ANSWERS TO OPTICAL dUESTlONS. 239 



to the relative pofition of its parts conlidered acrofs the con- 

 cavity, and the contrary along the lines oF convexity; — or, 

 while thefe lines have their pofition altered by rotation of the 

 mirror, the image itfelf w ill alio appear to revolve with twice 

 the velocity. Thefe effeds, as well as that of the extreme 

 diftortion of the image, feen when the objed is within the cen- 

 tre of concavity, will be eafily apprehended by the reader, 

 whofe notions of optics are fufficient to enable him to under- 

 ftand the preceding paragraph. 



3. The enlarged appearance of the fun and moon at low Why the (ky 

 altitudes or near the horizon, and the apparent flatnefs of the and^^the fun and 

 concavity of the fky have been explained, as my correfpondent moon large in 

 obferves, by reference to the diminithed light of the heavenly *^= ^onion. 

 bodies, when it pafles through a long portion of the atmof- 

 phere, and likewife to the notions of diftance which are ob- 

 tained by looking over a long row of terreftrial objeds. That 

 is to fay the explanation is grounded on the fuppoiition or no- 

 tion of greater diftance along the horizontal line deduced from 

 confideration of the aerial and the geometrical perfpe6tive. 

 The obfervations and objections of R. B. appear to me to be 

 perfectly well founded. I think the effedl is produced almoft 

 entirely by the geometrical perfpeflive, not of the houfes, trees, 

 and other land objeCls, but of the clouds themfelves. When 

 the clouds are in low diftinft flakes with clear openings be- 

 tween them, the angular magnitude both of the clouds and 

 their intervals will be greater near the zenith, and will dimi- 

 nifti as the zenith diftance encreafes, fo as on fome occafions 

 to exhibit the appearance of objeds running out to an immenfc 

 cli^ftftnce in the concavity of an extremely flat dome. On the 

 contrary, when lofty towering mafles of clouds rife from the 

 horizon to a confiderable elevation facing the fetting fun, or 

 otherwife fo circumflanced that the tint of the whole mafs fliall 

 be very little varied, the mind, fo far from adopting the notion 

 of an extended dome in that part, fliall receive the impreflion 

 of an immenfely elevated wall, with very little curvature as it 

 rifes. Both thefe conditions are not unfrequently to be feen at 

 one and the fame time, and the flcy is then,, according to the 

 obfervation of R. B. very far from exhibitinij; a figure of regular 

 dimenfions. A great viariely of intermediate forms and groups 

 of the clouds often prefent themfelves, by which the apparent 

 figure of the (ky will be made to difl^er from thefe extreme 

 9afes. When the Iky is perfectly and clofely clouded, we 



hav^ 



