\begin{table}%t2
\caption{ "Inside correlations'' between gradients from various sources, 
used in the estimation of probable errors in each dataset. 
The impartial correlation $Y=pX+a$ is given.  
\label{tab2}}
%\centerline
{\begin{tabular}{lllllrll}
\hline   \hline     
 X & Y & S & $N$ & \multicolumn{1}{c}{$p$} & \multicolumn{1}{c}{$a$} & $\rho$& $\sigma$  \\
 \hline
$\Delta_{BV}$&$\Delta_{UB}$&I& 39 & $2.10\pm .23$ & $-.005\pm .005$ & .75 & .029\\
$\Delta_{BV}$&$\Delta_{VI}$&I& 39 & $1.12\pm .16$ & $.022\pm .003$ & .51 & .020\\
$\Delta_{BV}$&$\Delta_{BR}$&I& 37 & $1.33\pm .10$ & $.004\pm .002$ & .90 & .012\\
$\Delta_{BV}$&$\Delta_{VI}$&G& 32 & $1.39\pm .30$ & $.030\pm .006$ & .45 & .035\\
$\Delta_{BR}$&$\Delta_{UR}$&P& 25 & $2.00\pm .29$ & $.083\pm .010$ & .73 & .051\\
\hline
\end{tabular}}
\par
\smallskip
(1) $X$ variable; (2) $Y$ variable; 
(3) S source reference abbreviated as I for IMP02 (enlarged from several OHP 
measurements), G for Gal94, P for Pal90 and Fal89 merged; 
(4) number of objects~$N$ after rejection of outliers; (5) slope~$p$; 
(6) zero point $a$; (7) coefficient of correlation $\rho$; (8) dispersion~$\sigma$.  
\end{table}