0 *5 analyser angle FIG. 3. The emission current plotted as a function of rotation of the analyzer. The scales are proportional to intensity and angle but have not been calibrated. (a) Copper (111); (b) tungsten <110>.and center of the dark area as shown in Fig. 2. (The projected diameter of the probe hole, in this case, was 30 A.) These five plots are the averages of four separate runs in which a smooth curve was drawn through any variations which could be attributed to random noise. However, no significance is attached to the structure in the curves. The emission current for the 10° and 13° plots was adjusted to 10 4 counts/sec while for the other three it was maintained at 300 counts/sec. This latter figure represented the maximum total current it was possible to draw without danger of rupturing the tip. The associated large shot noise was augmented by thermal migration of the prominent surface atoms since the tip was held at room temperature. The random noise is shown in Fig. 2 for the two count rates used. Figure 3 shows the effect of varying the alignment of the analyzer. The scales have not been calibrated in angle or intensity. In the case of the copper (111) direction, as the analyzer was rotated away from the straight-through direction the threshold in the energy distribution plots approached the Fermi energy.These results show that the upper limit of the conduction band in the (111) direction of copper is 0.4±0.05 eV below the Fermi level. This is in good agreement with the calculations of Burdick. 5 It also appears that an insignificant amount of diffuse electron scattering was caused by the thermally migrating surface, and that the band structure of the bulk copper persists to the surface monolayer. It is further concluded that the alignment of the analyzer is of great importance to the observation of these effects. 1 The influence of an electric field (~10 4 V rms /cm) on the luminescence of KBr:Tl for excitation in the A band has been investigated. The measured spectrum points out beyond any doubt that there are two different centers responsible for the emission.The emission spectra of thallium-doped alkalihalide phosphors have been investigated at different temperatures by several authors. 1 " 4 In general, emission bands can be ascribed to the optical transitions between localized electronic states and their shapes will be due to the interaction with the ions neighboring the impurities. The theoretical analysis of these bands becomes extremely difficult when electronic degenerate states are involved because the electron-lattice interaction will yield a coupling between different levels. This is just the case of the alkalihalide crystals activated by thallium. Particularly we have investigated the luminescence of KBr:Tl at 80 and 300°K on the entire A-band excitation as shown in Fig. 1.At 300°K the emission consists of two bands centered at 3.99 and 3.54 eV, while at 80°K the bands are shifted to 4.06 and 3.46 eV, respectively. In the latter case both bands show approximately Gaussian shape. At 30...