292 
PHYSICS: R. A. MILLIKA N 
Proc. N. A. S. 
d. In the case of fluorine a strong line has been found through the 
use of sodium fluoride at 657.2 A. This is in about the position pre- 
dicted by the foregoing mode of approach for its L a line. Another strong 
fluorine line appears at 607.2 A. These are the only lines thus far definitely 
identifiable as coming from the 7 L- ing electrons of the fluorine atom. 
The longer of these wave-lengths is taken as the L a line of fluorine. Fluor- 
ine probably has, however, other lines of shorter wave-lengths but of 
such intensity that we have not yet been able to obtain them. 
e. The spectrum, due to the 3 L-ring electrons of boron (atomic number 
5) is especially interesting because of its simplicity. It contains less than 
10 strong lines all told. It begins on the short wave-length side at 676.8 
A and has only the following strong lines: 676.8, 760.0, 1624.4, the two 
doublets at 2164.2, 2166.2 and 2496.9, 2497.8 and the single spark-line 
near the visible at 3451.5. According to the foregoing convention, its 
L a line should be the doublet at 2497 A. Since the K a line of boron is 
at 67.2 A, the ratio of the K a to the L a frequency in boron is about 37. 
/. The spectrum due to the 2 L-ring electrons of beryllium begins on 
the short wave-length side, according to all the data available up to date, 
at 2175 A and reaches its maximum, its L a doublet, at 3130.6 and 3131.2. 
The entire absence upon our plates of any lines whatever due to beryllium 
between 230 A and 2100 A is a conspicuous illustration of the wide gaps 
in spectra obtained with ordinary gratings in or near the visible region. 
g. Similiarly the present experiments with lithium (atomic number 3) 
reveal no lines whatever between the shortest wave-lengths measurable 
on the plates used in the case of lithium and the familiar series due to its 
single L-ring electron whose L a line is at 6708 A and whose convergence 
wave-length is at 2299 A. 
5. The graph of the L a lines of the elements from lithium up is shown 
in the accompanying figure, along with the corresponding graph for the 
K a lines from helium up taken from the work of others. The only point 
on the L a graph which has not been directly observed is that corresponding to 
neon which has been inserted from the resonance potential of that gas. 
It is to be observed that the method here employed gives, as I think, 
the characteristic spectra of the atoms of each element, not of the mole- 
cules. If it were possible to work with the atomic gas of each element, 
the L a lines given herewith would be the resonance potentials of these 
atoms. 
The progression thus revealed in these optical spectra is exceedingly interest- 
ing and simple, and very like that exhibited by X-ray spectra. The reason 
it has not been observed before is clearly because hitherto only the upper 
ends of these optical spectra have been observable, so that the unfolding 
of simple relationships between spectra and atomic number had to await 
the development of an ultra-violet technique. 
