We measured the surface state dispersions on Be(<10(1)over bar 0>) by
angle-resolved photoemission (ARUPS) and calculated the electronic str
ucture for both possible terminations of the bulk crystal from first p
rinciples. The experimental data agrees only viith the prediction for
the termination with the shorter first interlayer spacing. This is als
o supported by a comparison between our calculation of the surface cor
e-level shift spectrum with the experimental data from Johansson et al
. However, in contrast to their interpretation, our calculation predic
ts a larger surface core level shift for second-layer Be atoms than fo
r first-layer atoms.