\begin{table}%t1 
\caption{\label{tab1}Size of the samples and ratios.}                                                %%\centerline

{\small 
\begin{tabular}{l l l }                             
\hline\hline                                          
SN   & Larger number statistics        & Conservative  \\
Type & $V_{\rm HEL}< 5000$~km~s$^{-1}$ &  $V_{\rm HEL}< 2000$~km~s$^{-1}$ \\     
\hline                                                      
Ic    & 49  (41)    &   18 (15)   \\
Ib    & 32  (25)    &   15 (12)   \\
Ibc   & 98  (79)    &   43 (36)  \\
II    & 318 (239)   &  142 (96)  \\
Ia    & 166 (132)   &   56 (42)  \\
\hline                                             %inserts single line
$N$(Ibc)/$N$(II)  & 0.31  $\pm$ 0.04 (0.06) &    0.30  $\pm$ 0.05 (0.06 )\\
$N$(Ic)/$N$(Ib)   & 1.53  $\pm$ 0.35 (0.30) &    1.20 $\pm$ 0.42  (0.24) \\
$N$(Ia)/$N$(CC)   & 0.40  $\pm$ 0.04 (0.08) &    0.30 $\pm$ 0.05  (0.06) \\
\hline
\end{tabular}}
\par
\medskip
{\it Top part}: the first number is the number of SN of a 
  given type for which at least the magnitude of the host is available, 
  the second number (between parenthesis) is the number of SN for which 
  the position of the SN relative to the center of the galaxy is also known.
{\it Bottom part}: SN ratios computed from the first of the two numbers above, with statistical error. The data between parenthesis indicates what would be a 20{\%} error resulting from missing systematically under-luminous SN from one of the sub-types; it is also the order of magnitude of the Shaw (\cite{Shaw1979}) effect for $V_{\rm HEL} < 2000$~km~s$^{-1}$ (see text).
\end{table}