THE DISCOVERY OF THE ELEMENTS. 477 



minerals, as is the case, for example, with strontium and chromium, 

 and tungsten. Of course it is difficult to appraise the relative abun- 

 dance in Nature of different elements ; more especially from the cir- 

 cumstance of those -which are put to commercial uses being every- 

 where sought for, and those not put to commercial uses being habitually 

 neglected save indeed by the man of science, to whom the peculiar 

 properties of some of the less familiarly known elements, as palladium, 

 osmium, erbium, didymium, uranium, and thallium, render them ob- 

 jects of the highest interest. 



A very notable point with regard to the last-discovered four ele- 

 ments, namely, rubidium, ccesium, thallium, and indium, is their suc- 

 cessive discovery within a few years of each other, by one and the 

 same process, namely, that of spectrum analysis. This process, in- 

 vented and made available as a means of chemical research by Bunsen 

 and Kirchhoff in 1859, consists simply in allowing the light given off by 

 different ignited gases and vapors, limited by means of a fine slit, to 

 pass through a prism or succession of prisms ; and in observing the so- 

 produced, brightly-colored, widely-extended image of the slit. It has 

 been known from the days of Newton, that, by the passage of hetero- 

 geneous light through a prismatic, highly-dispersive medium, its differ- 

 ently refrangible constituents become widely separated from each 

 other, so as to furnish an elongated, colored spectrum. But, whereas 

 the spectra of incandescent solid and liquid bodies are continuous, and 

 not distinctive of the particular luminous bodies yielding them, the 

 spectra of incandescent, gaseous, or vaporized bodies, are found to be 

 discontinuous, and to consist of one or more bright lines of different 

 color, thickness, and position, according to the nature of the particular 

 incandescent gases or vapors from which the light through the slit is 

 proceeding. In this way it is found that the spectra of the different 

 chemical elements, alike when free and in combination, are perfectly 

 definite, and characteristic of the particular elements vaporized and 

 made incandescent. 1 And, in many cases, the spectra, or portions of 

 the spectra of particular elements, even when present in the most 

 minute proportion, are so extremely well marked and distinctive, that 

 the presence or absence of these elements is determinable with the 

 greatest ease and certainty, by a mere inspection of the emission spec- 

 tra yielded by the incandescent gases or vapors under examination. 

 Moreover, gases and vapors are further capable of affecting hetero- 

 geneous light which is passed through them ; and of thus yielding ab- 

 sorption spectra, in which the characteristic lines of the above-described 

 emission spectra are reversed, so as to appear, unaltered in position, as 

 black lines or intervals in an otherwise continuous band of color. 



Now, the salts of the alkali-metals, lithium, sodium, and potassium, 



1 For some qualifications of this statement, vide Eoscoe's " Spectrum Analysis." 



