58 FRAGMENTS OF SCIENCE. 
Kecent experiments have proved that elementary bodies, 
except under certain anomalous conditions, belong to the 
class of bad radiators. An atom, vibrating in the ether, 
resembles a naked tuning-fork vibrating in the air. The 
amount of motion communicated to the air by the thin 
prongs is too small to evoke at any distance the sensation 
of sound. But if we permit the atoms to combine chem- 
ically and form molecules, the result, in many cases, is an 
enormous change in the power of radiation. The amount 
of ethereal disturbance, produced by the combined atoms 
of a boSy, may be many thousand times that produced by 
the same atoms when uncombined. 
The pitch of a musical note depends upon the rapidity 
of its vibrations, or, in other words, on the length of its 
waves. Now, the pitch of a note answers to the color of 
light. Taking a slice of white light from the sun, or from 
an electric lamp, and causing the light to pass through an 
arrangement of prisms, it is decomposed. We have the 
effect obtained by Newton, who first unrolled the solar 
beam into the splendors of the solar spectrum. At one 
end of this spectrum we have red light, at the other, vio- 
let; and between those extremes lie the other prismatic 
colors. As we advance along the spectrum from the red 
to the violet, the pitcli of the light if I may use the ex- 
pression heightens, the sensation of violet being pro- 
duced by a more rapid succession of impulses than that 
which produces the impression of red. The vibrations of 
the violet are about twice as rapid as those of the red; in 
other words, the range of the visible spectrum is about an 
octave. 
There is no solution of continuity in this spectrum; 
one color changes into another by insensible gradations. 
It is as if an infinite number of tuning-forks, of gradually 
augmenting pitch, were vibrating at the same time. But 
turning to another spectrum that, namely, obtained from 
the incandescent vapor of silver you observe that it con- 
sists of two narrow and intensely luminous green bands. 
Here it is as if two forks only, of slightly different pitch, 
were vibrating. The length of the waves which produce 
this first band is such that 47,460 of them, placed end to 
end, would fill an inch. The waves which produce the 
second band are a little shorter; it would . take of these 
47,930 to fill an inch. In the case of the first baud, the 
