Astron. Astrophys. 328, 5-11 (1997)

1. Introduction

The abundance and nature of dark matter in the halo of our Galaxy is rapidly making the transition from theoretical hypothesis to observational science. This has been facilitated by the deep surveys that are now achievable with instruments such as the Hubble Space Telescope (HST) and by several gravitational microlensing searches that are currently in progress.

The 2nd-year results from the MACHO microlensing experiment (Alcock et al. 1997) towards the Large Magellanic Cloud (LMC), a direction which is sensitive to lenses residing in the dark halo, indicate that a substantial fraction of the halo () comprises objects with a typical mass in the range . ¹ These results appear to be broadly supported by the provisional findings from 4 years of MACHO observations (Axelrod 1997) which have uncovered at least 14 LMC microlensing candidates. Similar mass scales have also been implicated by the EROS I microlensing experiment (Renault et al. 1997), though with a somewhat lower inferred halo fraction. These results are consistent with the lenses being in the form of low-mass hydrogen-burning stars or white-dwarf remnants.

However, both of these candidates appear unattractive when other observational and theoretical results are taken into consideration. The number density, age and mass function of white dwarfs in the Galaxy is strongly constrained by number counts of high-velocity dwarfs in the Solar neighbourhood, and by their helium production (Carr et al. 1984; Ryu et al. 1990; Adams & Laughlin 1996; Chabrier et al. 1996; Graff et al. 1997). In particular, Chabrier et al. (1996) find that a halo fraction compatible with MACHO results requires that white dwarfs be older than 18 Gyr, though more recently Graff et al. (1997) have argued for a lower limit closer to 15.5 Gyr based upon reasonable white-dwarf model assumptions and a halo fraction of 30%. The situation for low-mass stars appears at least as pessimistic with recent HST results indicating that a smoothly distributed population of low-mass stars can contribute no more than a few percent to the halo dark matter density, regardless of stellar metallicity (Bahcall et al. 1994; Graff & Freese 1996; Flynn et al. 1996; Kerins 1997).

It has been suggested (Kerins 1997, hereafter Paper I) that if low-mass stars are clumped into globular-cluster configurations then HST limits can be considerably weakened, since this introduces large fluctuations in number counts and also may prevent a significant fraction of sources within the cores of clusters from being resolved. Motivation for the cluster scenario comes from the predictions of some baryonic dark matter formation theories, which are discussed in Paper I. However, such clusters are required to have masses and radii consistent with existing dynamical constraints on clusters and other massive objects residing in the halo. In Paper I it was shown that agreement between HST counts, dynamical limits and the central value for the halo fraction inferred by MACHO (40% for the halo model assumed) is possible if clusters have a mass around and radius of a few parsecs. However, HST, MACHO and dynamical limits are all dependent upon the unknown halo distribution function, so these results are valid only for the spherically-symmetric, cored isothermal halo model adopted in Paper I.

In this paper the model dependency of such conclusions is investigated using the same set of reference halo models employed in the MACHO collaboration's analysis of its results. One of these models is similar, though not identical, to the model investigated in Paper I, whilst the other models are constructed from the self-consistent set of `power-law' halo models presented by Evans (1994). All models are normalised to be consistent with observational constraints on the Galactic rotation curve and local column surface density. New data from the Hubble Deep Field (Flynn et al. 1996) and Groth Strip (Gould et al. 1997) are also incorporated, as well as two other new fields analysed by Gould et al., extending the analysis from 20 HST fields in Paper I to 51 in this study.

© European Southern Observatory (ESO) 1997

Online publication: March 24, 1998

helpdesk.link@springer.de