We have found by reanalyzing the data used in Hubrig et al. (2019a) that the rapid change of the spectrum and of the magnetic field reported for MJD 58166.0420 was due to an error in the spectral extraction from the two-dimensional raw data, very likely due to an insufficient signal-to-noise ratio – see also the discussion in Hubrig et al. (2019b). Table 1 of Hubrig et al. (2019a) presents 29 measurements. Only the measurement corresponding to MJD 58166.0420 is wrong and should be disregarded. In fig. 1 of Hubrig et al. (2019a), where all available magnetic field measurements are plotted, the measurements on MJD 58166.0420 – using the entire spectrum or exclusively the hydrogen lines – correspond to the low values at the right side of both plots. A similar plot, presenting all correct measurements, is shown in the top panel of fig. 1 in Hubrig et al. (2019b). Figs 2, 3, and 6 in Hubrig et al. (2019a) display the results of the wrong spectrum extraction and should be disregarded, also all references to the measurement on MJD 58166.0420 in the text, especially in Section 3. In other figures of Hubrig et al. (2019a), the wrong measurement is clearly denoted by red colour and appears along with the correct measurements in figs 4 and 5 (red dashed lines), and in figs 7–9 (red triangles). It has, however, no impact on the trends detected in these figures. Figs 10–14 and tables 2 and 3 are not affected by the wrong measurement. The observations reported in table 1, apart from the spectrum obtained on MJD 58166.0420, are used in Hubrig et al. (2019b) to discuss the evolution of the magnetic field, radial velocities, and equivalent widths.