PHYSICS: T. LYMAN 
369 
I have employed a vacuum spectroscope containing a concave diffrac- 
tion grating! arranged in such a manner that the Hght path from the 
source to the photographic plate is wholly in gas. Briefly, the apparatus 
may be described as a brass tube about 11 cm. in diameter and rather 
over a meter long. This tube is closed at each end by brass plates ground 
to fit suitable flanges. In my most recent experiments the Hght is 
generated electrically in a discharge tube of quartz provided with tung- 
sten electrodes. This discharge tube fits air tight on one of the two 
brass plates just mentioned; Hght from it having passed through a sHt, 
traverses the length of the apparatus and falls upon the diffraction grat- 
ing by which, having been analyzed into its component colors, it is 
brought to focus on a special photographic plate placed in close prox- 
imity to the sHt. 
As the discharge tube is in no way separated from the body of the 
spectroscope it is obviously necessary to choose for the experiment some 
gas which will not only 3deld radiations in the region under investigation 
but which will be transparent to these radiations. My earHer experi- 
ments were conducted with hydrogen, since it had shown the necessary 
characteristics in that part of the spectrum investigated by Schumann. 
With this gas at a pressure of 2 or 3 mm., and by employing a strong 
disruptive discharge, about eighteen months ago I was able to extend 
the spectrum to wave length 900. 
A tedious investigation having convinced me that nothing more was 
to be expected from the use of hydrogen, at least in this neighborhood, I 
turned my attention to heHum, being guided by some of my earlier ex~ 
periments which had proved that this gas possesses the necessary trans- 
parency. At the same time, I made some improvements in my appara- 
tus which, though they left its general form unchanged, resulted in 
making it far more air tight than ever before. The success of this im- 
provement may be judged by the fact that I have recently been able to 
leave the spectroscope for over fourteen hours at a pressure of about 
3 mm. without being able to detect any leak, either by a McLeod Guage 
reading to 0.007 mm. or by the appearance of impurities in the spec- 
trum of the gas content. I also took particular pains to purify the 
helium w^hich I employed. 
I have been rewarded for my trouble by a very considerable extension 
of the spectrum for with heHum free from nitrogen, at a pressure of 2 or 
3 mm., by the use of the disruptive discharge, and with an exposure of 
about ten minutes, I have repeatedly observed a number of new lines 
the most refrangible of which has a wave length of 600. All this has 
been accompHshed with a grating ruled on speculum and with photo- 
