THE SPECTRUM OF HYDROGEN IN THE REGION OF EXTREMELY 
SHORT WAVE-LENGTH. 
INTEODUCTIOK 
In a preliminary paper 1 the author has given the wave-length of more than one 
idred and thirty lines in the region of the spectrum lying between the values 1850 
and 1030 tenth metres 
It is the object of the present paper to compare the results 
obtained by the author with those given by Schumann; to describe the apparatus 
used in this research and to call attention to some new facts which have come to light 
since the publication of the first notice. The description has been made with some 
minuteness in the hope that an exact knowledge of the conditions necessary to success 
may prove of value to investigators who work in this field. Some attention has also 
been given to earlier and imperfect forms of the apparatus. For the author wishes, 
by flagging the pits into which he has fallen, to prevent other investigators from 
similar accidents. 
The improvement over the method of Schumann which characterizes this research 
consists in the introduction of a concave diffraction grating in place of fluorite prism 
and lenses, thus permitting the measurement of wave-lengths. The object of continu- 
ing the work has been to improve the accuracy of the measurements and to eliminate 
from the radiation obtained from a hydrogen tube those frequencies which were due 
to impurities. 
As it is unsafe to rely upon a process of extrapolation even with a grating 
ipectrum, the two-slit method described 
prev 
paper 
2 
ployed 
The 
pectroscope has been altered in construction to permit of all the adjustments 
quired for this method and finally the photog 
plate itself has been bent 
to agree in curvature with the 
arc of the circle on which the spectrum is in 
theory formed. Very considerable increase in accuracy has thus been gained 
grating with which the work has been done possesses one extremely stron 
1 Astrophysical Journal, Vol. XIX, No. 4, 1904. 
9 Physical Review, Vol. XVI, No. 5, 1903. 
g 
The 
first 
