144 LYMAN. — THE SPECTRUM OF IIYDIIOGEN. 
been limited by it. Besides this mechanical difficulty several other possible agents 
may exert an influence. Speculum metal may cease to reflect in the region near 
X 1000 ; that it reflects so well down to this point is surprising. The Schumann 
plates may cease to be sensitive. Small impurities in the hydrogen may exercise con- 
siderable absorption. Only experiments on metallic reflection, on the manufacture of 
plates and on the purification of gases can answer these questions. The author sees 
no insurmountable difficulty, however, to the still further extension of the spectrum. 
RESULTS. 
The results arrived at in this memoir may be set forth as follows : 
I. The spectrum has been extended from the limit obtained by Schumann to the 
value X 1030. 
II. The lines in the spectrum of hydrogen have been measured accurately from 
X 2000 to X 1228, and the values of the principal lines to X 1030 have been determined. 
III. The nature of the spectrum of air has been investigated. 
IV. The limit of transparency of certain specimens of white fluorite has been 
obtained. 
V. The effect of the disruptive discharge on the spectra of hydrogen and air lias 
been studied, and the absence of a secondary spectrum of hydrogen established in the 
O 
below X 2000 
VI Wave-lengths have been attached to the spectrographs obtained by Schu 
Much of this research has been carried on with the help of a grant from the Bache 
Fund. The permission to reproduce the plates from the Smithsonian Contributions is 
due to the kindness of the Secretary of that Institution. 
It is impossible to conclude this memoir without some tribute to the man whose 
name will be always associated with the region of short wave-lengths which he dis- 
covered, and it is with the greatest pi 
and inspiration he has received from the friendship of Dr. Victor Scl 
author acknowledges the help 
JEFFERSON PHYSICAL LABORATORY, 
Harvard University, Dec. 27, 1905. 
