\begin{table}%e4
  \caption {\label{abundances2}Abundances  for Arcturus and the three bulge giants relative to the Sun, where ${\rm [X/Fe]} = \{ {\rm \log\varepsilon(X)} - {\rm  \log\varepsilon(Fe)\}_{star}} - \{ {\rm  \log\varepsilon(X) } - {\rm  \log\varepsilon(Fe)\}_\odot }$.}
\small%\centerline
 {
  \begin{tabular}{l l c c c c c c c c c c }
  \hline\hline 
    Star &  Ref.   Star &  [C/Fe]   & [N/Fe] & [C+N/Fe] &  [O/Fe]  & [Ti/Fe]& [Si/Fe] & [S/Fe] &[Cr/Fe] & [Ni/Fe] & [Fe/H]\\
  \hline
 {Arcturus}&   this work$^{{a}}$ & 0.15 & 0.37 & 0.20  & 0.60  & 0.16 & 0.30 & 0.24 & 0.00 &0.03  & --0.50\   \\
 & \citeauthor{fulbright:07}$^{{b}}$ &  -- & -- & --  & 0.51  & 0.16 & 0.34 & -- &-- & -- & --0.50\\
 {Arp~4203} & this work$^{{a}}$ &  $-$0.54 & 1.15 & 0.48 & 0.30  & 0.21 & 0.45 & 0.31 & $-$0.14 & 0.12 & $-$1.25\\
  & \citeauthor{fulbright:07}$^{{b}}$ &  -- & -- &--  &  0.14  & 0.26 & 0.52 & -- & --& -- & $-$1.25 \\  
 {Arp~4329} & this work$^{{a}}$ & $-$0.09  & 0.33& 0.03 & 0.49  & 0.21 & 0.50 & 0.51 & 0.00  & 0.06 & $-$0.90\\
  & \citeauthor{fulbright:07}$^{{b}}$ &  -- & -- & --  & 0.40  & 0.18 & 0.49 & -- & -- & -- & $-$0.90 \\
 {Arp~1322} & this work$^{{a}}$ &  $-$0.25 & 0.53& 0.05 & 0.18 & 0.08 & 0.20 & 0.38 & 0.18 & 0.00& $-$0.17$^{{c}}$\\
  & \citeauthor{fulbright:07}$^{{b}}$ &  -- & -- & --  &  0.37   & 0.05 & 0.11 & -- & -- & -- & $-$0.23\\
    \hline
\end{tabular}}
\medskip
$^{{a}}$ We have used the following solar abundances:  ${\rm  \log\varepsilon(C)=8.41}$, ${\rm  \log\varepsilon(N)=7.80}$, ${\rm  \log\varepsilon(O)=8.66}$, ${\rm  \log\varepsilon(Fe)=7.50}$, ${\rm  \log\varepsilon(Ti)=5.02}$, ${\rm  \log\varepsilon(Si)=7.55}$, ${\rm  \log\varepsilon(S)=7.20}$, ${\rm  \log\varepsilon(Cr)= 5.67}$, and ${\rm  \log\varepsilon(Ni)=6.25}$. \\
$^{{b}}$ \citeauthor{fulbright:07} used the following solar abundances:  ${\rm  \log\varepsilon(O)=8.69}$, ${\rm  \log\varepsilon(Fe)=7.45}$, ${\rm  \log\varepsilon(Ti)=4.92}$, and ${\rm  \log\varepsilon(Si)=7.54}$. For the sake of comparison we have scaled their values to the solar values we have used. \\
$^{{c}}$  The [Fe/H] from this work  is determined from Fe{\sc i}~lines in the CRIRES~wavelength region, but not used as the metallicity in the models.
\end{table}