\begin{table}%t2
\caption{\label{tab:excited}Column densities of the fine-structure levels of the
  ground term and those of metastable levels of Fe$^+$ at the \grb-DLA
  redshift of $z_{\rm abs}=3.96857$.}
%\centerline
 {
\begin{tabular}{lllc}
\hline
\hline\\[-3.2mm]
Ion          &
Energy level &
Transition   &
${\log N\pm\sigma_{\log N}}^{b}$\\
&& lines used (\AA )&\\
\hline
\multicolumn{4}{l}{1st epoch of observations:}\\
Fe$^+$ & g.s.$^{a}$ ($^6{\rm D}_{9/2}$) & 1083, 1112             &     $\rm 15.31^{+0.04}_{-0.05}$       \\[0.5mm]
Fe$^+$ & $^6{\rm D}_{7/2}$                 & 1266, 1267             &     $\rm 14.39^{+0.10}_{-0.12}$       \\[0.5mm]
Fe$^+$ & $^6{\rm D}_{5/2}$                 & 1271                   &     $\rm 14.36^{+0.09}_{-0.10}$       \\[0.5mm]
Fe$^+$ & $^6{\rm D}_{3/2}$                 & 1275.7,  1275.8        &     $\rm 14.04^{+0.12}_{-0.16}$       \\[0.5mm]
Fe$^+$ & $^6{\rm D}_{1/2}$                 & 1277.64, 1277.68       &    ${<}13.60^{c}$             \\
\hline
\multicolumn{4}{l}{2nd epoch of observations:}\\
Fe$^+$ & g.s.$^{a}$ ($^6{\rm D}_{9/2}$) & 1611                   &     $\rm 15.34^{+0.06}_{-0.06}$       \\[0.5mm]
Fe$^+$ & $^6{\rm D}_{7/2}$                 & 1618, 1621             &     $\rm 14.41^{+0.03}_{-0.03}$       \\[0.5mm]
Fe$^+$ & $^6{\rm D}_{5/2}$                 & 1629, 1631             &     $\rm 13.96^{+0.06}_{-0.07}$       \\[0.5mm]
Fe$^+$ & $^6{\rm D}_{3/2}$                 & 1634, 1636             &     $\rm 13.79^{+0.09}_{-0.09}$       \\[0.5mm]
Fe$^+$ & $^6{\rm D}_{1/2}$                 & 1639                   &    ${<}13.65^{d}$             \\
Fe$^+$ & $^4{\rm F}_{9/2}$                 & 1559, 1566, 1569       &                                   \\
       &                              & 1610, 1612, 1658       &                                   \\
       &                              & 1702                   &     $\rm 14.31^{+0.03}_{-0.04}$       \\
Fe$^+$ & $^4{\rm F}_{7/2}$                 & 1563, 1625, 1659       &                                   \\
       &                              & 1712                   &     $\rm 13.64^{+0.10}_{-0.10}$       \\[0.5mm]
Fe$^+$ & $^4{\rm F}_{5/2}$                 & 1570, 1584, 1633       &     $\rm 13.50^{+0.11}_{-0.14}$       \\
Fe$^+$ & $^4{\rm F}_{3/2}$                 & 1574, 1640             & ${\le} 13.39^{+0.11d}_{-0.12}$\\
Fe$^+$ & $^4{\rm D}_{7/2}$                 & 1635                   &     $\rm 13.46^{+0.08}_{-0.08}$       \\[0.5mm]
Fe$^+$ & $^4{\rm D}_{5/2}$                 & 1641                   & ${\le} 13.29^{+0.13d}_{-0.15}$\\ 
\hline
\end{tabular}}
\medskip
\par
$^{a}$ Ground state. 
$^{b}$ Listed uncertainties were derived from the line fitting using the
best-fit continuum normalisation plus or minus 0.5~times the noise rms
in the adjacent continuum. 
$^{c}$ $3\sigma$ upper limit. 
$^{d}$ Possible blend or mis-identification (see text).
\end{table}