870 PHILOSOPHICAL TRANSACTIONS. [aNNO l687 . 



vation, I suppose, for I have never examined it myself, that the apparent mag- 

 nitude taken by instrument, however the fancy may apprehend it, is not greater 

 at the horizon than when higher. And Mr. Caswell here affirms the same 

 thing. And, though I have not myself made the observation, I do not doubt 

 but the thing is so. For it is agreed, that refraction near the horizon, though 

 as to appearance it alters the altitude of the thing seen, yet it alters not the azi- 

 muth at all. And it must needs be so : for, since this equally respects all points 

 of the horizon, let the refraction be what it will, the whole horizon can be but 

 a circle ; so that there is no room for the breadth of a thing, as to the angle 

 at the eye, to be made greater, whatever its tallness may, the refraction not 

 equally affecting all parts in the circles of altitude. Nor is there any reason why 

 this should rather thrust the other, than that the other thrust this, out of place. 

 But in the altitude it is otherwise, for while what is near the horizon is enlarged, 

 that which is further off is thereby contracted, which as to the azimuth or hori- 

 zontal position cannot be. 



In spectacles indeed it is otherwise, for they represent the object every way 

 enlarged, and do thereby hide the adjacent parts. But in refraction by vapours, 

 supposing all parts of the horizon equally affected by them, one part cannot be 

 expanded in breadth, whatever it may be as to the height, without thrusting out 

 another ; for the whole horizon can be but a circle, and why one part rather 

 than another ? Unless we would say, that the rays of a lucid body expand them- 

 selves every way, to the prejudice of the parts adjacent, by covering them. 



But supposing that the sun or moon's apparent horizontal diameter, taken by 

 instrument near the horizon, is the same as taken in a higher position, I take 

 its imaginary greatness, which is fancied near the horizon, to be only a decep- 

 tion of the eye, or rather the imagination from the eye. For the imagination 

 does not estimate the greatness of the object seen only by the angle which it 

 makes at the eye ; but by this compared with the supposed distance. True it is, 

 that, caeteris paribus, we judge that to be the greater object which makes at the 

 eye the greater angle, but not so if apprehended at different distances. For if 

 through a casement, or smaller aperture, we see a house at a 100 yards distance, 

 this house, though seen under a less angle, does not to us seem less than the 

 casement through which we see it, or this greater than that, because it makes 

 at the eye the greater angle ; but the imagination makes a comparative estimate 

 from the angle and distance jointly considered. So that, of two things seen 

 under the same or equal angles, if to one of them there be ought which gives 

 the apprehension of a greater distance, that to the imagination will appear 

 greater. ~ 



Now one great advantage for estimating of a thing seen is, from the variety 



