4. Methods of Recording Radiations.— Having separated the different lines in any spec trum, the next step, in general, is to record the intensity, position (wave-length) and physical character of each of these. Just how this is accomplished will depend partly upon the por tion of the spectrum under examination and partly upon the purpose of the work. If the physical character of a line be the matter in question, a glance of the eye may be sufficient to determine it, or it may require an elaborate investigation by means of the interferometer, according to the detail and accuracy required. To record the position of a line within one Angstrom unit is ordinarily the work of a few minutes; but to measure a wave-length with an accuracy of 0.01 Angstrom unit demands skill of the highest order and plenty of time. In general, however, it may be said that in the ultra-violet region, say from 7.4,000 to .1,000 photography is practically the only available means for obtaining a record. In the visible portion, say from 7.8,000 to 4,000 the eye and the photographic plate are each available. By photography Abney has recorded wave lengths as great as 27,000 Angstrom units.
In the infra-red region, that is for wave lengths greater than 8,000, the bolometer, the radiomicrometer, the radiometer and the linear thermopile have all been used to good purpose. Rubens employing a linear thermopile has pushed his measures to 7.61,000. But photog raphy is probably more used than all other methods put together.
5. Comparison and Interpretation of Radi ations.— Let us suppose that the lines in the spectrum of a certain body have been analyzed and recorded, it may be only for the moment on the retina of the eye, it may be permanently upon a photographic plate. The next step is to i apply this information to the purpose for which it was obtained. To this end, the observer gen erally makes one of the five following com parisons: (a) A comparison of the spectrum in question with itself when the radiation is produced under different conditions. Thus by comparing the ordinary spectrum of the sodium flame with that of the same flame in a strong magnetic field the Zeeman effect was discovered. By comparing spectra of the same star taken at short intervals of time Pickering discovered a new class of double stars called " spectro scopic binaries." (5) A comparison of the spectrum in question with the spectra of other bodies. Thus a spectrum of iron, placed alongside of solar spectrum of the same scale, convinces one that there are many lines in the two spectra which coincide in position; and that, therefore, iron is probably one con stituent of the sun.
(c) A comparison of the spectrum in question with some spectrum predicted by theory. Thus Keeler established Maxwell's view of the constitution of Saturn's rings by comparing the spectrum of these rings with the spectrum predicted by Doppler's principle.
(d) A comparison of the spectrum in question with the same spectrum recorded in other ways. Thus Lewis com pares the infra-red spectra of sodium as obtained from the photograph. the phosphorograph. the bolometer, the radio micrometer, with that predicted by Kayser and Runge's formula.
(e) A comparison of the spectrum in question with • standard scale of wave-lengths, that is, with an ideal spectrum in which each line differs from its nearest neighbor by ex actly one Angstrom unit. The most beautiful examples of this comparison are Rowbe found in the maps of the solar spectrum prepared byland and by Higgs.
Concerning the interpretation of spectra, it must be frankly admitted that this is the most difficult part of the entire subject, demanding as it does wide experience in the laboratory and Judgment of the highest order. Apparently no theory is so fanciful but it may find some sup port among the varied and complex phenomena presented by the spectroscope. The science of optics is controlled by a well-established theory, while spectroscopy must still be classed as an almost purely empirical science.
6. Summary of There are, however, some general principles which have been fairly well established. Following are the more important ones: Spectra of Gases.— The emission spectrum of a gas can be obtained, in general, only by passing an electric current through the gas. The emission spectrum of a gas is practically always a spectrum of bright lines. The emis sion spectra of solids and liquids are practically always continuous. As illustrating exceptions to this general law may be cited the fact that Paschen using the bolometer has obtained the characteristic radiations of carbonic acid gas heated by streaming through a hot platinum spiral; also the fact that the spectra of gases under very high pressures become nearly con tinuous.
Kirchhors Law.— Let us denote by H the amount of radiant energy, say heat or light, of any one wave-length which falls upon a body at temperature t° in one second. Let h denote the energy of the same wave-length which this body absorbs in one second when at the same temperature. The ratio h/H is what Kirchhoff calls the °absorption° of the body, and is de noted by A. A body which absorbs all the heat falling upon it is said to be °absolutely black*; its absorption is unity. Let us now denote by E the amount of energy of the same wave length which this same body would radiate in one second. This quantity is called the °emis sion° of the body. Now Kirchhoff, in 1859, proved that: (1) The ratio of the emission to the absorption of any body depends upon the temperature only; and (2) This ratio is numerically equal to the emission of an absolutely black body at the same temperature.