270 
PHYSICS: L. W. MCKEEHAN 
Proc. N. a. S. 
before collision. No drop is seen at 5.4 volts which would correspond 
to the loss of 10.4 volts which is normally assumed to take place at 
ionization. 
From curves of such nature for different velocities of impact, it has 
been possible to determine more surely than heretofore the energy losses 
at electron impact. The outstanding results of the investigation are: 
1. The discovery by Mohler, Foote and Meggers of a resonance poten- 
tiali nvolving an energy loss of 6.7 volts was confirmed. The work from 
which these investigators were able to deduce the existence of this critical 
potential seemed subject to some very serious difficulties in interpretation. 
These difficulties have been cleared up. The 6.7 volt type of collision 
does not occur for electrons with energies less than 8.5 volts, but at volt- 
ages above the ionization point it is the most important type of resonating 
collision. 
2. Indications were also found that a collision involving an energy loss 
of about 5.7 volts also occurs in mercury. This is in agreement with the 
fact that absorption lines have been found by previous investigators in 
the mercury spectrum in a region which through the quantum relation 
corresponds to this potential. 
3. Little has been hitherto known about the nature of the ionizing 
collision. The present method has cast some light on this matter. The 
results obtained indicate that at such a collision the impinging electron 
loses all of its energy, and also that the electron which is the product of 
the ionization leaves the parent atom with negligible energy. This con- 
clusion is at variance with the usual conception of the ionization process. 
THE CRYSTAL STRUCTURE OF BERYLLIUM AND OF 
BERYLLIUM OXIDE 
By L. W. McKbehan 
Research Laboratories of the American Tei^Ephone and Telegraph Company 
AND the Western KivEctric Company, Incorporated 
Communicated July 17, 1922 
The structure of the lighter elements and of their compounds should, on 
account of their greater simplicity, be of especial value in deciding what 
role the so-called valence electrons play in connecting atoms. Beryllium 
(Be; atomic number 4) is the lightest metal the crystal structure of which 
has not hitherto been determined by X-ray methods. Its accepted atomic 
weight is w = 9.10, although a recent determination, ^ appai-ently of high 
accuracy, gives w = 9.018 which is in much better agreement with the 
