THE SUN ITS CHEMICAL ANALYSIS. 179 



the solar light. Any flame may serve for this purpose. The spec- 

 trum of all sorts of artificial flames have been examined, as well as 

 that of the electric spark, of the luminous arch produced by a current 

 between two points of cliarcoal, even that of the more brilliant of the 

 stars. Observers have it in their power to vary indefinitely the nature 

 of artificial flames; nothing is easier than to place different substances 

 in suspension in a flame, either directly by immersing them at the end of 

 a wire of platina, or mixing them beforehand with the liquid which 

 supports the combustion. The study of artificial flames has revealed 

 a phenomenon at least as extraordinary as that of the dark lines of 

 the spectrum. When certain substances are in a state of ignition in 

 the flame, the spectrum is traversed by colored bands of a greater 

 brightness which are vividly depicted on the general ground of the 

 usual colors. This circumstance had not escaped Frauenhofer, who 

 had seen with astonishment bright lines traced in the spectrum pro- 

 duced by the flame of a wax candle. Different ph3^sicists, Brewster, 

 Miller, Schwann, submitted to analysis flames obtained by burning al- 

 cohol which contained various salts in solution, and they were thus 

 enabled to observe the bright stripes of artificial spectrums with 

 great distinctness. Thus it was ascertained that every flame con- 

 taining sodium furnishes a spectrum in which a yellow line of an ex- 

 traordinary brightness is delineated. Schwann even observed that it 

 was sufficient to mix a very small quantity of sea salt, or chloride of 

 sodium, with the combustible liquid, in order that the spectrum should 

 disclose its presence by the appearance of the yellow line. 



In this discovery there was the germ of a new method of chemical 

 analysis. Each metal in effect has bright lines of special colors, 

 which correspond with it, and which retain an invariable position 

 in the spectrum. A chemist may learn to distinguish these lines as 

 readily as he recognizes the precipitates obtained in the laboratory by 

 the ordinary reagents; but with the advantage that the light fur- 

 nishes a reaction far more delicate and perfect than any with which 

 chemistry was before acquainted. Arago, who accomplished so much 

 in behalf of optics, predicted that chemistry would some day derive 

 unexpected services from the progress of that science. A ray of 

 light proceeding from aflame discovers by its physical properties the in- 

 timate nature of the focus from which it emanates. 1 shall cite an 

 example, borrowed from M. Kirchofif, well calculated to excite sur- 

 prise : "The following experiment," writes the learned pliysicist, 

 ' proves that thus far chemistry has no reagents which can be placed 

 even in remote comparison with that of the spectrum as regards its 

 sensibility. We have caused three milligrams of chlorate of soda to 

 detonate in a part of the hall as far removed as possible from the ap- 

 paratus, while we observed the spectrum of the flame of a g>s-lamp 

 giving but slight illumination; the apartment in which the experi- 

 ment was made measures about sixty cubic metres. After some 

 minutes the flame, reflected in a tawny yellow hue, presented with 

 great intensity the characteristic line of sodium, and this line was not 

 completely eflaced till after the lapse of ten minutes. From the capa- 

 city of the hall and the weight of the salt employed in the experi- 



