then consist of a series of colored images of the slit arranged side by side. This is called a pure spectrum, to distinguish it from the spectrum shown in Fig. 449, in which no lens was used to bring the rays of each particular color to a particular point, and in which there was therefore much overlapping of the different colors. If the slit and screen are exactly at conjugate foci of the lens, and if the slit is sufficiently narrow, the spectrum will be seen to be crossed vertically by certain dark lines.

S

L

P

Red

Blue

These lines were first observed by the Englishman Wollaston in 1802, and were first studied carefully by the German Fraunhofer in 1814, who counted and mapped out as many as seven hundred of them. They are called, after him, the Fraunhofer lines. Their existence in the solar spectrum shows that certain wave lengths are absent from sunlight, or, if not entirely absent, are at least much weaker than their neighbors. When the experiment is performed as described above, it will usually not be possible to count more than five or six distinct lines. Kirchhoff explained these lines.

FIG. 457. Arrangement for obtaining a pure spectrum

477. Explanation of the Fraunhofer lines. Let the solar spectrum be projected as in § 476. Let a few small bits of metallic sodium be laid upon a loose wad of asbestos which has been saturated with alcohol. Let the asbestos so prepared be held to the left of the slit, or between the slit and the lens, and there ignited. A black band will at once appear in the yellow portion of the spectrum, in the place where the color is exactly that of the sodium flame itself; or, if the focus was sufficiently sharp to permit a dark line to be seen in the yellow before the sodium was introduced, this line will grow very much blacker when the sodium is burned. Evidently, then, this dark line in the yellow part of the solar spectrum is in some way due to sodium vapor through which the sunlight has somewhere passed.

The experiment at once suggests the explanation of the Fraunhofer lines. The white light which is emitted by the hot nucleus of the sun, and which contained all wave lengths, has had certain wave lengths weakened by absorption as it passed through the vapors and gases surrounding the sun and the earth. For it is found that every gas or vapor will absorb exactly those wave lengths which it is itself capable of emitting when incandescent. This is for precisely the same reason that a tuning fork will respond to, that is, absorb, only vibrations which have the same period as those which it is itself able to emit. Since, then, the dark line in the yellow portion of the sun's spectrum is in exactly the same place as the bright yellow line produced by incandescent sodium vapor, or the dark line which is produced whenever white light shines through sodium vapor, we infer that sodium vapor must be contained in the sun's atmosphere. By comparing in this way the positions of the lines in the spectra of different elements with the positions of various dark lines in the sun's spectrum, many of the elements which exist on the earth have been proved to exist also in the sun. For example, Kirchhoff showed that the four hundred and sixty bright lines of iron which were known to him were all exactly matched by dark lines in the solar spectrum. Fig. 458 shows a copy of a photograph of a portion of the solar spectrum in the middle, and the corresponding bright-line spectrum of iron each side of it. parison of solar (This was taken from a portion of the blue

FIG. 458. Com

and iron spectra

only of the spectrum.) It will be seen that each bright line of iron coincides exactly with a dark line of the solar spectrum.

478. Doppler's principle applied to light waves. We have noted (see "The Doppler effect," § 388, p. 353) that the effect of the motion of a sounding body toward an observer is to shorten slightly the wave length of the note emitted, and the effect of motion away from an observer is to increase the wave length. Similarly, when a star is moving toward the earth, each particular wave length emitted will be slightly less than the wave length of the corresponding light from a source on the earth's surface. Hence in this star's spectrum all the lines will be displaced slightly toward the violet end of the spectrum. If a star is moving away from the earth, all its lines will be displaced toward the red end. From the direction and amount of displacement, therefore, we can calculate the velocity with which a star is moving toward the solar system or receding from it. Observations of this sort have shown that some stars are moving through space toward the solar system with a velocity of 150 miles per second, while others are moving away from it with almost equal velocity. The whole solar system appears to be sweeping through space with a velocity of about 12 miles per second; but even at this rate it would be at least 70,000 years before the earth would come into the neighborhood of the nearest star, even if it were moving directly toward it.

SUMMARY. Continuous spectra are produced by light from incandescent solids and liquids.

Bright-line spectra come from incandescent vapors or gases. Absorption spectra are produced by light passing from an incandescent solid or liquid through an incandescent vapor or gas.

QUESTIONS AND PROBLEMS *

1. In what part of the sky will a rainbow appear if it is formed in the early morning?

2. Why is a rainbow never seen during the middle part of the day?

3. Explain the cause of the formation of a continuous spectrum when light from an incandescent lamp is passed through a prism.

* Supplementary questions and problems for Chapter XX are given in the Appendix.

4. How does a continuous spectrum differ from that obtained when the source of light is a Bunsen burner in which sodium is being vaporized?

5. Why do we believe that there is sodium in the sun?

6. What sort of spectrum should moonlight give? (The moon has no atmosphere.)

7. If you were given a mixture of a number of salts, how would you proceed, with a Bunsen burner, a prism, and a slit, to determine whether or not there was any calcium in the mixture?

8. Draw a diagram of a slit, a prism, and a lens, so placed as to form a pure spectrum.

CHAPTER XXI

INVISIBLE RADIATIONS

RADIATION FROM A HOT BODY

479. Invisible portions of the spectrum. When a spectrum is photographed, the effect on the photographic plate is found to extend far beyond the limits of the shortest visible violet rays. These so-called ultra-violet rays have been photographed down to a wave length of .00000136 centimeter, which is only one thirtieth the wave length of the shortest

violet waves.

The longest rays visible in the extreme red have a wave length of about .00008 centimeter, but delicate thermoscopes reveal a so-called infra-red portion of the spectrum, the investigation of which was carried, in 1912, by Rubens and Von Baeyer of Berlin, to wave lengths as long as .03 centimeter, 400 times as long as the longest visible rays.

FIG. 459. The Crookes radiometer

The presence of these long heat rays may be detected by means of the radiometer (Fig. 459), an instrument perfected by E. F. Nichols at Dartmouth. In its common form it consists of a partially exhausted bulb, within which is a little aluminum wheel carrying four vanes blackened on one face and polished on the other. When the instrument is held in sunlight or before a lamp, the vanes rotate in such a way that the blackened faces always move away from the source of radiation, because they absorb ether waves better than do the polished faces, and thus become hotter. The heated air in contact with these faces then exerts a greater pressure against them than does the air in contact with the polished faces.