\begin{table}%t8
\par
\caption{To evaluate the mean energies of the three analysed bands, 
we used the spectral indexes of the best-fit power-law models
in the \mbox{[0.6--10]}~keV ranges 
($\alpha = 1.15\pm0.01$ and $\alpha = 1.535\pm0.003$). 
We do not report the errors on the mean energies which are of the order 
of $10^{-4}$~keV.
The notations $\delta_{10}$ and $\xi_5$ mean ($\delta/10$) and
$\xi/10^5$, respectively. 
\label{tab6}}
%\centerline
{\begin{tabular}{ccccccc}
\hline\hline
$E_1$ & $E_2$ & $E_3$ & $\tau_{12}$ & $\tau_{13}$ & $B_{12}$ &
 $B_{13}$ \\
(keV) & (keV) & (keV) & (s) & (s) & (G) & (G) \\
\hline
\multicolumn{7}{c}{November 14 soft lag}\\
\hline
0.42 & 1.44 & 5.19 & 5800 & 11~600  & $0.24\pm0.02~\delta_{10}^{-1/3}$ & $0.21\pm0.01~\delta_{10}^{-1/3}$ \\
\hline
\multicolumn{7}{c}{December 1 hard lag}\\
\hline
0.40 & 1.38 & 4.87 & $450\pm200^*$ & $950\pm200^*$ & $0.53\pm0.12 ~\delta_{10}^{-1} \xi_5^{2/3}$ & $0.65\pm0.08 ~\delta_{10}^{-1} \xi_5^{2/3}$\\
\hline
\end{tabular}}
\smallskip
$^*$ Since the uncertainties reported
 in Table~\ref{tab5} are probably underestimated, we assumed more 
conservative errors of 200~s.
\end{table}