482 
PHYSICS: DUANE AND STENSTROM Proc. N. A. S. 
1. Dr. R. A. Patterson cooperated with us in making the measurements 
of these wave-lengths. We used an X-ray tube with a molybdenum 
target in these experiments. 
It appears from the data that the wave-length of the 7 line and that of 
the critical absorption differ from each other by about V2 of 1%- In 
the K series spectrum of rhodium (see Physical Review, Oct., 1919, p. 
369) the two wave-lengths differ from each other by about V4%- If 
we adopt the theory of electrons revolving in atomic orbits we can ex- 
plain this difference in wave-length as follows: According to the theory 
the critical absorption frequency equals the amount of energy required 
to carry an electron from the K orbit out to the periphery of the atom 
divided by Planck's constant, h. The y line is due to electrons falling 
into the vacancies in the K orbit from orbits outside of those that we may 
class together and call the M orbits. The amount of energy set free 
and radiated during one of these transfers of an electron is less than the 
amount of energy required to lift the electron all the way from the K 
orbit to the periphery of the atom. Since the frequency of the X-ray 
emitted during the transfer equals this energy divided by h, it follows 
that the frequency must be less than that of the critical absorption, and, 
therefore, its wave-length greater than that of the critical absorption. 
The 7 line is probably complex, and the wave-length we measure must 
be a kind of centre of gravity of the wave-lengths of its components. 
In the second order spectrum (fig. 1) a drop occurs in the curve at 
a certain point marked "a." This does not belong to the K series of 
tungsten. It represents the critical ionization of iodine in the first order 
spectrum, the wave-length of which is 0.3737 X 10~^ cm. We used 
methyliodide in the ionization chamber of the spectrometer. 
In a research described in the Physical Review for July, 1919, p. 67, 
Mr. Shimizu and one of us obtained experimental evidence showing 
that the difference between the K critical absorption frequency and any 
one of the L critical absorption frequencies equals the frequency of one 
of the emission lines in the K series. Theoretically this relation ought 
not to be exact, if orbits in some of the atoms are elliptical and in other 
atoms of the same chemical element, circular. The atoms with elliptical 
orbits contain amounts of energy that are different from the amounts of 
energy in atoms with circular orbits, and hence the change in energy 
when an electron is removed from the K orbit will not be the same in 
the two cases. ^ This means that the K critical absorption frequency 
should be complex, and that what we measure is a kind of average fre- 
quency. Strictly, according to the theory, the frequency difference 
law should apply to those atoms only that are exactly alike. 
To test this point we have taken the absorption measurements made on 
the ly series of tungsten by Dr. R. A. Patterson and one of us. The law 
applies, of course, to the wave-numbers as well as to the frequencies. 
