4 The colour-magnitude diagram

The M_V vs. (B-V)₀ diagram for the giant members or probable members of NGC2354 is plotted in Fig. 3. By adopting the true distance modulus (m-M)₀ = 10.80 from Ahumada & Lapasset (1996) and our mean E(B-V) colour excess, the age of NGC2354 may be estimated by fitting theoretical isochrones computed by Bertelli et al. (1994), which include mass loss and moderate overshooting. In Fig. 3 we plotted the isochrones for Y = 0.25, Z = 0.008, equivalent to [Fe/H] = -0.4, and $\log t$ = 8.8, 9.0 and 9.2. A reasonable fit has been found for the isochrone of $\log t$ = 9.0 both for the Bertelli et al. (1994) and Schaller et al. (1992) isochrones. A better age determination is awaiting more precise CCD data to replace the old, photographic data of Dürbeck (1960) and the determination of membership of main-sequence stars. Neither set is able to reproduce correctly the distribution of all red giants, the Italian isochrones being too blue and the Swiss ones too red. The peculiar appearance of the red giant clump is partly due to the position of the binaries: binarity has a tendency to move points towards bluer B-V colours. A photometric decomposition of the two most pronounced binaries (#113, 179) has been attempted. The primary of #179 falls with the other single clump stars, while that for #113 falls close to #183 and would define the ascending giant branch.

From this, we could conclude that the shape of the isochrone is in good agreement with the overall distribution of the member red giants when the effect of binarity on the joined colours is properly taken into account. However, the positions of the four stars (#59, 152, 184 and 200) is quite puzzling. If the case of #184 may be explained by differential reddening as mentioned above, no convincing explanation have been found for the three others. The agreement of the radial velocities with the cluster mean ($V_{\rm r} = 33.4$ km s^-1), in particular that for #152 (33.84 km s^-1) and #200 (33.08 km s^-1), cannot be fortuitous. Red giants with colours redder that those of the bottom of the ascending red giant branch have been found in NGC 6940 (Mermilliod & Mayor 1989) and in NGC 2360 (Mermilliod & Mayor 1990), but both stars are binaries, which does not seem to be the case of stars #59 and 152. Five stars in an anomalous position with respect to the isochrone have been identified in the somewhat older cluster NGC 752 (Mermilliod et al. 1998), but they are bluer than the isochrone, not redder.

In many photometric analysis of the colour-magnitude diagrams, stars located outside the regular sequences are generally rejected, which gives the impression that sequences are clean. Independent kinematical data sometimes support the membership of stars located in an anomalous position. Clearly, additional observations are necessary before one can conclude that the four anomalous stars in NGC 2354 are really non-members. Meanwhile, the observed distribution of the red giants cannot be reproduced by any available model.

  

Figure 3: Mv vs. (B-V)0 diagram for NGC2354. Confirmed or possible red giant members and red field stars are represented by filled and open circles, respectively. Spectroscopic binaries (underlined) are indicated. Dots represent photographic data from Dürbeck (1960). Isochrones corresponding to $\log t$ = 8.8, 9.0 and 9.2 (Bertelli et al. 1994) are plotted

Acknowledgements

We are grateful to the referee, Dr J. Andersen, for his valuable remarks. We would like to thank ESO, CTIO and CASLEO for the generous allocation of observing time. Thanks are also due to M.A. Nicotra for his help with the software support. This work was partially supported by grants from the Argentinian institutions CONICET and CONICOR. The radial velocity program has been supported by continuous grants from the Swiss National Foundation for Scientific Research (FNRS).