76 THE PROBLEM OF THE GALLOPING HORSE 



invariably — departs greatly from scientific truth, and it 

 is a question as to whether he is justified in what he does. 

 Take first the case of the low-lying moon near the horizon 

 as contrasted with the high moon. Everyone knows that 

 the moon (and the sun* also) appears to be much bigger 

 when it is low than when it is high. Everyone who has 

 not looked into the matter closely is prepared to maintain 

 that the luminous disc in the sky — whether of moon or 

 of sun — not merely seems to, but actually does, occupy a 

 bigger space when it is low down near the horizon than 

 when it is high up, more nearly overhead. Of course, 

 no one nowadays imagines that the moon or the sun 

 swells as it sinks or diminishes in volume as it rises. 

 Those who think about it at all, say that the greatei 

 length of atmosphere through which one sees the low sun 

 or moon, as compared with the high, magnifies the disc as 

 a lens might do. This, however, is not the case. If we 

 take a photograph of the moon when low and another with 

 the same instrument and the same focus when it is high, 



* What we may call " the visual size " of the sun happens to be owing to 

 its far greater size and its far greater distance from us— very nearly the 

 same as that of the moon — and is subject to the same numerical law of 

 apparent diameter, viz. a disc of any given measurement in diameter will 

 cover it exactly when held at a distance from the eye which is 115 times 

 that measuremept. 



Plate V. — The traclc of the rising moon registered by continuous exposure 

 of a photographic plate. It is given here in order to show that the dia- 

 meter of the visible disc of the moon does not diminish as it rises. The 

 slight increase in the breadth of the track registered by the moon's disc 

 is probably due to a little distortion caused by the side portion of the 

 lens. After M. Flammarion. The actual width of the moon's disc 

 as printed here is a little over one eighth of an inch, which, if we regard 

 it as " a picture'' and not merely as a mechanical record, implies that 

 the observer's eye is only about 144 inches distant from the picture 

 plane instead of the more usual 18 inches, which corresponds to a 

 diameter of the pictured moon's disc of between |th and |th of an 

 inch ('156 inch). 



