A1 was anodized in 2.4M H2SO4 at 5 mA/cm 2 and electronmicroscopic studies made of breakdown and formation of pits beneath the film after various surface treatments or after additions of up to 0.05M Cl-to the acid and selection of those specimens which had remained in the film-growth stage before stable pits grew. An electron-opacity technique was developed for studying film growth. Even in the absence of Cl-, there was a continued incidence of breakdown and a substantial time-dependent direct-dissolution current. With CI-, cell voltage tended to be lower. Film formation efficiency and geometrical structure were correlated directly with average cell voltage and not directly with C1-concentration. Cl-was interpreted as having only an indirect effect on film growth by stabilizing concurrent pit growth, wl~ch reduced the filmforming current. Possible modes of pitting beneath the film are discussed.In a recent paper (1), the authors presented electronmicroscopic evidence that during galvanostatic anodizing of aluminum in sulfuric acid many small breakdown-repair (B-R) or breakdown-pitting-repair (B-P-R) events occurred at the base of the film concurrently with porous film growth and that these events were not generally associated with substrate lattice defects or impurity clusters. As part of a program concerned with the mechanism of pit initiation, this was followed in the present work by studies of: (i) the possible influence of the surface preparation on the incidence of breakdown and (ii) the effects of C1-additions to the anodizing bath. The previous results and those of Part (i) directed attention not to the production by C1-of breakdown, which occurred in its absence, but to its reduction of the effectiveness of repair. This might be due to an effect of C1-on film growth rate or structure. Conditions favorable for examining this question were obtained by adjusting the C1-concentration in the bath to the highest level at which film growth would not be immediately short-circuited by growth of optically visible stable pits. Observations had been made identifying the transition in behavior with increasing C1-concentration (2).The surface treatment before anodizing was of concern specifically because with the perchloric acidethanoi electropolishing bath employed a residual film was known to be present after rinsing and this film presumably contained C1, which has been detected by Auger analysis on surfaces polished in this bath (3). Such surfaces have been reported to yield films of low ionic resistance after forming (4).The processes occurring in this system were such that the microtopographic features most accessible to observation were not those of growing pits but those of repair zones in the film and of the pore structure at various stages of growth. Comparisons were effectively made by examining the film after anodizing times at which it contained evidence of the previous occurrence of a high density of small (~0.1 ~m) B-P-R or B-R "events," and the nature of the geometry deduced is illustrated schematically ...