Real crystals as a source of error

Specimen-dependent properties of real crystals which influence the accuracy of intensity measurement are size, shape, homogeneity, stability and environment. Variations of any of these will affect the corrections necessary to allow for absorption and extinction in a given experimental measurement; these in turn will affect the evaluation of the absolute intensity, and the two principal tests for internal and external consistency, namely variation within any form {hkl}, and variation of mean values I{hkl} between different specimens. They may also affect the relative values of I{hkl} within one data set, leading for example to spurious anharmonicity in the temperature factors derived for an ellipsoidal crystal. Examples of practical solutions of actual problems involve waxes instead of glues for crystal mounting, miniature films or intensifying screens for use in unstable situations or with very small crystals, integrated oscillation photographs for intensity measurements from poor specimens, rotation photographs for high symmetry crystals, the use of Laue photographs for the detection of order-disorder phenomena, and the incorporation of an iron-55 source in an automated diffractometer to provide an internal standard for intensity measurements.