218 On some Recent Spectrosco2)ic Researches. [April, 



terrestial light relatively to the mirror was ensured by making it 

 pass through a small hole in a plate of ebonite. This plan pos- 

 sessed two other important advantages. The two identical spectra 

 of comparison were seen, one above the spectrum of the star, the 

 other below it ; and further, as the rays from the terrestrial light 

 were divergent, they encroached a little upon the pencils from the 

 object-glass at the slit, and caused the lines in the two spectra of 

 comparison to overlaj) for a little distance the spectrum of the star. 

 This state of things greatly assisted the eye in forming a judgment 

 as to the absolute coincidence or not of a line in a star with that of 

 a substance compared with it. 



The comparisons, even when effected in this way, were not 

 trusted to alone, but were checked by independent observations 

 made when the light of comparison was placed before the object- 

 glass of the telescope. 



To return to the observations of Sirius. The only line in the 

 star that could be successfully observed is the strong line at the 

 position of Fraunhofer's F, which is due to hydrogen. Now this 

 line of hydrogen is subject to great modifications under different 

 conditions of density. Hydrogen at the atmospheric pressure 

 gives a broad band of diffused light ; it was, therefore, necessary 

 to employ the hydrogen at a small tension. The light of rarefied 

 hydrogen is resolved by the spectroscope into three bright fines. It 

 is with the line at the blue extremity of the green that the com- 

 parisons were made. 



In Fig. 4 of the Plate these lines are represented. 



As the result of numerous observations on many nights, it was 

 found that the narrow line of rarefied hydrogen did not coincide 

 precisely with the dark fine of Sirius, but appeared to be a little 

 more refrangible. 



After the excessive care that had been taken to ehminate every 

 conceivable source of error, it is believed that the want of coinci- 

 dence of the line in the star with that of hydrogen, may be accepted 

 as a shift produced by the star's motion. The amount of the 

 displacement represents a velocity of separation between the star 

 and the earth of about 41 miles per second. 



The writer then obtained evidence from experiment that the 

 want of coincidence of the narrow bright line of rarefied hydrogen 

 with the centre of the band in Shius was not due to an unsymme- 

 trical expansion of the line as the tension of the gas is increased. 

 For this purpose a modified form of Sprengel's aspirator was con- 

 structed, and also a condensing apparatus, with which the spectra 

 of hydrogen and some other gases were examined under different 

 pressures. It is obvious that the hydrogen in Sirius is at a pres- 

 sure considerably less than tliat of our atmosphere at the surface of 

 the earth, but is more dense than the hydi'ogcn in the solar atmo- 



