Analysis of total ozone data from Syowa (69 ¿S, 40 ¿E) for the years 1982 to 1994 reveals several new features in the behavior of ozone depletion. There is no significant year-to-year trend or seasonal variation of total ozone during the polar night (June and July), and the mean value for these months provides a baseline for determining the springtime ozone decrease. Also, the constancy of the polar night data shows that the ozone depletion in spring does not affect the following wintertime total ozone. September mean total ozone relative to the mean total ozone during June and July varies linearly with that for August showing that the ozone destruction process is active in August and September; this relationship does not appear to continue through October when corruption or displacement of the polar vortex occurs. The magnitude of springtime ozone depletion has increased from the 1980s to the 1990s. Moreover, analysis of temperature at the 100-hPa level shows that final warming at Syowa has been delayed 1.2 days per year. This suggests that the polar vortex is stabilized by the development of the Antarctic ozone hole.¿ 1997 American Geophysical Union