454 LECTURE XXXVIII. 



at sea, is smaller than its true distance. Tiie city of Lo rid on is unquestion- 

 ably larger than Paris; but the ditTerence appears at first sight much greater 

 than it really is; and the smoke, produced by the coal fires of London, is proba- 

 bly the principal cause of the deception. 



The sun, moon, and stars, are much less luminous when they are near 

 the horizon, than wdien they are more elevated, on account of the greater 

 quantity of their light that is intercepted, in its longer passage through the 

 atmosphere: we also observe a much greater variety of nearer objects almost 

 in the same direction: we cannot, therefore, help imagining them to be 

 more distant, when they rise or set, than at other times; and since they sub- 

 tend the same angle, they appear to be actually larger. For similar reasons 

 the apparent figure of the starry heavens, even when free from clouds, is 

 that of a flattened vault, its summit appearing to be much nearer to us than 

 its horizontal parts, and any of' the constellations seems to be considerably 

 larger when it is near the horizon than when in the zenith. (Plate XXX. 

 Fig. 438.) 



The faculty of judging of the actual distance of objects is an impediment 

 to the deception, which it is partly the business of a painter to produce. Some 

 of the effects of objects at different distances may, however, be imitated in 

 painting on a plane surface. Thus, supposing the eye to be accommodated to 

 a given distance, objects at all other distances may be represented with a 

 certain indistinctness of outline, which would accompany the images of the 

 objects themselves on the retina: and this indistinctness is so generally 

 necessary, that its absence has the disao:reeable efltcct called hardness. The 

 apparent magnitude of the suSjects of our design, and the relative situations 

 of the intervening objects, may be so imitated by the rules of geometrical 

 perspective as to agree perfectly with nature, and we may still further im- 

 prove the representation of distance hy attending to the art of aerial perspec- 

 tive, which consists in a due observation of the loss of light, and the bluish 

 tinge, occasioned by the interposition of a greater or less depth of air between 

 us and the ditfcrent parts of the scenery. 



We cannot indeed so arrange the picture, that either tlie focal length of 

 the eye,- or the position of the optical axes, may be such as would be required 



