INCANDESCENT SAPORS AND GASES 541
nescence rather than to a high temperature. For the heating of the gas to 300° C. could only appreciably change the molecular velocity if the temperature $ were low, for example 50° C.
Although the results of the above calculation are in good agreement with the facts, neveretheless the considerations here presented do not completely cover the case. For on the one hand, according to Ebert,* the distance between two lines in the solar spectrum which can still be resolved is smaller than is consistent with Doppler's principle, and on the other hand, according to Lord Rayleigh,t the consideration of the rotation of the molecules would reduce the ability of the transmitted light to produce interference much more than the consideration of their motion of translation. To be sure the revolution of the molecules would have to be considered only in the case of molecules composed of more than one atom; hence the explanation given above of the great capacity for interference shown by the mercury lines would still stand.
7. Other Causes of the Broadening of the Spectral Lines. —The motion of the molecules is not the only cause of the broadening of the spectral lines. The change in the period of the ionic vibrations due to damping must set a limit to the ability to produce interference, and hence must broaden the spectral line,^ since the ability to produce interference and the homogeneity of the spectral lines are closely connected. When a stationary condition of emission has been reached the ions are continually set into vibration by the collisions of the molecules. The more frequently these collisions occur, the smaller becomes the ability of the emitted light to produce interference. Since now the number of collisions increases
*Sitz.-Ber. d. phys. med. Sex:. Erlangen, 1889. Wied. Beibl. 1889, p. 944. fPhil. Mag. (5) 34, p. 410, 1892.ference of path over which interference can bein the flame of a substance of little mass, small conducting power, large surface, and large emissive power. But according to Rubens (Wied. Ann. 69, p. 588, 1899) the Auer burner is probably chemically active for