\begin{table}%t1
\caption{Best-fit parameters of the multi-component model \wabs (\comptb\ + \comptb\ + \gaussian).}
%\centerline
 {\small
\begin{tabular}{p{2cm}p{2cm}p{2cm}p{2cm}p{2cm}p{2cm}}
\hline \hline  
Parameter   &  $UHB$  & $LHB$  &  $UNB$  &   $MNB$  & $LNB$ \\
\hline
$N_{\rm H}^{a}$ &[0.15]   &[0.15]   & [0.15]   &  [0.15] &  [0.15]\\
\hline
\multicolumn{6}{c}{\comptb\  (thermal: $\log$[A]=8, $\delta$=0)}\\
%\hline
$\kts$ (keV) & $\rm 0.28^{+ 0.02}_{- 0.03}$ & $\rm 0.29^{+  0.02}_{-0.02}$ & $\rm 0.31^{+0.01}_{-0.01}$ & $\rm 0.30^{+  0.01}_{-0.01}$ & $\rm 0.31^{+0.01}_{-0.01}$\\
$\kte$ (keV) & $\rm 2.72^{+ 0.06}_{-0.08}$ & $\rm 2.65^{+  0.08}_{-0.10}$ &$\rm 2.41^{+0.04}_{-0.04}$  &  $\rm 2.23^{+0.06}_{-0.05}$   & $\rm 2.51^{+ 0.06}_{-0.08}$\\
$\alpha$ & $\rm 0.80^{+0.03}_{-0.03}$ & $\rm 0.82^{+0.07}_{-0.03}$  & $\rm 0.80^{+0.03}_{-0.03}$ & $\rm 0.82^{+  0.04}_{-0.04}$& $\rm 1.00^{+0.08}_{-0.02}$\\
$\tau^{b}$ & 6.5 &  6.4 &  6.9 &  7.1  &  5.8\\
$\rm CAF^{c}$&   2.8  & 2.6  & 2.5  & 2.4  &  2.2\\
\noalign{\smallskip}\hline
\multicolumn{6}{c}{\comptb\ }\\  
$\kts$ (keV) &  $\rm 1.00^{+0.08}_{-0.11}$ & $\rm 1.15^{+0.05}_{-0.14}$ & $\rm 1.11^{+0.02}_{-0.02}$&  $\rm 1.08^{+0.03}_{-0.03}$& $\rm 1.16^{+ 0.03}_{-0.04}$ \\
$\rbb^d$ (km) & 3.4  & 3.2 & 4.6  & 4.9  & 4.6\\
$\kte$ (keV) & [$\kte$] & [$\kte$]& [$\kte$]  & [$\kte$]& --\\
$\log(A)$&  0.11~($>$--0.12)& --0.40~($>$--0.65) & [--0.4] & [--0.4] &[--8]\\
$\alpha$ & [2.5] & [2.5] &  [2.5] & [2.5] & -- \\
$\delta$ &  75 ($>$40) & $\rm 49^{+ 39}_{- 18}$ & $\rm 25^{+  5}_{-4}$ & $\rm 22^{+5}_{-4}$& -- \\
\hline
\multicolumn{6}{c}{\gaussian\ }   \\  
%\hline
$E_{l}$ (keV) &  $\rm 6.8^{+ 0.1}_{- 0.1}$ &  6.7 $\rm ^{+ 0.1}_{- 0.1}$ & $\rm 6.7^{+  0.1}_{-0.1}$  & $\rm 6.6^{+  0.1}_{-0.1}$ & $\rm 6.7^{+  0.1}_{-0.1}$ \\
$\sigma_{l}$ (keV) & $\rm 0.3^{+ 0.2}_{- 0.1}$ & 0.2~($<$0.3) & 0.1~($<$0.2) & 0.12~($<$0.2)& 0.1 ~($<$0.2) \\
$I_{l}$ & $\rm 2.5^{+ 0.9}_{- 0.7}$ & $\rm 1.4^{+0.7}_{-0.6}$ & $\rm 1.9^{+0.7}_{-0.7}$ & $\rm 1.8^{+0.9}_{-0.8}$& $\rm 3.0^{+1.7}_{-0.9}$	\\
$EW_{l}$ (eV) & $\rm 38^{+ 18}_{-9}$ & $\rm 17^{+8}_{-8}$ & $\rm 15^{+5}_{-5}$ & $\rm 14^{+7}_{-6}$& $\rm 29^{+16}_{-9}$	\\
%\hline
$L_{\rm bb}^e/L_{\rm tot} $ &0.25 & 0.26 & 0.25 & 0.25 & 0.28\\
%\hline
$L_{\rm th}^f/L_{\rm tot} $ &0.69 & 0.67 & 0.63 & 0.61 & 0.63\\
$L_{\rm tot}^{\rm g} $ &0.9 & 1.1 &1.7 & 1.7 & 1.5\\
%\hline
${\chi^{2}}$/d.o.f. & 175/152 & 162/154  &   159/148  & 185/144  & 197/143 \\
\hline
\end{tabular}}
\medskip
Errors are computed at 90\% confidence level for a single parameter. Additional emission lines at $\sim$1~keV and $\sim$2.5~keV were included in all the five spectra (see Sect.~2).  $^{a}$ In units of 10$^{22}$~cm$^{-2}$; $^{b}$~computed from  $\kte$ and $\alpha$ using Eqs.~[17] and~[24] for slab geometry in Titarchuk \& Lyubarskij (1995); $^{c}$~Compton Amplification Factor (see text for the definition); $^{d}$ computed using the simple relation $L = 4\pi \sigma R^2 T^4$;  $^{e}$~ratio between the direct BB~component flux of the second \comptb\ and the total flux in the 0.1--200~keV energy range; $^{f}$~ratio between the TC~component flux and the total flux in the 0.1--200~keV energy range; $^{g}$~in units of 10$^{38}$~erg~s$^{-1}$ computed in the energy range 0.1--200~keV assuming a source distance of 8~kpc. 
\label{tab_fit}
\end{table}