\begin{table}%t2
\par
\caption{\label{tab2}PSD model fitting results.}
\small%\centering
\par
\begin{tabular}{llccccccccc}
 \hline \hline \noalign{\smallskip}
\multicolumn{1}{l}{PSD} &
\multicolumn{1}{l}{Model} &
\multicolumn{1}{c}{$A$\tablefootmark{a}} &
\multicolumn{1}{c}{$\alpha$} &
\multicolumn{1}{c}{$C$} &
\multicolumn{1}{c}{$\nu_{\rm b}$} &
\multicolumn{1}{c}{$\alpha_l$} &
\multicolumn{1}{c}{$\alpha_h$} &
\multicolumn{1}{c}{$f_{\rm L}$} &
\multicolumn{1}{c}{$R$} &
\multicolumn{1}{c}{$\chi^2/$d.o.f.}  \\
 &   & (Hz$^{-1}$) &  & (Hz$^{-1}$) &  ($\times 10^{-6}$~Hz) & & 
 & ($\times 10^{-5}$~Hz) & (\%) & \\
\noalign{\smallskip} \hline \noalign{\smallskip}
{\it XMM}/soft   & PL~  & $18^{\rm +4}_{-3}$  & $2.30\pm 0.14$ & $0.14\pm 0.01$ & -- & -- & -- & -- & -- & 43.6/42 \\
\noalign{\smallskip}
{\it XMM}/soft   & PL+L & $13^{\rm +7}_{-6}$  & $2.24\pm 0.20$ & $0.14\pm 0.01$ & -- & -- & -- & $8.2^{\rm +4.9}_{-2.1}$ & 
$4.9\pm 2.5$ & 34.9/40 \\
\noalign{\smallskip}
{\it XMM}/hard   & PL~  & $19^{\rm +4}_{-3}$  & $2.11\pm 0.18$ & $0.94\pm 0.05$ & -- & -- & -- & -- & -- & 41.2/42\\
\noalign{\smallskip}
\rxte+{\it XMM}  & BPL~ & $64^{\rm +24}_{-8}$ &            --  &
$0.94$\tablefootmark{b} /906\tablefootmark{c} 
& $6.7^{\rm +4.4}_{-2.9}$ & $0.40\pm 0.17$ & $2.21^{\rm +0.24}_{-0.20}$ & -- & -- & 51.7/53 \\
\noalign{\smallskip}
\hline
\end{tabular}
\tablefoot{Following Lampton \etal\ (\cite{Lampton1976}),  all
errors indicate the 68\% confidence intervals, for 3, 5 and 4 parameters of
interest in the case of the PL, BPL~and PL+L models, respectively.\\
\tablefoottext{a}{PL~normalization which is equal to the PSD value 
at $10^{-4}$~Hz. In the case of the BPL~model, $A$ is
defined as in Eq.~(2).}
\tablefoottext{b}{Poisson noise level for the ``high-frequency'' part of the PSD.  This
value is  equal to the best-fit value of the PL~model fit to the  \xmm\ hard
band PSD, and it was kept constant during the model fit. }
\tablefoottext{c}{Expected Possion noise level for the low-frequency, \rxte\ part of the
PSD.}
} 
\end{table}