Recent observations of the dwarf elliptical galaxy Scl-dE1 (Sc22) in the Sculptor group of galaxies revealed an extended globular cluster (Scl-dE1 GC1), which exhibits an extremely large core radius of about 21.2 pc. The authors of the discovery pape
r speculated on whether this object could reside in its own dark matter halo and/or if it might have formed through the merging of two or more star clusters. In this paper, we present N-body simulations to explore thoroughly this particular formation scenario. We follow the merger of two star clusters within dark matter haloes of a range of masses (as well as in the absence of a dark matter halo). In order to obtain a remnant which resembles the observed extended star cluster, we find that the star formation efficiency has to be quite high (around 33 per cent) and the dark matter halo, if present at all, has to be of very low mass, i.e. raising the mass to light ratio of the object within the body of the stellar distribution by at most a factor of a few. We also find that expansion of a single star cluster following mass loss provides another viable formation path. Finally, we show that future measurements of the velocity dispersion of this system may be able to distinguish between the various scenarios we have explored.
We present new spectroscopic data for twenty six stars in the recently-discovered Canes Venatici I (CVnI) dwarf spheroidal galaxy. We use these data to investigate the recent claim of the presence of two dynamically inconsistent stellar populations i
n this system (Ibata et al., 2006). We do not find evidence for kinematically distinct populations in our sample and we are able to obtain a mass estimate for CVnI that is consistent with all available data, including previously published data. We discuss possible differences between our sample and the earlier data set and study the general detectability of sub-populations in small kinematic samples. We conclude that in the absence of supporting observational evidence (for example, metallicity gradients), sub-populations in small kinematic samples (typically fewer than 100 stars) should be treated with extreme caution, as their detection depends on multiple parameters and rarely produces a signal at the 3sigma confidence level. It is therefore essential to determine explicitly the statistical significance of any suggested sub-population.
