Solar Eclipse. 303- 



clouds and vapours are never found above the highest moun- 

 tains on the earth, none of which exceed -j-^-g- of the moon's 

 diameter, it is to be concluded that the grosser part of our at- 

 mosphere, capable of producing any sensible refraction of the 

 rays of light, does not extend beyond that elevation. Conse- 



dia. dia> 



quently, as 408 : 139 : : -^^ : xsVg' which, on the same prin- 

 ciple of analogy as is recognised in the theory itself, gives a 

 very near approximation to the true extent of lunar atmo- 

 spheric refraction ; a quantity, nevertheless so small as not to 

 subtend an angle of 1", and which does not exceed one mile in 

 perpendicular elevation, whereas the circles of what is termed 

 refracted light, appear to have extended at least 160 miles 

 beyond the moon's surface. 



The second theory is found on a property of matter, by 

 which it deflects, or attracts towards the perpendicular, the rays 

 of light which pass very near its surface. From experiment, 

 however, the quantity of light thus deflected is so inconsider- 

 able as to be apparently inadequate to the production of the 

 effects in question. If the observations of Maraldi are to be 

 considered as accurate, the body of the moon must be regarded 

 as even altogether destitute of tliis property of terrestrial 

 matter. But we are not left to doubtful speculation on this 

 subject. Since it has been determined that the aberration of 

 the rays produced by the attraction of the moon's surface does 

 not exceed 3.5", a quantity which certainly appears insufficient 

 to account for the phenomena above described*. Such are the 

 limits prescribed, both by analogy and experiment, to the opera- 

 tion of those causes, which are assumed as perspectively pro- 

 ducing the efi'ects in question. That they cannot far exceed 

 these limits also appears, for if the inflection of the solar rays, 

 whether effected by atmospheric refraction, or by the attraction 



• M. du Sejour troiiva qu'il falloit faire 1' inflexion (de rayons que 

 rasent les bonds dc la hine) d'cnviron 3|". M. Mechaiu, et M. Lexellc ont 

 tro(iv6 le m^me resiiltat, — La Landc Aslron toni. ii. ji. 445. Sejuur. 

 Traiti dea Monvem., &c. 



