of the Heavenlj/ Bodies. 17 



therefore to be correctly determined, any estimate can or ought 

 to be depended upon, except derived from observations made 

 at the time, and upon the occasion, during the existing state of 

 the atmosphere, and independently of the principal observations, 

 will hereafter appear. Many modes of making these observa- 

 tions will occur to practical astronomers, to whom we defer 

 with all due respect in matters purely astronomical. 



The apparent and extraordinary changes of dimensions, and 

 of figure, in the sun and moon in stations just above the horizon, 

 have been referred by various conjectures to various causes. 



Des Cartes, Wallis, and others, suppose that a better judg- 

 ment being formed of the distance of the moon by comparison 

 with objects in the horizon, she is considered as more remote, 

 and therefore appears larger in the horizon than in the zenith. 



Ptolemy, considered the effect, as in part a fancied, in part 

 an actual enlargement of apparent disc, and conformably to 

 this latter opinion Roger Bacon ascribes the enlargement to 

 refraction. Gassendus ascribes the appearances to an in- 

 creased dilatation and flatness of the pupil of the eye in 

 less light producing a larger picture on the retina ; Berkeley 

 to the diminished horizontal light of the moon ; Smith to the 

 apparent figure of the sky as being less than an hemisphere, 

 the moon retaining her size unchanged, and appearing, upon 

 the principles of perspective, larger at supposed remoter 

 distances. 



The horizontal diameters of sun and moon have been sub- 

 jected to actual measurements with varying conclusions. 

 Riccioli afiirms, that together with Grimaldo, having, with a 

 sextant, carefully and repeatedly measured the horizontal dia- 

 meter of the sun, one by the right, the other by the left limb, 

 they distinctly ascertained the increase to be what the naked 

 eye exhibits. Almost all other philosophers have considered 

 the appearances to be delusive, to be optical deceptions, and 

 have affirmed that measurements by instruments give no in- 

 crease of dimensions. Almost all are of opinion that the re- 

 fractions, which, as they suppose, produce these appearances, 

 can only be in vertical circles. 



V^OL. X. C 



