202. On Refraction. 
a refracted image we see, is evident from our being able to see it 
in every direction floating on the surface of the water: if re« 
fracted, we could only see it in the direction of the refracted rays. 
When the eye is placed immediately over the half-crown, looking 
down into the water, we see the image, not the piece of money, 
one-fourth nearer to the eye: here there can be no refraction, as 
the rays coming to the eye must be at right angles to the surface 
of the water: here there is no angle of incidence; no angle of 
refraction; no ratio of 3 to4. In fact, this simple experiment 
rebels against almost all the laws of optics. Snellius was the first 
who supposed he discovered a constant ratio in refraction; he 
used the secants of the complements instead of the sines used 
by the celebrated Des Cartes. As his doctrines are founded on 
this experiment, [ think it necessary to make a few observations. 
Supposing the surface of the water to be A B (fig. 9), and an 
object under it at D, which to the eye at F appeared as it were 
in the line TC. He produced T C till it met in G with the per- 
pendicular D A to the surface AB. Then he argued, that the 
image of the object D appeared at G, and that C D was toC G 
in a certain given ratio as 4 to 3 in water. 
The following objections may be made: 1. The images can 
be seen by an eye at B on a plane with the surface of the water. 
2. This image can be perceived in every direction above, below, 
and on a plane with the surface of the water, which could not be 
the case with a refractedimage. 3. There is no reason whatso- 
ever that the ray D C should be refracted in the diagonal at plane 
surfaces, except for the purpose of supporting the theory. On 
the contrary, there is every reason to prove that the rays move 
parallel, for the image is perceived at A immediately over the 
piece of coin, An eye at A looking down into the tumbler sees 
the piece of money one-fourth nearer. Here, according to opti- 
cians, the rays are not refracted; yet they cannot deny that the 
piece of money appears nearer the eye, and somewhat magnified. 
If it were the object and not an image they saw, it would appear 
at the same distance as in air. It is agreed on all hands, that 
every refracting surface forms a reflected image; why resort to 
any other means? I shall now proceed to extend this experi- 
ment toa medium terminated by two plane surfaces inclined to 
one another, such as an equilateral prism. 
Having placed a sovereign under the plane of a prism (fig. 10.) 
resting on the table, I found that two reflected and not refracted 
images were formed in each plane, as represented in the follow- 
ing figure. a The Ssvereign placed under the plane dc of an 
equilateral prism, forms an image at a; which image sends 
images to b and f. That these are reflected and not refracted 
images, is so evident as scarcely to require remark. According 
‘ to 
