The variability of total ozone at mid-latitudes and its relation to dynamical and chemical mechanisms in the stratosphere have been widely studied using recent satellite data. However, relatively little is still known about total ozone variability in earlier, "chemically unperturbed" time periods. For this purpose, we have re-evaluated historical total ozone data from a network of Dobson instruments in Europe from 1952 to 1957 (supplemented up to 1963 using the WOUDC data base). For four locations between 45ºN and 60ºN we have analyzed the total ozone variability; and the relation to atmospheric circulation is studied using NCEP/NCAR reanalysis data. The variability of total ozone is explored mainly with respect to the lower tail of the frequency distribution, i.e., ozone mini holes or, more generally, low-ozone events. The results from the earlier time period are compared to results from a more recent time period (1979-2000) using TOMS total ozone data. The re-evaluated historical ozone data from 1952 to 1963 show many low-ozone events (less than 225 DU). However, the events were confined almost exclusively to the late autumn season (Oct.-Dec.). A case study is presented of an event in November 1953, when total ozone dropped below 200 DU at two sites in the UK in conjunction with a strong upper ridge and low temperatures at 30 hPa. Only one low-ozone event was recorded in January or February. In contrast, in the more recent time period (TOMS data), low-ozone events occurred frequently also in the late winter months. Apart from the chemical ozone depletion that has taken place since the 1970s, changes in atmospheric circulation probably also contributed to this difference. Ozone mini hole events over the Atlantic-European sector occur more often in the positive phase of the North Atlantic Oscillation (NAO) than in the negative one. For October to December, the NAO index was similar for both periods, whereas for January and February, it was significantly higher in the later period compared to the earlier period. Several attempts are presented to discern between changes in the circulation and chemically induced changes in the stratospheric ozone amount in order to explain the differences in the frequency distributions and in the frequency of occurrence of low-ozone events.