The authors found no correlation between crystal growth angle and twinning incidence. This is contrary to the predictions of the Hurle model [1]. A possible explanation for this is presented here. It is shown that the spatial resolution of their measurement of growth angle was insufficient to determine its value at the point at which twinning was initiated. Their measured value represents an average over a growth time, which was large compared to the likely characteristic temperature fluctuation time. Therefore the author's conclusion that "the role of slope angle seems to be obscured by much stronger forces...." cannot properly be inferred from their data. (C) 2010 Elsevier B.V. All rights reserved.