140 
LYMAN. 
TIIE SPECTRUM OF HYDROGEN 
METHODS OF MEASUREMENT. 
The methods used were two in number. The values of all the lines were first ob- 
tained by the two-slit method and these values were then checked by obtaining the 
stronger lines in the second spectrum and comparing their positions with known iron 
lines in the first spectrum. For this last purpose the first and second spectra obtained 
from the left hand slit were employed. 
The two-slit method has been described elsewhere, but a brief account of its theory 
d its limitations may not be out of place h 
If 
8 and #', be placed 
whose diameter is the grating's radius of curvature, the illumination of these 
slits by white light will give rise to the images / 
and /'. 
To each of these images a set of spectra 
will correspond. For the present purpose 
suf- 
Fig. 3. 
ficient to concentrate the attention on the two first 
spectra. It is evident that these two spectra will 
be shifted with respect to each other by an amount 
depending on the distance between the slits. If a 
photographic plate be placed between S and S', and 
if the height of these slits be properly adjusted, one 
of these spectra will be superposed upon the other. 
At a given point, P on the plate, the light brought 
to focus f 
S will be of a sho 
wave-length 
than that from S'. If the sources of light be so selected that the wave-lengths in both 
spectra arriving at P have known values, then the shift of one spectrum with respect 
may be determined by comparison of 
If the apparatus 
adjustment both spectra are in focus upon the same circle and the amount by wh 
one spectrum is shifted over the other is a constant quantity ; that is to say, if 
shift is determined by comparing known lines at one end of 
the same value at the other end. 
method depends. 
the plate, it must have 
It is upon this property that success in the use of 
It next becomes of importance to inquire 
influence the 
extent small 
tancy of the shift 
of adjust 
Here the nature of the method upon 
remembered 
the observer must rely in determining the perfection of this adjustment must be 
bered. I he onlv tipap.t.inn.l t ao f ~^ -;-*.. • ,i -, . . . , . • 
pectra 
nly practical test consists in the sharpness of focus of the 
tw 
It is the object then to so manage matters that both spectra shall be in per- 
fect focus throughout the plate's length and 
The vital questior 
