358 On the Refraction of Light. 
‘ : 
5. The bed of a river, ora picture behind a plate of glass, would 
be seen beautifully enamelled with spectra. 
These, and many more phenomena of a similar nature, would be 
manifest, if the colorific rays possessed different degrees of refran- 
gibility ; but as no such appearances are observed to take place, it 
follows that the colorific rays are all equal in point of refrangibility. 
In order, therefore, to account for the prismatic analysis of the solar 
rays, we must seek for some other cause than the different refrangi- 
bilities of the colors of light. feb 
- A theory which appears to me to be less objectionable than any 
already advanced, more consistent with phenomena, and more con- 
formable to facts, is the following. ty 
It is manifest, even to the organ of vision, that there exists, on th 
solar spectrum, a gradation of densities from the red to the violet rays} 
the former being most dense, the latter most rare, and the colors in- 
cluded between these extremes possessing i ate degrees of den 
sity.* As a-corollary to this, it follows that the attraction existing be- 
tween homogeneous colorific atoms is most energetic in the red, an 
most feeble in the violet rays. This density, or concentration of color- 
ific atoms, varies not only in rays of different colors, but in different rays 
of the same color. This may readily be proved. Into a dark cham- 
_ ber admit a sun beam through red glass, and fix the prism in sucha 
‘Manner as to refract the rays; receive the image on a white surface 
at a distance, and it will be seen that the red color is most conceD- 
trated at the base, and becomes gradually less dense from the lower 
to the upper extremity of the image. Lest this phenomenon might 
have arisen from any inequalities in the substance or polish of the, 
colored medium, I admitted into a dark chamber, the red rays 
emerging from a prism, and having refracted them by a second prism; 
I received the image on a white paper. It differed in no respect 
from that produced by red glass. In this experiment it is to be re- 
marked that the image is not circular, as Newton imagined, 
quadrilateral. The quantity of heat, which accompanies the differ- 
ent colors of light, appears to depend on the density of the colorific 
ray, and to be proportional to it; for, as we proceed upward on the 
| solar spectrum with the differential thermometer, a decrement of sen- 
= 
1° 
* 
* Sir Isaac Newton was of opinion that the particles of red light were not only 
more contiguous to one another, buf larger than those of the other colors, and cou- 
sequently that they possessed a greater specific gravity. 
