1869.] On some Recent Spectroscopic Besearches. 215 



this particular part of the light again in the spectrum of the star's 

 light. 



Spectrum analysis enahles us to obtain the information and to 

 make the recognition which are necessary. When, by means of a 

 group of dark or bright lines, we learn the presence of a terrestrial 

 substance in the star, we possess the means of knowing with the 

 greatest exactness the original refrangibility of those lines ; and 

 further, these dark or bright lines are objects which can be easily 

 recognized, and can be compared dhectly with the hues produced 

 by the same substance on the earth. By this method an observer 

 can detect any minute shift from their original position which the 

 hues may have suffered from the motions of the star and the earth. 



For such an investigation it was necessary to choose a brilliant 

 star, in the spectrum of which there were strong and very distinct 

 hues, and which, by their coincidence as a group with the bright 

 line of some terrestrial substance, showed that they were produced 

 by the vapour of that substance in the star. Of all the stars which 

 the writer had compared with terrestrial elements, when working 

 with his distinguished friend. Dr. W. Allen Miller, Treas. K.S., 

 Sirius appeared to fulfil most nearly these conditions. 



The spectrum of Sirius (see woodcut) contains four very strong 

 lines; two of these were found to be coincident, as seen in a 

 spectroscope of two prisms, with bright lines of hydrogen ; a third 

 line is probably also to be regarded as a line of hydrogen. In 

 addition to these strong lines, the spectrum is furrowed from end 

 to end by numerous fine lines. Among these were found lines 

 indicating the presence of sodium, magnesium, and iron. Nearly 

 all stars of which the light is white or nearly so give a similar 

 spectrum. 



Spectrl'm of Sirius. 



Of all these lines, however, there is only one (the strong hne 

 due to hydrogen and coincident with Fraunhofer's F) which could 

 be seen with the necessary distinctness when the powerful battery 

 of prisms required for this dehcate investigation was employ al. 



It may be well here to give an answer to an objection which 

 may suggest itself to some readers. If the line in the star is not 

 absolutely coincident with the line of hydrogen, would not the 



Q 2 



