly bodies 

 could not continue through ages to revolve in the fame peri- 

 odic times J iheir momentum would be diminithed by refift- 

 ance, and the periodic times of their revolution would change. 

 The blue flcy therefore canr.ot be placed beyond the atmofphere 

 of our earth. The fmalleft parts of bodies, that are coloured 

 are blue, and the blue f]<y is therefore either the atmofphere 

 itfelf or the fmallefl and mofl: elevated vapours afcending in 

 it, or both together. The heavenly bodies Qiine through it, 

 and therefore it cannot be opaque; it is itfelf of a blue colour, 

 and therefore is not perfectly tranfparent. 



Though we are ignorant of the exad height of the atmof- 

 phere, yet we may take it for granted, that it does not extend 

 as far as the moon, and therefore that the diftance of its far- 

 ihefl points from the centre of the earth has a finite ratio to the 

 femi-diameter of the earth. That the ratio is probably lefs 

 than 2: 1. 



If therefore the blue heavens which furround the earth, and 

 are concentric with it, have a femidiameter not double that of 

 the earth, their horizontal points as viewed from the earth, 

 iHuftbe farther from us than any that are nearer the zenith. 



For let A C F (PlateVU. Fig, ].) reprefent a great circle 

 of the earth, and A C be its radius, and let the circle D BE G 

 reprefent a great circle of the atmofphere drawn with a radius 

 AB not =2. AC, the line C D yB C; HCy BC. 

 —which is not ^ ^'^^ ^^^ ^^ ^^'^ folution of the flattened appearance of the 

 always alike. heavens, by obferving that, when the fky is uniformly over- 

 caft with clouds, the concave fuperficies appears coniiderably 

 flatter than when the (ky is ferene. In the former cafe, the 

 two concentric circles in the figure approach nearer to each 

 other, the clouds being nearer to the earth than the fky is, and 

 the ratio of D E to C B muft of courfe increafe. 

 Explanation ^^^ lo proceed. When any bodies fituated behind a feml- 



tranfparent fcreen are feen through it, they will appear to be 



fi>ved 



fiom a diagram 

 ot the Iky, that 



