1893.] and its Relation to those of Indium and Gallium, 371 



In all my observations on the spectrum of thallium I found that 

 the red line was slightly more refrangible than the hydrogen line. 

 The difference, however, was so small from the nebulous character of 

 hydrogen C at atmospheric pressure, that, with the spark spectrum 

 in moist air, the thallium line overlapped C so much as to be undis- 

 tinguishable ; but when the intensity of the hydrogen line was re- 

 duced, by passing the spark through partially dried air, the two lines 

 could be distinctly recognised by the difference in their bright- 

 ness. 



A closer comparison of the red lines of thallium and hydrogen was 

 now made by placing the narrow part of a hydrogen vacuum tube 

 behind the phial containing the thallium electrodes. 



The sparks through the vacuum tube and electrodes were taken 

 simultaneously from separate induction coils. The red lines were 

 now distinctly separated by a dark space of less width than the in- 

 terval between the double sodium line J) l D % . The spectrum of C was 

 less bright than the thallium line, owing to the greater distance of the 

 vacuum tube from the slit of the spectroscope ; but a more distinct 

 line was obtained by. substituting for the hydrogen tube one of carb- 

 onic acid, in which the C line appeared by the electrolysation of the 

 residual aqueous vapour contained in the gas. 



The wave-length of the thallium red line, as estimated by the dif- 

 ference of refrangibility of the hydrogen line, from the same electrodes 

 in moist air, is 6560 ; or, 6558, as estimated from the distinct line of 

 hydrogen in a vacuum tube. 



A comparison of the arc spectrum of thallium may now be made 



with the similar arc spectra of its analogues, indium and gallium. I 

 have already pointed out that the characteristic lines of the alkaline 

 metals and their homologues of position in the thallium series 



2 d 2 



