By means of these two general principles Kirchhoff was enabled to explain the fact that a sodium flame placed between one carbon pole of an electric arc and the slit of the spectroscope will produce two dark absorption lines in the orange of the continuous spectrum exactly where, without the arc, it would produce two bright lines. But what is more important, Kirchhoff succeeded in explaining, on these same principles, the dark lines which Fraunhofer mapped in the spectrum of the sun. The con tinuous part of the solar spectrum is supposed to be due to the bright, underlying layers of the sun, which are under great pressure; while the dark lines the so-called Fraunhofer lines are due to absorption of the cooler gases which surround the sun. In the same manner may be explained the dark lines which appear in several classes of stellar spectra.
Doppler's Principle.If we denote by v the speed with which any radiant source is ap proaching the observer, by V the speed of light, and n the number of waves of length ? emitted per second by the source, then it is evident that the n vibrations which the source emits in any one second will be distributed over a dis tance numerically equal to Vv. Hence the wave-length of the light, X, which reaches the observer will be V v = V ° ) It is evident, therefore, that the wave-lengths of light coming from a star which is approaching the observer will be shortened in the ratio V In like manner, if a star be receding V from the spectroscope the wave-lengths will be increased in the ratio V + This principle, V first enunciated by Doppler in 1843, enables the astrophysicist to measure rates of ap proach and recession. Recent work by Frost and Adams shows that the speed of a star in the line of sight can be determined with an error not exceeding one-half mile per second. In like manner the relative motion of the two limbs of the sun and of several of the planets, and hence their periods of rotation have been determined by the spectroscope. Even in the very faint spiral nebulas, the rotation and the nature of their motions in space can in some cases be found, thus throwing a flood of light on the nature of these bodies and on the pos sible development of stars from them.
Effect of Pressure. In 1895 it was proved by Humphreys and Mohler, from measurements made in the physical laboratory of Johns Hop kins University, that the wave-length of a line in the spectrum of any element is dependent upon the pressure of the medium surrounding the source. Thus when the cadmium arc is worked under a pressure of 10 atmospheres the cad mium lines are shifted toward the red end of the spectrum about 0.07 of an Angstrom unit.
In general, the amount of this shift is toward the red, i directly proportional to the wave lengths n any one element, and directly pro portional to the excess of pressure above one atmosphere.
of Atmosphere It has been shown by Crew, Basquin, Porter and Hartman that the spectrum of an electric arc between two metallic electrodes is very distinctly affected by surrounding it with an atmosphere other than air. In the case of hydrogen, it has been shown that the lines affected by the hydrogen atmosphere belong to the spark spectrum of the metal. It is also true that the lines which belong to Kayser and Runge's series are not affected by the change from air to hydrogen. It has been suggested that this effect may account for the peculiar spectra of many of the so-called "hydrogen stars;' that is, stars in which the hydrogen spectrum is very strong.
The Zeeman Effect. In 1896 Zeeman, then at Leyden, made the capital discovery that if a source of radiation be placed in a magnetic field each individual line in its spectrum will in general become a triplet with the two side components circularly polarized and the mid dle component plane polarized.
Temperature Effects. The role which tem perature plays in the production of line spectra is very little understood. It is, at present, im possible to say to what extent the various char acteristics of flame, arc or spark, spectra depend upon temperature, upon chemical action and upon electrical conditions. But in the case of solid bodies, which yield a continuous spectrum, Stefan showed, in 1879, that the total radiation varies as the fourth power of the absolute tem perature. And it has since been established by Wien that, if we denote absolute temperatures by T, 4nT = constant and = constant where indicates the wave-1 at which the radiation is a maximum and ,n represents the value of the maximum radiation. These two important laws are merely inferences from a still more general expression which Wien has established.
Law of Spectral Series. Soon after the dis covery was made that there exist in the spectra of the elements certain series of lines distin guished by certain common characteristics, Bal mer succeeded in devising a single formula which gives, in a very exact manner, the wave length of each of the hydrogen lines known to him at that time. The expression,is as follows: X= where is a constant and m denotes the nat ural numbers beginning with 3. This for mula has since shown Itself competent to de scribe accurately the 29 hydrogen lines which are now known.