1 Introduction

The majority of spiral galaxies have a warped plane, as has been revealed in the neutral gas extended component, through HI-21 cm observations (e.g. Bosma 1981; Briggs 1990), and in a lesser extent through optical observations (Sanchez-Saavedra et al. 1990; Reshetnikov & Combes 1998). This dynamical feature raises the problem of its origin and maintenance, and the numerous mechanisms that have been proposed and explored have not yet given a definitive and satisfactory answer (e.g. the review by Binney 1992).

Differential precession should be very quick to wrap up any warp perturbation even in the outer parts of the galaxies (Kahn & Woltjer 1959), unless the potential is nearly spherical (Tubbs & Sanders 1979). But most warps are observed while the disk is still a significant part of the potential, which cannot therfore be spherical. It has been shown that coherent bending modes cannot be sustained, since the oscillations spectrum is continuous, for realistic disks that have no sharp edges (Hunter & Toomre 1969). Models then tried to consider a non-spherical dark halo, misaligned with the inner visible disk of the galaxy (Sparke 1984; Sparke & Casertano 1988; Dubinski & Kuijken 1995). However, these structures can only be transient, since the inner disk is bound to align with the dark halo (New et al. 1998; Binney et al. 1998). Alternatively, the warp could be the consequence of continuous accretion of gas with a slewed angular momentum, due to cosmic infall, as suggested by Ostriker & Binney (1989) and Binney (1992). It is not excluded either that a large part of warps are due to interactions or mergers: the prototypical warped galaxy NGC 5907 (Sancisi 1976) that was long thought isolated, might have experienced a minor merger recently (Lequeux et al. 1998), and is currently interacting with two dwarf companions (Shang et al. 1998).

To progress about the puzzle of the origin of warps, it is important to have a sample of optically strongly warped galaxies, to perform new observations and statistical studies. Recently, we have presented a survey of optical warps in a sample of 540 galaxies, about 5 times larger than the previous samples (Reshetnikov & Combes 1998). The galaxies were selected from the Flat Galaxy Catalogue of Karachentsev et al. (1993) (FGC) and we studied their optical images extracted from the Digitized Sky Surveys. We identified three classes of galaxies, those without observable warps (30%), and those with U-shaped (37%) and S-shaped (33%) warps. We have considered the artefacts due to projection effects, that could be severe in nearly edge-on galaxies, when there are spiral arms or m=2 perturbations. Through numerical simulations, it was found that the U-shape are more affected by projection effects, but that no more than 15% of S-shape warps could be geometrical artefacts. On the other hand, intrinsic warps could be missed through projection effects (but no more than 20%).

We therefore select a sample of 60 S-shape warped galaxies, the strongest and clearest among the 174 found. The selection is subjective, based on isophotal maps from the DSS. This sample should be a suitable material for future detailed HI and optical works on galaxy warps.