SPECTROSCOPE AND STUFF OF COSMOS 347 



showed only a series of bright lines, and not the whole rain- 

 bow spectrum, like the sunlight. Moreover, these bright bands 

 seem to vary with the different substances employed. Finally, 

 solid substances, rendered incandescent, gave varying dark-line 

 spectrums. It seemed as if each substance produced its own 

 characteristic apparition. 



Evidently here was a capital discovery. The light of the 

 spectrum comes from a distance. That distance may be half 

 a foot, a few feet, a few miles, it may be millions of miles. So 

 long as the light passes through no absorbing medium the 

 spectrum remains the same. What if this were true even of 

 the light of the sun ! It was first needful, of course, to sift 

 out all the disturbances due to purely earthly conditions, to 

 atmospheric absorption, and the like. If this could be done, 

 then it might be possible to compare the spectrum of the sun 

 with the spectrum of the various elements, and so determine 

 the materials which make up that glowing disk of light. 



Such a possibility floated before more than one mind, notably 

 Sir John Herschel's, but the solution came from a German 

 physicist, Kirchhoff, working in conjunction with the celebrated 

 Bunsen. One day, passing a ray of sunlight through vaporised 

 sodium, one of the constituents of common salt, he noted that 

 the characteristic bright lines of the spectrum of sodium had 

 disappeared. He saw in a trice that he had made an observa- 

 tion of profound import. Following it up, he was able to show 

 that when a gas grows cold it will absorb the same rays of light 

 which it emits when it is incandescent. 



The dark lines of the solar spectrum mean, therefore, that in 

 the sun are many of the earthly elements sending us their charac- 

 teristic light, but sending it through an atmosphere of cooler gas, 

 which absorbs these rays, and produces the dark lines we see. 



This was the essence of Kirchhoffs discovery. The broad 

 facts were simple enough ; the details were infinitely puzzling. 

 Of a truth, spectrum analysis, as this study has come to be 

 called, has developed a wondrous and ofttimes precarious 

 delicacy. Sometimes it has seemed as if the apparent contra- 

 dictions, the inexplicable anomalies that have appeared as the 

 study developed, were likely quite to vitiate any certainty of 

 inference. Patience and persistence, however, have told, and 

 the typical German textbook on the subject now comprises 

 three huge volumes. It is to-day a special science. 



