...2003)

Due to a program error the full line width given in this report was divided by a factor of two. In addition the equivalent width of the investigated spectral line was given as 81 instead of 57 mÅ (Table I of that paper).

.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.

... 0.26''

This is because of the finite width of the spectral-line formation-region, which translates to a horizontal sampling for an inclined line-of-sight. The depression contribution-functions for the Fe II 6456.38 Å line peak at $\tau_{\rm c} = 0.5$ to $\tau_{\rm c} = 0.05$ when going from line core to the line wing with a spread of more than three decades, so that the full line forms over a region from $\tau_{\rm c} = 2$ to less than $\log\tau_{\rm c} = -3$.

.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.