SENSE OF SIGHT. 585 



were just below the horizon it would appear just on the horizon. 

 Further, if the moon were just on the horizon the lower part of 

 the moon would be raised (by refraction) more than the upper 

 part, because refraction diminishes rapidly after leaving the hori- 

 zon, so the vertical (apparent) diameter of the moon would be /ess, 

 but the horizontal diameter would not change i. e., the area (size) 

 of the moon at the horizon would be smaller by refraction. 



"As to the apparent greater size of the moon when near the 

 horizon, I believe that as stated in most astronomy books, say 

 Charles Young's or Todcl's New Astronomy it is due to the fact 

 that in looking at the moon in that position we see many well- 

 known objects as well : houses, trees, etc. ; and as we have a rough 

 idea of the size of these objects and also know they are nearer us 

 than the moon is, we get the idea of largeness. In the zenith we 

 have no objects for comparison, and as angular magnitude alone 

 conveys no meaning, we have nothing to guide (or misguide) our 

 judgment. Of course, when in the zenith, the moon's angular size 

 is very slightly greater, as we are nearly 4000 miles nearer i. e., 

 by the radius of the earth. I may remark that the common notions 

 as to the moon's size are curious : one person may say it looks as 

 1 big as a dinner-plate/ and another, i as big as a half-dollar/ but 

 in no case which I have examined have these thoughtful people 

 made any mention of how far from them the dinner-plate was sup- 

 posed to be at the time." 



Prof. C. A. Young, the well-known astronomer, of Princeton 

 University, was written to for his explanation of the phenomenon ; 

 his answer is as follows : "As to the phenomenon to which you 

 refer, the apparent enlargement of celestial objects near the horizon, 

 I am satisfied with the generally received explanation that it is 

 due to the fact that while the angular diameter of the object re- 

 mains practically the same as at the zenith (really a little less), the 

 object is instinctively referred to a greater distance because of the 

 number of intervening objects by which we judge of distance. 



" If the sun or moon be looked at through a smoked glass, 

 which leaves the disc visible while hiding the horizon and inter- 

 vening objects, the sun or moon immediately shrinks to the same 

 apparent diameter as if higher up. At least it is so with me, and 

 with most persons whom I have seen try the experiment. Look- 

 ing through a tube like a piece of gas-pipe 12 inches long is just 

 as effectual ; what is necessary is to cut out all objects and the 

 horizon line while still leaving the moon (or sun) visible. But 

 some persons say they do not get the effect. In their case I judge 

 that the habitual sense of the horizon as distant comes in to main- 

 tain the illusion, even though they cannot actually see it, but I am 

 aware that some physiologic psychologists decline to accept the 

 explanation" 



