Vol.. 8, 1922 
PHYSICS: L. W. MCKEEHAN 
271 
evidence indicating that the element consists of a single atomic 
species.^ 
Through the kindness of one of my colleagues, Dr. H. E. Ives, a small 
supply of metallic beryllium, recently prepared for him in this laboratory 
by Mr. G. O. Smith, was made available to me for this study. The method 
of preparation was by the electrolysis of 2BeF2.NaF with from 7 to 15 
amperes at 80 volts. This is a modification of the method of Fichter and 
Jablczynski^ who used a lower voltage. The product of the process is a 
fine gray powder interspersed with thin flakes up to half a millimeter in 
greatest dimensions, which show good metallic lustre. The principal im- 
purity to be expected is beryllium oxide (BeO) , and this should not exceed 
in amount a few per cent by weight. 
A thin-walled pyrex glass tube about one millimeter in diameter was 
packed loosely with the beryllium powder, the looseness of packing being 
intended to reduce the chance of anisotropy.^ This sample was not thick 
enough to cast a very sharply defined shadow on the film, a circumstance 
which somewhat reduces the accuracy of the results.^ A water-cooled 
molybdenum target, bombarded by about 30 milliamperes of electrons at 
about 30,000 volts, and provided with ZxO^ filters^ gave an intense source 
of X-rays, the photographically important wave-lengths being^ 0.71212 
X 10-8 cm. (K as) and 0.70783 X 10-^ cm. (K on). The radius of the 
photographic film was 20.36 cm. The time of exposure was about 18 
hours. In taking the first photograph the cylindrical sample was kept 
stationary; in taking the second photograph the sample was rotated 
about its axis through a small angle at intervals of one hour. The two 
photographs were quite similar in general appearance, but the second had 
smoother lines in its pattern, so that it alone was used in the final measure- 
ments and calculations. 
In order to eliminate from consideration any lines certainly due to the 
presence of BeO a sample of that substance in the form of a fine powder was 
examined in the same way. The results are presented in Table I, and show 
that the diffracting centers lie at the points of two symmetrically inter- 
penetrating hexagonal space lattices, each with the axial ratio c/a = 1.63 
and with a = 2.696 X lO'^cm. There are three faint lines which must be 
attributed to the presence of impurities. Assuming that one molecule, BeO, 
is associated with each lattice point and taking8A^*= 6.0594 X 10^^, 
w = 9.018, this gives a computed density 2.99 gm./cm^. Both structure 
and density are in agreement with published values,^ although a recent 
paper by Gerlach^^ contains a very different result. A careful examina- 
tion of his data, however, shows that the resolving power of his apparatus 
was not high, and that all the lines which he reported as intense are within 
a short distance of positions accountable for by the structure here pro- 
posed. 
