The electronic excitation spectra at arbitrary momentum values can be deter
mined in principle by (e,2e) spectroscopy. These experiments are performed
using high energy electrons (1 keV up to 20 keV) meaning that refraction of
electrons at the surfaces, a complicating factor for low energy photoemiss
ion measurements, can be avoided. However at such high energies the geometr
y of the experiment has to be known very accurately. Here we describe a pro
cedure to check the actual geometry of the experiment, and how to adjust fo
r misalignments, if they are present. Spectral momentum densities obtained
in this way for graphite along low symmetry directions are compared with fu
ll potential linear muffin tin orbital calculations. The measured intensiti
es agree quite well with the calculations, however the observed dispersion
of the graphite sigma band exceeds the calculated one by 1.5-2 eV, and the
separation between s and p band is somewhat smaller in the calculation comp
ared to the experiment.