86 . KANSAS UNIVERSITY QUARTERLY. 



the 50 volt alternating current circuit. In this manner the tube 

 could be run for hours without appreciable heating and the steadi- 

 ness and brilliancy of the light were all that could be desired. It is 

 not known how high the vacuum was, as the tube when made was 

 merely exhausted to the point of maximum brilliancy. 



The light coming from the tube was subjected to prismatic analysis, 

 the spectrum falling upon a slit plate. A slight motion of the 

 prisms given by a screw permitted any desired radiation to be thrown 

 upon the slit, from which it passed on into the Interferometer. 



Visibility curves were then observed and the results derived by 

 means of the usual formulas.* The mean of several closely agree- 

 ing visibility curves taken by eacfi of the writers for the yellow 

 helium line is given very approximately by the equation 



/, 1. 01 4-. 2 cos — 7^ \ X'^ 



v= ^.-J^^ 2-36-^ 



The line is therefore a doublet, the distance between the compo- 

 nents being .351 Angstrom units and the "half breadth" "8" of 

 each line being .021 Angs. The ratio of the intensities is about 

 1:10. 



The lines at 6678, 5016, and 4471.7 were also analyzed, the re- 

 sult for the blue line (4471.7) being also approximate, however, on 

 account of its extreme faintness. 



The red and green lines appear to be single, of half width .027 

 and .017 respectively: while the blue line is a doublet, the distance 

 between the components being .235 Angs. and the half width of 

 each line about .02 Angs. 



It will be seen that the result obtained for the distance between 

 the components of the yellow line differs less than i per cent, from 

 the mean of the available measures upon the solar helium line. 



*See PLil. Mag. 34 or K. U. Quarterly VI, No. 3. p. 77. 



