1880.] 



On the Spectra of Magnesium and Lithium. 



99 



at 55107, 5527-5, the b group, 4703'5, 4570-5, 4351-2, 4166, 4057'3, 

 and the triplet beyond H. 



On the Spectrum of Lithium. 



In the spectrum of lithium we have not heretofore been able to 

 obtain the wave-lengfch of the violet line (at about 4130, Boisbaudran) 

 by reason of its very diffuse character, but we have recently found 

 that by a slight modification of our apparatus we could obtain this 

 line more sharply defined and even reversed. This modification con- 

 sists in making the opening in the lime small, 5 or 6 millims. in 

 diameter, using small carbons as electrodes, and the current from a 

 De Meritens machine. The line can then be seen close to but on the 

 less refrangible side of the barium line in that region. We make its 

 wave-length 4131 - 7. 



Besides this line we find that lithium gives another line in the arc 

 at 4273. This line comes out conspicuously nearly midway between 

 one of the brightest chromium lines and an iron line when a fragment 

 of lithium is dropped into the crucible. The same line may also be 

 detected in the spark taken between points of fused lithium carbonate ; 

 and as a sample of the carbonate very carefully purified shows this 

 line, we have no doubt that it really belongs to lithium. 



Both these lines are remarkably strengthened when a fragment of 

 magnesium is dropped into the crucible. 



On taking a photograph of the arc when a fragment of lithium is 

 introduced, two more lines come out on either side of H and K. These 

 are generally diffuse, and barely discernible by the eye in the arc when 

 glass prisms are used, but easily seen with a calcite prism and quartz 

 lenses. In the powerful arc of a Siemens machine they are enormously 

 expanded and reversed. From measurements of the photographs the 

 wave-lengths appear to be 3913 and 3984. A yet more refrangible 

 diffuse line appears in the photographs at about 3799. 



The lithium lines form a progression not unlike that of the sodium 

 lines, the alternate lines at 4603, 4131*5, 3913 are much more diffuse 

 than the lines at 4972, 4273, and 3984, and approach nearer together 

 as they become more refrangible. There is probably another narrow 

 line at about 3838, amongst the magnesium triplet, which we have not 

 yet made out with certainty, and we may expect to find fainter lines 

 still more refrangible, but as our photographs have been taken with a 

 train of glass prisms they do not show out. 



There is a line in the " normal solar spectrum " corresponding to 

 the line at 4273. 



While the ratios of the wave-lengths of the red, orange, and green 

 Hues of lithium are nearly : ' 2V5 those of tbe green and blue 

 lines and the two violet lines first above mentioned are nearly 

 1 . . _i_ . 1 



25 ' 27 • 29 • 30' 



H 2 



