We show unusual dependence of the total electron yield on the illumina
ted surface orientation of TiO2. This requires care in recording angul
ar dependencies and forces one to use always the same surface orientat
ion. This in fact led us to publish wrong results in the past which we
correct here.