524 
PHYSICS: DUANE AND PATTERSON Proc. N. A. S. 
other lines belonging to the second group, is produced by electrons falling 
into the second, L2, orbit. 
On the experimental and theoretical evidence we have reached the 
conclusion that the emission line /Ss belongs to the atomic mechanism 
that produces the lines in the Li group, and that 72 belongs to that pro- 
ducing the lines in the L2 group. On the*other hand our experiments 
prove conclusively that the wave-length of /55 is shorter than that of the 
Lai critical absorption, and that the wave-length of 72 is shorter than that 
of La2. A somewhat similar situation occurs in the spectra of ordinary 
light. Here we have resonance and ionization potentials. There is a 
difference, however, between the character of the absorption of X-rays 
and that of ordinary light. 
The best way of explaining these phenomena may be somewhat as fol- 
lows : Let us suppose that the critical absorption Lai corresponds to the 
transfer of an electron from the Li orbit to the periphery of the atom — 
to an orbit there where it can find a place to stick. This does not mean 
a transfer to an indefinitely great distance from the centre of the atom. 
There may be many orbits outside that at the periphery, even in addi- 
tion to those that belong to visible spectra. With the electron in the 
orbit at the periphery the atom is in a state such that by the transfer of an 
electron back to fill the vacancy in the Li orbit it can radiate any one of 
the emission lines 1, 0:2, ai, ^2, etc., but not /Js- If, now, the frequency 
of the incident X-ray beam is higher than that of the critical absorption, 
if the energy hv is greater than that required to lift the electron to the 
periphery of the atom, it may carry the electron to an orbit further out 
or even outside of the atom altogether. In this case the atom would 
be in a state such that the return of an electron from one of these outside 
orbits would emit X-radiation of higher frequency and shorter wave- 
length than that of the critical absorption. This would correspond to 
the emission line jSs. A similar explanation applies also to the emission 
line 72, and the L2 orbit. 
From this point of view we would expect the critical potential for /Ss 
to be slightly higher than that for the other lines belonging to the Li 
orbit, and the critical potential for 72 to be slightly higher than that for 
the lines associated with the L2 orbit. 
Further, on this theory, the critical absorption does not correspond to 
the limit of the group, or series. It becomes incorrect to speak of the 
"absorption limit," as these words are usually employed. 
An accurate, quantitative test for our point of view cannot be obtained 
without making special assumptions. According to our measurements, 
however, the difference in frequency between the emission line and the 
critical absorption for tungsten, each divided by the Rydberg constant, 
is 6.0 for ^h-Lai, and 5.6 for yT-La^. These are of the order of magnitude 
of frequencies associated with the peripheries of atoms. 
