\begin{table}%t4
\caption{\label{tab:var}Number and percentage of variable X-ray counterparts.}
%\centerline
{
\begin{tabular}{c c| c c c | c c c c}
\hline\hline
Q. flag & Sp. Type & \multicolumn{3}{c|}{Within one exposure} & \multicolumn{3}{c}{Between exposures} &\\     
& & total & single & binaries & total & single & binaries & not tested\\     
\hline                        
0 & O  & 0 & 0 & 0 & 4 (25\%)& 3 (25\%) & 1 (25\%) & 19 (54\%)\\ 
0 & B  & 4 (4\%) & 4 (5\%) & 0 & 7 (32\%)& 7 (33\%)& 0 & 69 (76\%)\\ 
0 & WR & 0 & 0 & 0 & 2 (50\%)& 1 (50\%)& 1(50\%) & 3 (43\%)\\ 
\hline                                   
$\neq$0 & O  & 4 (4\%)& 3 (4\%)& 1 (5\%)& 44 (69\%)& 32 (67\%)& 12 (75\%)& 34 (35\%)\\ 
$\neq$0 & B  & 4 (6\%)& 4 (6\%)& 0 & 14 (67\%)& 12 (63\%)& 2 (100\%) & 70 (49\%)\\ 
$\neq$0 & WR  & 0 & 0 & 0 & 6 (86\%)& 0& 6 (100\%) & 2 (22\%)\\ 
\hline                                   
\end{tabular}}
\par
\medskip
When comparing several exposures, the percentages noted for non-tested objects were calculated with respect to the entire sample, whereas those of the variable sources refer only to the tested subsample.  
\end{table}