1 Introduction

It was first suggested by Bok & Reilly (1947) that small, compact and isolated dark regions in the sky (now called "Bok Globules'') may be undergoing gravitational collapse and may eventually form low mass stars. Barnard (1927) had earlier prepared a list of such dark regions in the sky. Lynds (1962) later published a catalogue, with a larger number of such dark objects. Bok (1956) himself studied many such clouds. Subsequently the importance of these clouds as possible sites for star formation were felt by many workers (papers in Lynds 1971; Villere & Black 1980; Leung et al. 1982; Vrba et al. 1981; Keene et al. 1983; Joshi et al. 1985 etc.). Recently Clemens & Barvainis (1988) (henceforth CB) have compiled a list of 248 small (mean size $\sim 4'$) and nearby (distance < 1 kpc) molecular clouds.

Bok's speculations got firmly established through Infra-red (IRAS survey) and millimeter studies (¹²CO map) on these clouds (Keene et al. 1983; Goldsmith et al. 1984; Yun & Clemens 1990; Clemens et al. 1991; Yun & Clemens 1992 etc.). These clouds undergo gravitational collapse and magnetic force plays a key role in collapse dynamics by mediating accretion, directing the outflows and collimating the jets. The determination of the strength and geometry of the magnetic field in these clouds help to understand the dynamics of the star formation processes (Goodman et al. 1989; Myers & Goodman 1991; Kane et al. 1995 etc.). However, it is not very easy to determine magnetic field geometry in a cloud. When light from a star passes through the interstellar (henceforth IS) medium it gets linearly polarized due to forward scattering by the dichroic IS grains. The IS magnetic field is believed to be responsible for the alignment of these dichroic grains (Davis & Greenstein 1951).

In the past there have been many studies of polarization with a view to trace the geometry of the magnetic field in dark clouds. These studies have shown that the relation between the extinction and polarization of the background star is far from unique and depends on the environment of the cloud. Thus although quite large values of polarization could be associated with large extinction (Hodapp 1987; Zaritsky et al. 1987; Jones 1989), the polarization efficiency appears to be low for dark clouds (Jones et al. 1984; Klebe & Jones 1990). It was also found that the polarization did not increase linearly with extinction (Jones 1989; Jarrett et al. 1994). Jones (1989) and Goodman et al. (1990) suggested that a combination of regular and random magnetic fields could be responsible for the lack of linear increase in polarization with extinction. Myers & Goodman (1991) and Jones et al. (1992) modeled the magnetic fields in clouds to explain the polarization-extinction relation. On the other hand, Goodman et al. (1995) and Creese et al. (1995) have argued that the grains in dark clouds are poorly aligned and contribute very little to the polarization. These conclusions have been confirmed by theoretical modeling by Lazarian et al. (1997) and further observation by Arce et al. (1998).

There is a need to study further: (i) the relation between physical environment in dark clouds and alignment of grains and (ii) alignment mechanism for the polarization vectors and its connection the ambient magnetic fields in the clouds. With these aims we have carried out imaging polarimetric observations of eight clouds selected from the catalogue by CB. From our observations we construct the polarization maps and report our results with some preliminary discussions.