\begin{table}%t5
\caption{\label{tabfits}Fits to the data  ($b = 0.25$).} 
\small%\centerline
 {
\begin{tabular}{lrrrr}        
\hline\hline                 
%&&&& \\[-8pt]
Relationship & Slope & Y intercept & $\chi^2$ & Corr.? \\
\hline
$T$ vs. 1 + 
$z$            &$      -2.00   \pm     0.2             $&$     6.1     \pm 
   0.5     $&      90.6    &Y \\
Log $T$ vs. Log (1 + 
$z$)  &$      -1.80   \pm     0.2             $&$     0.9     \pm     0.06 
  $&      100.8   &Y \\
$T$ vs. $r$              &$      -4.00   \pm     1.4     \times 10^{-3}$ 
&$      2.5     \pm     0.3     $&      181.6   &N
\\
Log $T$ vs. Log 
$r$      &$      -0.25   \pm     0.11            $&$     1       \pm 
0.2     $&      183     &N \\
$E_{\rm T}$ vs. 1 + 
$z$          &$      1.70    \pm     0.3             $&$     -1.1    \pm 
0.5     $&      92.8    &Y \\
Log $E_{\rm T}$ vs. Log (1 + 
$z$)        &$      1.80    \pm     0.2             $&$     -0.1    \pm 
0.06    $&      100.8   &Y  \\
$E_{\rm T}$ vs. $r$            &$      1.40    \pm     1.4     \times 10^{-3}$ 
&$      1.5     \pm     0.2     $&      138     &N\\
Log $E_{\rm T}$ vs. Log 
$r$            &$      0.25    \pm     0.11            $&$     -0.2    \pm 
     0.2     $&      183     &N \\
$n_{\rm a}$ vs. $r$           &$      -1.90   \pm     0.9     \times 10^{-4}$ 
&$      0.09    \pm     0.04    $&
347.3   &N\\
Log $n_{\rm a}$ vs.  Log 
$r$          &$      -1.90   \pm     0.2             $&$     3.4     \pm 
   0.5     $&      561.9   &Y \\
Log $n_{\rm a}$ vs.  Log (1 + 
$z$)      &$      -1.90   \pm     1.1             $&$     -0.3    \pm 
0.3     $&      1213.6  &N \\
$v/c$ vs.        $r$      &$      1.30    \pm     0.3     \times 10^{-4}$ 
&$      0.006   \pm     0.003   $&
741.1   &N \\
Log $v/c$ vs. Log 
$r$             &$      0.70    \pm     0.1             $&$     -3 
\pm     0.2     $&      740.9   &N \\
$v/c$ vs. 1 + 
$z$           &$      0.02    \pm     0.004           $&$     -0.01   \pm 
  0.006   $&      747.5   &N\\
Log $v/c$ vs. Log (1 + 
$z$)         &$      1.50    \pm     0.4             $&$     -1.88   \pm 
0.1     $&      966.9   &N \\
$L_j$ vs. 
$r$           &$      0.01    \pm     0.03            $&$     16      \pm 
    5       $&      364.8   &N \\
Log $L_j$ vs. Log 
$r$           &$      0.60    \pm     0.2             $&$     0.3     \pm 
    0.5     $&      710.6   &N \\
$L_j$ vs. 1 + 
$z$                 &$      36.50   \pm     9               $&$     -38 
\pm     13      $&      280.1   &N \\
Log $L_j$ vs. Log (1 + 
$z$)        &$      3.80    \pm     0.5             $&$     0.72    \pm 
0.11    $&      361.4   &Y \\
\hline
\end{tabular}}
\medskip
The best fits to the radio source relationships for $b = 0.25$. Column~1: the relationship. $v/c$~is the lobe expansion velocity, $r$~is the core-hotspot separation, $L_j$ ($\times$$10^{44}$~erg~s\mone) is the beam power, $n_{\rm a}$ ($\times 10^{-3}$~cm \mthree) is the ambient density, $T_{\rm T}$ is the source lifetime, and $E_{\rm T}$ is the total energy. Column~2: the best fit slope. Column~3: the best fit Y intercept. Column~4: the  $\chi^2$ for 60~points and 58~degrees of freedom.
Column~5: is there a correlation? The uncertainty of the best fit parameter is multiplied by $\sqrt{(\chi^2/58)}$ to obtain the normalized uncertainty in the best fit parameter, and this is
compared with the value of the best fit parameter to determine whether the correlation is significant.
\end{table}