\begin{table}%t2
\caption{\label{tab,tb}Brightness temperatures$^{{a}}$  of the VLBI sources in Arp 299-A.} 
\small%\centerline
 {
\begin{tabular}{lllccllrr}
\hline \hline
Source & \multicolumn{4}{c}{8 April 2008}                  & \multicolumn{4}{c}{5 December 2008}\\
name   & $S_\nu$ ($\mu$Jy) & $a$(mas) & $b$(mas)& $T_{\rm B}$(K) & $S_\nu$ ($\mu$Jy) & $a$(mas) & $b$(mas) & $T_{\rm B}$(K) \\
\hline
&&&&&&&& \\[-8pt]
A0 &  318 $\pm$ 42 & ${\leq}2.8$& $\cdots$    & $\cdots$ & 446 $\pm$ 33 &5.3 & 3.0 & $1.8\times 10^{6}$ \\
A1 &  855 $\pm$ 58 & ${\leq}3.6$& $\cdots$    & $\cdots$ & 901 $\pm$ 51 &2.1 & 1.8 & $1.6\times 10^{7}$ \\
A2 &  708 $\pm$ 53 & ${\leq}2.5$& ${\leq}1.5$  & ${\geq}1.3\times 10^{7}$ & 713 $\pm$ 43 & 2.1 & 1.6 & $1.4\times 10^{7}$ \\
A3 &  398 $\pm$ 44 & ${\leq}2.2$& $\cdots$    & $\cdots$ & 353 $\pm$ 30 &$\leq6.5$ & ${\leq}0.3$ & ${\geq}1.2\times 10^{7}$ \\
A4 &  558 $\pm$ 48 & ${\leq}3.6$& ${\leq}2.3$  & ${\geq}1.0\times 10^{8}$ & 628 $\pm$ 40 &$\leq2.8$ & ${\leq}0.5$ & ${\geq}3.0\times 10^{7}$ \\
A5 &  278 $\pm$ 41 & ${\leq}11.1$& ${\leq}1.2$  & ${\geq}1.3\times 10^{6}$ & 143 $\pm$ 25 & $\cdots$ &  $\cdots$ &  $\cdots$ \\
A6 &  208 $\pm$ 40 & ${\leq}7.6$& ${\leq}1.4$  & ${\geq}1.3\times 10^{6}$& $\leq$72   & $\cdots$ &  $\cdots$ &  $\cdots$ \\
A7 &  496 $\pm$ 46 & ${\leq}2.5$& ${\leq}3.3$  & ${\geq}4.0\times 10^{6}$& 468 $\pm$ 34  & 3.4    & 2.0 & $4.6\times 10^{6}$ \\
A8 &  226 $\pm$ 41 & ${\leq}4.8$& $\cdots$    & $\cdots$& 264 $\pm$ 27 & ${\leq}3.3$ &  ${\leq}0.7$ & ${\geq}7.6\times 10^{7}$ \\
A9 &  294 $\pm$ 42 & ${\leq}3.6$& ${\leq}4.4$  & ${\geq}1.2\times 10^{6}$ & 282 $\pm$ 28 & $\cdots$ &  $\cdots$ &  $\cdots$ \\
A10&  550 $\pm$ 48 & ${\leq}3.2$& ${\leq}3.9$  & ${\geq}2.9\times 10^{6}$ & 436 $\pm$ 32 & $\cdots$ &  $\cdots$ &  $\cdots$ \\
A11&  300 $\pm$ 42 & ${\leq}5.5$& $\cdots$    & $\cdots$ & 351 $\pm$ 30 &${\leq}4.5$ & ${\leq}1.8$ & ${\geq}2.0\times 10^{6}$ \\
A12&  449 $\pm$ 45 & ${\leq}3.1$& ${\leq}3.7$  & ${\geq}2.6\times 10^{8}$ & 639 $\pm$ 40 &2.3 & 1.5 & $1.2\times 10^{7}$ \\
A13&  251 $\pm$ 41 & ${\leq}3.3$& ${\leq}4.2$  & ${\geq}1.2\times 10^{6}$ & 118 $\pm$ 25 & ${\leq}5.3$ & $\cdots$ & $\cdots$ \\ 
A14&  292 $\pm$ 42 & ${\leq}5.0$& ${\leq}3.7$  & ${\geq}1.1\times 10^{6}$ & 260 $\pm$ 27 & ${\leq}4.9$ &  $\cdots$ &  $\cdots$ \\
A15&  159 $\pm$ 40 & $\cdots$ &  $\cdots$ &  $\cdots$ & 304 $\pm$ 28 & $\leq2.8$ & ${\leq}1.6$ & ${\geq}4.5\times 10^{6}$ \\
A16&  $\leq$117  & $\cdots$ &  $\cdots$ &  $\cdots$ & 147 $\pm$ 25 & 5.6 & 3.2 & $5.5\times 10^{5}$ \\
A17&  $\leq$117  & $\cdots$ &  $\cdots$ &  $\cdots$ & 179 $\pm$ 26 & 7.3 & 3.9 & $4.2\times 10^{5}$ \\
A18&  151 $\pm$ 40 & $\cdots$ &  $\cdots$ &  $\cdots$ & 129 $\pm$ 25 & ${\leq}5.1$& ${\leq}5.5$ & ${\geq}3.1\times 10^{5}$ \\
A19&  $\leq$117  & $\cdots$ &  $\cdots$ &  $\cdots$ & 191 $\pm$ 26 & ${\leq}4.0$& ${\leq}2.0$ & ${\geq}1.6\times 10^{6}$ \\
A20&  $\leq$117  & $\cdots$ &  $\cdots$ &  $\cdots$ & 146 $\pm$ 25 & ${\leq}7.9$& ${\leq}3.5$ & ${\geq}3.5\times 10^{5}$ \\ 
A21&  $\leq$117  & $\cdots$ &  $\cdots$ &  $\cdots$ & 133 $\pm$ 25 & ${\leq}7.4$& ${\leq}2.9$ & ${\geq}4.1\times 10^{5}$ \\
A22&  173 $\pm$ 40 & $\cdots$ &  $\cdots$ &  $\cdots$ & 217 $\pm$ 26 & ${\leq}8.6$& ${\leq}4.3$ & ${\geq}3.9\times 10^{5}$ \\
A23&  $\leq$117  & $\cdots$ &  $\cdots$ &  $\cdots$ & 137 $\pm$ 25 & ${\leq}8.8$& ${\leq}2.9$ & ${\geq}3.6\times 10^{5}$ \\
A24&  $\leq$117  & $\cdots$ &  $\cdots$ &  $\cdots$ & 166 $\pm$ 25 & ${\leq}8.6$& ${\leq}7.0$ & ${\geq}1.8\times 10^{5}$ \\
A25&  132 $\pm$ 40 & $\cdots$ &  $\cdots$ &  $\cdots$ & 209 $\pm$ 26 & ${\leq}8.6$& ${\leq}4.4$ & ${\geq}3.7\times 10^{5}$ \\
\hline
\end{tabular}}   
\medskip 
$^{{a}}$ The brightness temperatures shown for the 5.0~GHz VLBI source components in Arp~299-A were calculated using the flux densities in Table~\ref{tab,evn} and the angular sizes quoted here.  We derived the brightness temperatures from the general formula: $T_{\rm B} = (2~c^2/k)~B_\nu~\nu^{-2}$, where $B_\nu$~is the intensity, in erg~s$^{-1}$~Hz$^{-1}$~str$^{-1}$.  Since $B_\nu$ depends on the measured flux density, $S_\nu$, and on the deconvolved angular
size of each VLBI~component (obtained by fitting them to elliptical Gaussians, characterized by their major and minor semi-axis, $a$ and $b$). Therefore, the above formula can be
rewritten as $T_{\rm B} = (2~c^2/k)~B_\nu~\nu^{-2} = 1.66$~$\times$ $10^{9}~~S_\nu ~ \nu^{-2} (a~b)^{-1}$, where $S_\nu$~is in mJy, $\nu$~in~GHz, and~$a$ and~$b$ are in milliarcseconds, respectively.
\end{table}