An error occurred in Philosophical Magazine, Volume 90, Issues 15–16, 21–28 May 2010, pp. 2161–2190, in the article ‘Statistical analyses of deformation twinning in magnesium’, by I.J. Beyerlein, L. Capolungo, P.E. Marshall, R.J. McCabe and C.N. Tomé.
In Figure 11b, an error was made in calculating the number of twinned grains as a function of grain diameter and thus also twinned frequency. The original and corrected figures are shown below. As you can see, the correct trend as a function of Grain Diameter (Figure 11b corrected) is similar to the trend as a function of Grain Area (Figure 11a), and it is considerably different than the trend in the published figure (Figure 11b original). Perhaps fortuitously, Figure 11b was added primarily to show an alternative representation of grain size, while the results, discussion, and conclusions are based on the results shown for grain area in Figure 11a. Our primary conclusion based on Figure 11a was that the propensity for a grain to twin has little dependence on grain size, except perhaps among the smallest grains in the sample (ag
< 1000 µm2, which corresponds to a diameter of about 35 µm). This conclusion does not change based on the corrected figure.
Figure 12a intends to plot the number of twins per grain (all variants) as a function of grain size, but only plots the number of variant 1 twins per grain. The primary conclusion based on Figure 12 is that larger grains have more twins. This conclusion is still valid based on only variant 1 twins as they account for around half of all of the observed twins. However, the magnitude of the variation of variant 1 twins with grain size is obviously less than it is for twins of all variants.
The authors would like to apologise for these mistakes.