The grain $( {E}^{\prime}_{cc} \hbox{ and }{E}^{\prime}_{ca}) $ and bulk (E _mm and E_mt) enhancement factors should be raised to a power of 2/(n + 1) throughout, so that

$$\eqalign{& E^{\prime}_{cc} \rightarrow {E}^{\prime 2/( {n}^{\prime} + 1)}_{cc} , \cr & {E}^{\prime}_{ca} \rightarrow {E}^{\prime 2/( {n}^{\prime} + 1)}_{ca} , \cr & E_{mm} \rightarrow E_{mm}^{2/( n + 1) } , \;\cr & E_{mt}\rightarrow E_{mt}^{2/( n + 1) } , \;}$$

to be consistent with the enhancement-factor definition (11) for n ≠ 1. Only if the orientation-independent fluidity (5) is used instead of (2) do the equations stand correctly in the original paper for n ≠ 1.

The conclusions of the original paper are unaffected by this correction, but the best-fit grain enhancements factors, needed to reproduce bulk behaviour for n′ = 3, are different. Figure 1 shows the revised version of Figure 3 (panel a is unchanged) where ${E}^{\prime}_{ca} = 10^4$ is found to best fit bulk behaviour, as opposed to ${E}^{\prime}_{ca} = 10^2$ in the original paper.

Figure 1.

Revised Figure 3.