SPECTRUM ANALYSIS. As is explained in the article on SPECTROSCOPY, the spectrum of a definite substance under definite conditions is always the same. The spectrum of the radia tions emitted by iron vapor, for instance, when rendered luminous, is characterized by certain definite trains of waves, or, as ordinarily ex pressed, by certain 'lines.' the conditions under which the spectra arc produced are varied, for instance, if for one ease the spark was used, and in another the arc, there are differences in the spectra, consisting mainly in variations in the intensity of the lines. Further, if other sub stances are mixed with the iron, or if the iron exists as an impurity in some other substance, or if the pressure of the surronnding atmosphere is varied, there are corresponding alterations in the iron spectrum. But under definite conditions there are definite lines which are characteristic of iron. Extremely small traces of a substance
may be thus made evident by the presence of its spectrum; and in a mixture of many substances the presence of the various parts may often be ascertained by a study of the spectrum emitted. This constitutes the science of spectrum analysis and has proved useful to the chemist in many cases. In fact the discovery of several elements, viz. emsium, rubidium, and gallimn, was due to the detection in certain spectra of lines which could not be ascribed to any known substances. The application of spectrum analysis to the study of the spectra of the sun, stars, and other heaven ly bodies has proved most important, and some of the results of various observations and investiga tions are given in the article on SPECTROSCOPY (q.v.).