On Refraction. 203 
to the present theory, two images could not possibly be formed by 
refraction at b and f; for a being at right angles to the plane dc, 
the rays should suffer no refraction, but proceed on to the vertex, 
The very same mistake which induced optical writers to suppose 
from analogy that rays converged in the body of a convex lens, 
made them also suppose that rays were turned in the body of a 
prism to the thickest part as well within as without the medium. 
Let us now examine this experiment according to the present Jaws 
of optics. . . 
Let the angle CAI (fig. 11) bea right angle; then the whole 
refraction is at C; and in this‘ease, DC A:ACD:: m:m—n. 
Also, since the right angle D C | is equal to the’sum of the two 
ACI, AIC, take away the common angle ACI, and the re- 
‘maining angles DC A, AIC are équal. Consequently AIC 
“:ACI::m:m—n, Now I would beg leave to ask, Does any 
light in this experiment pass through the plane YZ? The ray 
Q A is undoubtedly turned to the thickest part of the prism 5 not, 
as Newton and his followers suppose, from any principle of at- 
traction, but simply because it strikes the plane I Z obliquely, 
and there forms at: image, which moves downwards. Let us 
vary this experiment. I placed the plane of the prism on a small 
hole, cut in a large sheet of pasteboard, and perceived two images 
of the hole formed in the planes, as already described with the 
sovereign. I now removed this sheet of pasteboard with the 
prism into the sun-beams, as represented (see fig. 12), and 
found that the rays passed through both planes 8. The sun 
passes through a hole in the pasteboard, and, striking the plane 
AB perpendicularly, forms an image at d, which image sends 
rays to form other images at f and g. Here we have two spec- 
tra at f and g, the one ascending the other descending in conse- 
quence of striking the planes obliquely. In this experiment op- 
ticians are necessarily obliged to relinquish one of their favourite 
laws, “ that rays striking at right angles to plane surfaces suffer 
no refraction; for it cannot possibly be denied, Ist, that the rays 
strike the plane AB at right angles; and 2dly, that the rays di- 
verge: otherwise they could not come through the planes A C and 
BC. Are the rays refracted in opposite directions? or are they 
attracted and repelled in opposite directions? But if we admit 
that an image is formed at d, we can easily account for the two 
reflections at f and g. Had Sir Isaac Newton been acquainted 
with the formation of two spectra {and I cannot but express sur- 
prise that he was not), he never could haye maintained the doc- 
trines he did. Here I cannot but notice a curious fact in regard 
to the prism, although not immediately connected with the doc- 
trine of refraction, When the sun-beams are passed through the 
lower refracting angle, as it is called, on emergence they ascend 
c2 and 
