Astron. Astrophys. 325, 881-892 (1997)

1. Introduction

The large scale structure of the Universe, as derived from recent redshift surveys, reveals us large inhomogeneities: we see the galaxies forming clusters and superclusters, or even larger structures, filaments and sheets of galaxies, or walls, which surround the empty regions that are the voids. But this view is primary based on the high luminosity galaxies - the giants. There has been a long debate in the literature whether the giant galaxies are fair tracers of the large scale structure and whether the voids are really empty. From an observer point of view, the galaxy maps may reflect special observational selection effects in surface brightness, integral magnitude or diameter. It is a straightforward idea to think that going fainter we will eventually start to fill the voids. We know that the dwarf galaxies dominate the galaxy number in the Universe, but only very few are contained in the present galaxy catalogues. These objects can be very small, intrinsically faint and can also have very low surface brightness. The dwarf galaxies are thus good candidates to fill up the voids. There are also some theories of galaxy formation (Dekel & Silk 1986) that were worked out in the frame of the cold dark matter (CDM) scenarios including biasing, which predict that the dwarf galaxies should originate from the 1 fluctuations, expecting thus to be more evenly distributed than the high rare density peaks which form the giants. In these scenarios the dwarf galaxies should trace the underlying dark matter and are expected to fill the voids. It is therefore a fundamental task to search for the existence of a homogenous void population. Whether or not such a population exists would be of extreme importance for our understanding of galaxy formation and large scale structure formation and evolution.

Several studies of the spatial distribution of dwarf galaxies were carried out to answer these questions. Different projects were conceived to search for galaxies in voids or to see if the clustering properties of dwarfs galaxies are different from that of the normal giant galaxies. Most of the studies were based on redshift surveys of low-surface brightness dwarfs (Bothun et al. 1986, Thuan, Gott & Schneider 1987, Eder et al. 1989, Salzer, Hanson & Gavazzi 1990, Binggeli, Tarenghi & Sandage 1990, Weinberg et al. 1991, Thuan et al. 1991). The results of these studies did in general not favour biased galaxy formation and no void population was found. More recently, a new search for faint galaxies in voids was accomplished, with the intention to overcome the limitations of previous surveys in magnitude, diameter and surface brightness (Hopp 1994, Hopp et al. 1995, Kuhn et al. 1997). Other studies were based on emission-line objects (Tifft et al. 1986, Moody et al. 1987, Weistrop & Downs 1988, Weistrop 1989, Salzer 1989, Rosenberg et al. 1994) and there has been the suggestion that a few emission-line galaxies have been found in the voids. Much effort has concentrated on the study of the Bootes void, a huge low density region with a volume of , first discovered by Kirshner et al. (1981) (see also Kirshner et al. 1983 a,b). Up to now 58 galaxies were identified in the Bootes void, most of them coming from objective-prism surveys (Sanduleak & Pesch 1982, Tifft et al. 1986) but also from IRAS (Dey et al. 1990) and HI surveys (Szomoru et al. 1996a). The studies of the properties of these galaxies (Weistrop et al. 1995, Szomoru et al. 1996b) showed that they were strong emission-line objects with significant amount of star formation. They were also luminous galaxies, not the low mass, low surface brightness galaxies predicted to be found in voids. On the other hand it is important to notice that the Bootes void is beyond the distance at which the large scale structure is well defined by the present catalogues.

Having in mind that no definitive conclusion was drawn from previous studies and no faint void population was found in literature, we have undertaken a programm to search for dwarf galaxies in voids (Popescu et al. 1995, Popescu 1996). We especially selected nearby voids which were very well defined in the distribution of normal galaxies, in order to overcome some of the limitations of previous surveys like those in Bootes. Since there was a hint that emission-line galaxies were found in voids, we chose to search for emission-line objects, but with the aim of finding mainly dwarf HII galaxies or BCDs. The observational data of our survey as well as the location of our void regions are given in Popescu et al. (1996) (Paper 1). The candidates were selected on the objective prism plates taken from the Hamburg Quasar Survey (HQS)(Hagen et al. 1995) and therefore they have the name built with the prefix HS.

In this paper we will give the results of our search for galaxies in voids and we will describe the spatial distribution of the ELGs in comparison with the normal galaxies. The paper is organised as follows. In section 2 we estimate the completeness of our survey in terms of fluxes and equivalent widths. In section 3 we give a qualitative description of the spatial distribution of our ELGs, by using cone-diagrams in both Right-Ascension and Declination projection, and we also quantify the results by calculating the spatial densities and the nearest neighbour distributions. In section 4 we analyse the main properties of the galaxies we found in voids and in the last section we summarise and give the conclusions.

© European Southern Observatory (ESO) 1997

Online publication: March 26, 1998

helpdesk.link@springer.de