Abstract
A recently published assessment of the reproducibility of experimental LEED intensity data from the tungsten (100)(1*1) surface by Stevens and Russell (1981) is evaluated. I-V curves are presented which suggest that the independently obtained LEED spectra which they compared, including their own, were not at precisely normal incidence, this being the major cause of the differences among the curves. The importance of a valid criterion for establishing normal incidence is emphasised, especially when studying subtle changes in surface-structure symmetry. This is demonstrated by data which suggest the W(100) surface above 370K shows a very subtle loss of mirror-plane symmetry in the direction parallel to the steps existing on the crystal surface studied. Just as below 370K the observed 2 mm point group symmetry of the ( square root 2* square root 2)R45 degrees structure can be attributed to a preferred orientation of p2mg domains, this apparent 1 m point group symmetry of the (1*1) structure is suggested to be a direct step-induced preference for one p1m1 orientation rather than a consequence of deviations from normal incidence.