locations had propagated from several lines of damage into the surrounding regions during annealing, and the driving force for these dislocations was considered to be the associated stresses.However, although strains are present beneath the "visible" damage, i.e., the chips, cracks, and dislocations, little appears to be known about their nature or extent. The strains are elastic and arise from the stresses imposed by the "visible" damage, and also perhaps from the presence of point defects, e.g., vacancies. Such point defects could be present without being detected by the transmission electron microscope method.When such specimens are chemically polished with CP4, the transmission electron micrographs show that the attack commences preferentially at the lines of damage. After 1 sec, the damage is partly removed and grooves begin to form along the lines of damage. After 3 sec, no damage can be detected and the grooves are appreciably wider. Thereafter, the grooves continue to increase in width, but not apparently to change much in depth. The latter may also be deduced from the interference micrographs of Fig. 1. The sequence is illustrated diagrammatically in Fig. 6. The reason for the preferential attack is undoubtedly due to the strains associated with the lines of damage.The question still remains as to why the scratch-