16 
tinued; and if it is still heated, fresh sets of colours are formed; which 
is not only a proof of a sort of permanence of colours brought on 
by heat, but of their regularity in successively forming rings, which 
is rather universal vviiere colours are formed, especially as they de- 
pend upon some sort of prismatic agency: but the yellow and red 
are not very bright, as the steel does not reflect them so well as 
many other substances, although the blue is generally beautifully 
reflected. Nature, however, produces them all most brilliantly on 
some iron ores and other oxides. 
Having gone thus far, I found it might give much satisfaction if 
I were to examine the light from the full sun's rays, as an original 
source, in the open air. I therefore took the usual three-sided 
prism, and was highly gratified by observing white in the refracted 
reflexion of the sun, as seen of an oval form, light on the face of the 
prism, and the three primitive tints, a fine yellow, a rich red, and 
a light blue: (see Tab. 2. Jig. 1 and 2,) and I found, by turning 
the prism, very beautiful phaenomena took place within it, scarcely, 
if at all, before noticed in open light. The rays, as they become 
more oblique, spread over each other, the yellow over the white to 
the blue forms green, while the blue on tlie darker side is changed 
into violet and indigo. The red, by the same motion, passes over 
some of the yellow spreading into orange, and thus are formed the 
regular seven prismatic tints or spectrum. At the same time as the 
above, may be seen lefracted immediately from the sun's most bril- 
liant rays, upon any object within a few inches, a fine image of the 
prism bordered lengthwise, by the same three tints: {see Tab. 2, 
fig. 3.) This shows the primitive tints on a larger scale, and will 
therefore give a fuller exjjlanation of the whole. Thus the middle 
I call white, as the more direct light, the yellow is below it, the red 
lowest, and the blue on the uppermost or opposite side, blending 
but a little with each other, and least so when most vivid, or when 
the oval reflexion. Jig. 2, is nearest in form to a circle. When a 
strip of white paper is seen through a prism, at a proper distance, on 
a mahogany or dark table, blue will terminate one edge, and red 
