[Illustration: Fig. 6.] Fig. 6 represents the result of this experiment: the ray of sunlight or electric light, _r_, passes through a quantity of sodium vapour, and is then decomposed by the prism; the spectrum produced is marked by the absence of light (or by a dark line) where the yellow line, Y, was before noticed. These are the fundamental facts of spectroscopic analysis: sunlight is decomposable into a band of many colours, that is, into a spectrum; light emitted by a glowing vapour is characterized by the presence of coloured lines, each of which occupies a definite position with reference to the various parts of the solar spectrum; sunlight--or the electric light--when allowed to pass through a mass of vapour, furnishes a spectrum characterized by the absence of those bright lines, the presence of which marked the spectrum of the light obtained by strongly heating the vapour through which the sunlight has passed. The spectrum obtained by decomposing the light emitted by glowing vapour of potassium is characterized by the presence of certain lines--call them A and B lines. We are asked what element (or elements) is present in a certain gas presented to us: we pass a beam of white light through this gas and then through a prism, and we obtain a continuous spectrum (_i.e._ a spectrum of many colours like the solar spectrum) with two dark lines in the same positions as those occupied by the lines A and B. We therefore conclude that the gas in question contains vapour of potassium. The solar spectrum, when carefully examined, is found to be crossed by a very large number of fine black lines; the exact positions of many hundreds of these lines have been carefully determined, and, in most cases, they are found to correspond to the positions of various bright lines noticed in the spectra of the lights emitted by hot vapours of various elementary bodies. Assume that the sun consists, broadly speaking, of an intensely hot and luminous central mass, formed to a large extent of the elementary substances which build up this earth, and that this central mass is surrounded by a cooler (but yet very hot) gaseous envelope of the same elements,--and we have a tolerably satisfactory explanation of the principal phenomena revealed by the spectroscopic study of the sun's light. On this assumption the central mass of glowing iron, chromium, magnesium, nickel, cobalt, hydrogen, etc., is sending out light; a portion of the light emitted by the glowing iron is quenched as it passes through a cloud of cooler iron vapour outside the central mass, a portion of the light emitted by the glowing chromium is quenched as it passes through a cloud of cooler chromium vapour, and so on; the black lines in the spectrum are the records of these various quenchings of this and that light.