All Figures

$\begin{figure} \par\includegraphics[width=7.1cm,clip]{2967f1.eps} \end{figure}$ Figure 1: The $T_{\rm eff}$ versus $\log g$ Kiel diagram for the sample. The diameter of the circle for each star is proportional to [C/Fe]; cases where C is undetected are shown with an asterisk. The mean error bar is shown in the bottom right.
Open with DEXTER

In the text

$\begin{figure} \par\includegraphics[width=7.4cm,clip]{2967f2.eps} \end{figure}$ Figure 2: The distribution of analysed stars in [Fe/H]. The [Fe/H] values are from the final analysis of the spectra.
Open with DEXTER

In the text

$\begin{figure} \par\includegraphics[width=7.7cm,clip]{2967f3a.eps}\par\vspace*{3mm} \hspace*{2mm}\includegraphics[width=8cm,clip]{2967f3b.eps} \end{figure}$ Figure 3: Change in E(B-V) ( upper panel) and $T_{\rm eff}$ ( lower panel) if we adopt the reddening maps of Burstein & Heiles (1982) instead of those of Schlegel et al. (1998). The reddening corrections are usually reduced and thus the temperatures become somewhat cooler.
Open with DEXTER

In the text

$\begin{figure} \par\includegraphics[width=6.5cm,clip]{2967f4.eps} \end{figure}$ Figure 4: Histogram of the slopes of fits to the trends of Fe I line abundances with excitation for the sample.
Open with DEXTER

In the text

$\begin{figure} \par\includegraphics[width=7.5cm,clip]{2967f5a.eps}\par\vspace*{3mm} \includegraphics[width=7.5cm,clip]{2967f5b.eps} \end{figure}$ Figure 5: Difference between the final [Fe/H] as determined from the spectrum and the initial guess determined from medium-resolution follow-up spectroscopy, where $\Delta$ [Fe/H] = [Fe/H](initial)-[Fe/H](final), plotted against $T_{\rm eff}$ and [Fe/H](final).
Open with DEXTER

In the text

$\begin{figure} \par\includegraphics[width=7.4cm,clip]{2967f6.eps} \end{figure}$ Figure 6: Difference between the final $\log g$ as determined from the spectrum and the initial guess as described in the text, where $\Delta \log g = \log g {\rm (initial)} - \log g {\rm (final)}$ .
Open with DEXTER

In the text

$\begin{figure} \par\includegraphics[width=7.6cm,clip]{2967f7.eps} \end{figure}$ Figure 7: Plot of final microturbulence values $\xi$ as derived from the spectrum against gravity. Note that the stars lying higher than the general trend are the red horizontal branch stars (see Fig. 1).
Open with DEXTER

In the text

$\begin{figure} \par\includegraphics[width=6.7cm,clip]{2967f8a.eps}\hspace*{3mm} ... ...c.eps}\hspace*{3mm} \includegraphics[width=6.7cm,clip]{2967f8d.eps} \end{figure}$ Figure 8: Plots showing trends of abundances found from individual Fe and Ti lines with excitation potential $\chi$ ( left column) and wavelength $\lambda$ ( right column) for two comparison stars, CS 22186-025 $S/N\sim 32$ and CS 31082-001 $S/N\sim 90$ . Circles and crosses indicate lines from the neutral and singly ionised species respectively. The dashed line in each case shows the determined abundance from the simultaneous best fit to all lines.
Open with DEXTER

In the text

$\begin{figure} \par\includegraphics[width=7.4cm,clip]{2967f9a.eps}\hspace*{3mm} ... ...mm} \includegraphics[width=7.4cm,clip]{2967f9e.eps}\hspace*{7.2cm}\end{figure}$ Figure 9: Comparison of abundances for stars previously studied in the literature. Differences are in the sense of $\Delta \log \epsilon = \log \epsilon_{\rm this\; work} - \log \epsilon_{\rm literature}$ . Error bars indicate the estimated absolute error in our result. The mean difference and standard deviation $\sigma$ are reported.
Open with DEXTER

In the text

$\begin{figure} \par\includegraphics[width=7.4cm,clip]{2967f10a.eps}\includegraphics[width=7.4cm,clip]{2967f10b.eps} \end{figure}$ Figure 10: Comparison of abundances as in Fig. 9, where we have adopted the same temperature as the literature study.
Open with DEXTER

In the text

$\begin{figure} \par\includegraphics[width=7.4cm,clip]{2967f11.eps} \end{figure}$ Figure 11: Comparison of abundances as in Fig. 9, for HD 20 where we have adopted the same temperature, gravity and microturbulence as Burris et al. (2000).
Open with DEXTER

In the text

$\begin{figure} \par\includegraphics[width=7.4cm,clip]{2967f12a.eps}\hspace*{3mm}... ...\vspace*{3mm} \par\includegraphics[width=16.4cm,clip]{2967f12c.eps} \end{figure}$ Figure 12: Examples of the spectra and fits for HE 1127-1143 (an r-process rich star), which has average $S/N\sim 49$ , $T_{{\rm eff}}=5224$ K, $\log g\sim2.6$ , ${\rm [Fe/H]}\sim -2.7$ , and is thus reasonably typical for the sample. The double line is the observed spectrum and the single line the fitted model spectrum (noting that it is not only the plotted observations which are being fit, but all considered lines of the element). The shaded regions are the comparison windows used for fitting.
Open with DEXTER

In the text

$\begin{figure} \par\hspace*{2mm}\includegraphics[width=7.6cm,clip]{2967f13a.eps}\par\vspace*{3mm} \includegraphics[width=7.3cm,clip]{2967f13b.eps} \end{figure}$ Figure 13: A plot of [Ba/Eu] vs. [Fe/H] (upper panel) and [C/H] (lower panel). The horizontal full line shows the pure solar r-process value, computed from the solar r-process fractions of Arlandini et al. (1999), and the dashed line the solar s-process value. The pure r-II stars are shown as stars, the r-I stars as diamonds, the pure r-process stars without excess r-process elements as triangles, the s-II stars as asterisks, and the two remaining s-process-rich stars as squares. The average relative error bar is shown in the bottom left.
Open with DEXTER

In the text

$\begin{figure} \par\includegraphics[width=7.1cm,clip]{2967f14.eps} \end{figure}$ Figure 14: Histogram of [Eu/Fe] values. The dashed line includes all stars where Eu was detected, including the four comparison stars. The full line includes the pure r-process stars only (i.e. ${\rm [Ba/Eu]}<0$ noting Ba must be detected; for a small number of stars where Eu was detected, Ba was not detected due to cosmic ray pollution of the Ba II resonance line) and excludes the comparison stars.
Open with DEXTER

In the text

$\begin{figure} \par\includegraphics[width=4.7cm,clip]{2967f15a.eps}\hspace*{1.5m... ...hics[width=4.7cm,clip]{2967f15h.eps}\hspace*{1.5mm}\hspace*{4.7cm}\end{figure}$ Figure 15: Abundance patterns for neutron-capture elements in the r-II stars of the sample. The full line shows the scaled solar system r-process abundances based on the r-process fractions from Arlandini et al. (1999), meteoritic abundances from Grevesse & Sauval (1998), scaled to match our observed Eu abundance. Plotted error estimates are relative errors.
Open with DEXTER

In the text

$\begin{figure} \par\includegraphics[width=4.7cm,clip]{2967f16a.eps}\hspace*{2mm}... ....eps}\hspace*{2mm} \includegraphics[width=4.4cm,clip]{2967f16c.eps} \end{figure}$ Figure 16: Abundance patterns for neutron-capture elements in the s-process rich stars of the sample with strong Eu enhancement, ${\rm [Eu/Fe]}> 1.5$ , the s-II stars. The full line shows the scaled solar system r-process abundances based on the r-process fractions from Arlandini et al. (1999), meteoritic abundances from Grevesse & Sauval (1998), scaled to match our observed Eu abundance. The dashed line shows the equivalent scaled solar system s-process abundances, scaled to match our observed Ba abundance. Plotted error estimates are relative errors.
Open with DEXTER

In the text

$\begin{figure} \par\includegraphics[width=5cm,clip]{2967f17a.eps}\hspace*{2mm} \... ...egraphics[width=5cm,clip]{2967f17q.eps}\hspace*{2mm} \hspace*{5cm}\end{figure}$ Figure 17: Abundances ratios [X/Fe] plotted against [Fe/H] for elements with significant numbers of detections. Full lines show estimated $1\sigma$ scatter. The average relative error bars are shown in the bottom left. Note differing scales on the y-axes.
Open with DEXTER

In the text

$\begin{figure} \par\includegraphics[width=5cm,clip]{2967f18a.eps}\hspace*{2mm} \... ...egraphics[width=5cm,clip]{2967f18q.eps}\hspace*{2mm} \hspace*{5cm}\end{figure}$ Figure 18: Abundances ratios [X/Mg] plotted against [Mg/H] for elements with significant numbers of detections. Full lines show estimated $1\sigma$ scatter. The average relative error bars are shown in the bottom left. Note differing scales on the y-axes.
Open with DEXTER

In the text

$\begin{figure} \par\includegraphics[width=7.4cm,clip]{2967f19.eps} \end{figure}$ Figure 19: A plot of [C/H] versus [C/Fe] following Bromm & Loeb (2003). Giants and red horizontal branch stars are plotted as open circles, while the subgiants and dwarfs are plotted as filled circles.
Open with DEXTER

In the text

$\begin{figure} \par\includegraphics[width=7.6cm,clip]{2967f20.eps} \end{figure}$ Figure 20: A plot of [La/Eu] vs. [Fe/H]. Symbols and lines have the same meanings as in Fig. 13.
Open with DEXTER

In the text

$\begin{figure} \par\includegraphics[width=7.4cm,clip]{2967f21.eps} \end{figure}$ Figure 21: A plot of ${\rm [Ba/Fe]}_{r\rm -process}$ vs. [Fe/H]. Symbols have the same meanings as in Fig. 13. For the closed symbols (pure r-process stars) ${\rm [Ba/Fe]}_ {r\rm -process} = {\rm [Eu/Fe]}$ . Open circles show stars where ${\rm [C/Fe]}<0.3$ and ${\rm [Fe/H]}<-2.5$ , for which we assume ${\rm [Ba/Fe]}_{r \rm -process} = {\rm [Ba/Fe]+0.721}$ . Note the scatter is underestimated at ${\rm [Fe/H]} > -2.5$ ; see text.
Open with DEXTER

In the text

$\begin{figure} \par\includegraphics[width=7.4cm,clip]{2967f22a.eps}\par\vspace*{3mm} \includegraphics[width=7.1cm,clip]{2967f22b.eps} \end{figure}$ Figure 22: A plot of [Sr/Ba] vs. [Fe/H] ( upper panel) and [C/H] ( lower panel). Symbols and lines have the same meanings as in Fig. 13, with circles representing stars where Eu is undetected.
Open with DEXTER

In the text

$\begin{figure} \par\includegraphics[width=7.7cm,clip]{2967f23a.eps}\par\vspace*{3mm} \includegraphics[width=7.4cm,clip]{2967f23b.eps} \end{figure}$ Figure 23: A plot of [Sr/Eu] vs. [Fe/H] ( upper panel) and [C/H] ( lower panel). Symbols and lines have the same meanings as in Fig. 13, with circles representing stars where Ba is undetected. The solar s-process value is 1.08.
Open with DEXTER

In the text

$\begin{figure} \par\includegraphics[width=5.6cm,clip]{2967f24.eps} \end{figure}$ Figure 24: A plot of [Sr/Ba] vs. [Ba/Fe]. Symbols and lines have the same meanings as in Fig. 13, with circles representing stars where Eu is undetected.
Open with DEXTER

In the text

$\begin{figure} \par\includegraphics[width=5.9cm,clip]{2967f25.eps} \end{figure}$ Figure 25: A plot of [Sr/Eu] vs. [Eu/Fe]. Symbols and lines have the same meanings as in Fig. 23.
Open with DEXTER

In the text

$\begin{figure} \par\includegraphics[width=7.5cm,clip]{2967f26a.eps}\par\vspace*{3mm} \includegraphics[width=7.2cm,clip]{2967f26b.eps} \end{figure}$ Figure 26: A plot of [Y/Ba] vs. [Fe/H] ( upper panel) and [C/H] ( lower panel). Symbols and lines have the same meanings as in Fig. 13, with circles representing stars where Eu is undetected.
Open with DEXTER

In the text

$\begin{figure} \par\includegraphics[width=5.6cm,clip]{2967f27.eps} \end{figure}$ Figure 27: A plot of [Y/Ba] vs. [Ba/Fe]. Symbols and lines have the same meanings as in Fig. 26.
Open with DEXTER

In the text

$\begin{figure} \par\includegraphics[width=5.6cm,clip]{2967f28.eps} \end{figure}$ Figure 28: A plot of [Ba/Fe] vs. [C/Fe]. Symbols have the same meanings as in Fig. 22.
Open with DEXTER

In the text

$\begin{figure} \par\includegraphics[width=7.1cm,clip]{2967f1.eps} \end{figure}$	Figure 1: The $T_{\rm eff}$ versus $\log g$ Kiel diagram for the sample. The diameter of the circle for each star is proportional to [C/Fe]; cases where C is undetected are shown with an asterisk. The mean error bar is shown in the bottom right.
Open with DEXTER

$\begin{figure} \par\includegraphics[width=7.4cm,clip]{2967f2.eps} \end{figure}$	Figure 2: The distribution of analysed stars in [Fe/H]. The [Fe/H] values are from the final analysis of the spectra.
Open with DEXTER

$\begin{figure} \par\includegraphics[width=7.7cm,clip]{2967f3a.eps}\par\vspace{3mm} \hspace{2mm}\includegraphics[width=8cm,clip]{2967f3b.eps} \end{figure}$	Figure 3: Change in E(B-V) ( upper panel) and $T_{\rm eff}$ ( lower panel) if we adopt the reddening maps of Burstein & Heiles (1982) instead of those of Schlegel et al. (1998). The reddening corrections are usually reduced and thus the temperatures become somewhat cooler.
Open with DEXTER

$\begin{figure} \par\includegraphics[width=6.5cm,clip]{2967f4.eps} \end{figure}$	Figure 4: Histogram of the slopes of fits to the trends of Fe I line abundances with excitation for the sample.
Open with DEXTER

$\begin{figure} \par\includegraphics[width=7.5cm,clip]{2967f5a.eps}\par\vspace*{3mm} \includegraphics[width=7.5cm,clip]{2967f5b.eps} \end{figure}$	Figure 5: Difference between the final [Fe/H] as determined from the spectrum and the initial guess determined from medium-resolution follow-up spectroscopy, where $\Delta$ [Fe/H] = [Fe/H](initial)-[Fe/H](final), plotted against $T_{\rm eff}$ and [Fe/H](final).
Open with DEXTER

$\begin{figure} \par\includegraphics[width=7.4cm,clip]{2967f6.eps} \end{figure}$	Figure 6: Difference between the final $\log g$ as determined from the spectrum and the initial guess as described in the text, where $\Delta \log g = \log g {\rm (initial)} - \log g {\rm (final)}$ .
Open with DEXTER

$\begin{figure} \par\includegraphics[width=7.6cm,clip]{2967f7.eps} \end{figure}$	Figure 7: Plot of final microturbulence values $\xi$ as derived from the spectrum against gravity. Note that the stars lying higher than the general trend are the red horizontal branch stars (see Fig. 1).
Open with DEXTER

$\begin{figure} \par\includegraphics[width=7.4cm,clip]{2967f9a.eps}\hspace{3mm} ... ...mm} \includegraphics[width=7.4cm,clip]{2967f9e.eps}\hspace{7.2cm}\end{figure}$	Figure 9: Comparison of abundances for stars previously studied in the literature. Differences are in the sense of $\Delta \log \epsilon = \log \epsilon_{\rm this\; work} - \log \epsilon_{\rm literature}$ . Error bars indicate the estimated absolute error in our result. The mean difference and standard deviation $\sigma$ are reported.
Open with DEXTER

$\begin{figure} \par\includegraphics[width=7.4cm,clip]{2967f10a.eps}\includegraphics[width=7.4cm,clip]{2967f10b.eps} \end{figure}$	Figure 10: Comparison of abundances as in Fig. 9, where we have adopted the same temperature as the literature study.
Open with DEXTER

$\begin{figure} \par\includegraphics[width=7.4cm,clip]{2967f11.eps} \end{figure}$	Figure 11: Comparison of abundances as in Fig. 9, for HD 20 where we have adopted the same temperature, gravity and microturbulence as Burris et al. (2000).
Open with DEXTER

$\begin{figure} \par\includegraphics[width=4.7cm,clip]{2967f15a.eps}\hspace{1.5m... ...hics[width=4.7cm,clip]{2967f15h.eps}\hspace{1.5mm}\hspace*{4.7cm}\end{figure}$	Figure 15: Abundance patterns for neutron-capture elements in the r-II stars of the sample. The full line shows the scaled solar system r-process abundances based on the r-process fractions from Arlandini et al. (1999), meteoritic abundances from Grevesse & Sauval (1998), scaled to match our observed Eu abundance. Plotted error estimates are relative errors.
Open with DEXTER

$\begin{figure} \par\includegraphics[width=5cm,clip]{2967f17a.eps}\hspace{2mm} \... ...egraphics[width=5cm,clip]{2967f17q.eps}\hspace{2mm} \hspace*{5cm}\end{figure}$	Figure 17: Abundances ratios [X/Fe] plotted against [Fe/H] for elements with significant numbers of detections. Full lines show estimated $1\sigma$ scatter. The average relative error bars are shown in the bottom left. Note differing scales on the y-axes.
Open with DEXTER

$\begin{figure} \par\includegraphics[width=5cm,clip]{2967f18a.eps}\hspace{2mm} \... ...egraphics[width=5cm,clip]{2967f18q.eps}\hspace{2mm} \hspace*{5cm}\end{figure}$	Figure 18: Abundances ratios [X/Mg] plotted against [Mg/H] for elements with significant numbers of detections. Full lines show estimated $1\sigma$ scatter. The average relative error bars are shown in the bottom left. Note differing scales on the y-axes.
Open with DEXTER

$\begin{figure} \par\includegraphics[width=7.4cm,clip]{2967f19.eps} \end{figure}$	Figure 19: A plot of [C/H] versus [C/Fe] following Bromm & Loeb (2003). Giants and red horizontal branch stars are plotted as open circles, while the subgiants and dwarfs are plotted as filled circles.
Open with DEXTER

$\begin{figure} \par\includegraphics[width=7.6cm,clip]{2967f20.eps} \end{figure}$	Figure 20: A plot of [La/Eu] vs. [Fe/H]. Symbols and lines have the same meanings as in Fig. 13.
Open with DEXTER

$\begin{figure} \par\includegraphics[width=7.4cm,clip]{2967f22a.eps}\par\vspace*{3mm} \includegraphics[width=7.1cm,clip]{2967f22b.eps} \end{figure}$	Figure 22: A plot of [Sr/Ba] vs. [Fe/H] ( upper panel) and [C/H] ( lower panel). Symbols and lines have the same meanings as in Fig. 13, with circles representing stars where Eu is undetected.
Open with DEXTER

$\begin{figure} \par\includegraphics[width=7.7cm,clip]{2967f23a.eps}\par\vspace*{3mm} \includegraphics[width=7.4cm,clip]{2967f23b.eps} \end{figure}$	Figure 23: A plot of [Sr/Eu] vs. [Fe/H] ( upper panel) and [C/H] ( lower panel). Symbols and lines have the same meanings as in Fig. 13, with circles representing stars where Ba is undetected. The solar s-process value is 1.08.
Open with DEXTER

$\begin{figure} \par\includegraphics[width=5.6cm,clip]{2967f24.eps} \end{figure}$	Figure 24: A plot of [Sr/Ba] vs. [Ba/Fe]. Symbols and lines have the same meanings as in Fig. 13, with circles representing stars where Eu is undetected.
Open with DEXTER

$\begin{figure} \par\includegraphics[width=5.9cm,clip]{2967f25.eps} \end{figure}$	Figure 25: A plot of [Sr/Eu] vs. [Eu/Fe]. Symbols and lines have the same meanings as in Fig. 23.
Open with DEXTER

$\begin{figure} \par\includegraphics[width=7.5cm,clip]{2967f26a.eps}\par\vspace*{3mm} \includegraphics[width=7.2cm,clip]{2967f26b.eps} \end{figure}$	Figure 26: A plot of [Y/Ba] vs. [Fe/H] ( upper panel) and [C/H] ( lower panel). Symbols and lines have the same meanings as in Fig. 13, with circles representing stars where Eu is undetected.
Open with DEXTER

$\begin{figure} \par\includegraphics[width=5.6cm,clip]{2967f27.eps} \end{figure}$	Figure 27: A plot of [Y/Ba] vs. [Ba/Fe]. Symbols and lines have the same meanings as in Fig. 26.
Open with DEXTER

$\begin{figure} \par\includegraphics[width=5.6cm,clip]{2967f28.eps} \end{figure}$	Figure 28: A plot of [Ba/Fe] vs. [C/Fe]. Symbols have the same meanings as in Fig. 22.
Open with DEXTER