\begin{table}%t8
\caption{\label{t8}Ammonia column densities derived from the Medicina spectra.}
\par
%\centerline
{
\begin{tabular}{cccc}
\hline\hline\noalign{\smallskip}
Name & $T_{\rm rot}$ & $N_{\rm 22}$\tablefootmark{a} & $N_{\rm tot}$\tablefootmark{a} \\
     &         (K)      & (10$^{12}$ cm$^{-2}$)  & (10$^{14}$ cm$^{-2}$)  \\
\hline
G12.22--0.12\tablefootmark{b} &  26(6) & 22.7(2.3) &  1.5(0.1) \\
G12.22--0.12M  &    20(4) & 13.4(4.3) &  1.4(0.1) \\
G23.44--0.18   &    15(2) & 27.5(6.0) & 29.8(2.8) \\
G31.28+0.06    &    17(3) &  7.0(1.3) & 10.8(1.4) \\
G32.74--0.07   &    21(6) &  9.0(2.7) &  6.9(1.2) \\
G35.03+0.35    &    21(3) &  5.4(1.5) &  0.5(0.1) \\
G35.20--0.74\tablefootmark{b} &  21(2) & 32.5(1.6) &  9.2(0.6) \\
G35.20--1.74   &    23(2) & 10.1(1.1) &  0.8(0.3) \\
G43.89--0.38   &    23(4) & 13.1(3.8) &  1.1(0.1) \\
G48.61+0.02    &    18(4) &  1.3(0.7) &  0.2(0.1) \\
G75.78+0.34   &    20(3) & 17.4(4.0) &  1.8(0.1) \\
\hline
\end{tabular}}
\par
\tablefoot{\tablefoottext{a}{For all the sources but G23.44--0.18, G31.28+0.06, and G35.20--1.74 (where the optical depths
of both ammonia transitions have been derived, see Table 5), we have assumed optically thin emission for
the NH$_3$(2, 2) line.} 
\tablefoottext{b}{Given the spatial resolution of the Medicina antenna, the spectra collect the contributions due to
both A and B cores (see Figs.~7--10).}}
\end{table}