and on the Elementary Colors of the Solar Spectrum. 
server when the spectrum arises from luminous agencies so little 
energetic as are those of the first periods of incandescence, © ~~ 
To a perfectly sensitive eye the variations of length would ev- 
idently have taken place in the direction of the more refrangible 
rays only; and all the spectra would have commenced at the ex- 
treme limit of the red rays. 
It results from all these observations, that when the incandes- 
cence of a body becomes more and more vivid and brilliant by 
the elevation of its temperature, there is not only an augmenta- 
tion in the intensity of the resulting light, but also in the variety 
of elementary colors which compose it; there is too an addition 
ol rays so much the more refrangible as the temperature of the 
incandescent body is higher. In this there is therefore established 
an intimate analogy between the progressive development of light 
and that of heat. Indeed, as soon as I convinced myself of 
the immediate transmission of every variety of radiant heat 
through rock salt, I availed myself of that valuable property to 
study the refraction of heat from different sources; and I discov- 
ered that radiations coming from those of a high temperature con- 
tain elements more refrangible than those which are derived from 
sources that are not so hot. 
_ After having pointed out how very important for the theory of 
the identity of calorific and luminous radiations is the analogy or 
rather absolute equality which exists between his experiments 
and mine on the successive appearance of the elementary rays, 
Professor Draper passes forward to investigate the law of the in- 
crease of the luminons and calorific radiations, according as the 
temperature of the source of heat is elevated. i 
_ Bouguer has shewn, that a difference of ;', in the quantity of 
light which acts on the eye is insensible ; and that thus this frac- 
Hon constitutes the limit of perceptible variations. From this it 
arises that we easily perceive differences in the intensity of light 
when they exceed even to the smallest extent the sixtieth part of 
the entire quantity. Let there be placed, for example, an opaqt 
cylinder between the red hot platina, or other source of light, and 
a sheet of white paper arranged at such a distance that its surface 
tay be illuminated all over by the light of the shining body, ex- 
The on those parts where the shadow of the cylinder is thrown. 
then let there be a lamp furnished with a metallic tube, in which 
there is a small opening ; let the rays of the lamp which go out 
through the aperture fall on the paper enlightened by the lumin- 
ous source of which we wish to measure the intensity; let the 
lamp be gradually approached until the shadow of the cylinder is 
no longer sensible ; let this experiment be repeated in each of the 
cifferent cases we wish to compare with one another, measuring 
each time the distance from the lamp to the paper. It is then 
evident that the values sought will be in the inverse ratio of the 
r; 
