2. OBSERVED PROPERTIES

The global observational understanding of galaxy warps has been summarized in two pairs of Laws.

• Bosma's Laws (Bosma 1991, on overall statistics of warps)
  1. At least half of all galaxies are warped
     Implication: Warps are long-lived or continuously generated
  2. Galaxies with small dark halo core radii (as determined from a rotation curve decomposition) are less likely to be warped
     Implication: Link between warps and the dark halo potential
• Briggs's Laws (Briggs 1990, on structure of individual warps)
  1. Disks are generally flat inside radius R₂₅. Out to radius R_26.5 the line of nodes of a warp is straight
     Implication: Self-gravity of the disk is important (it keeps the different parts of the disk precessing synchronously and hence the line of nodes straight - cf. the winding problem of spiral waves)
  2. The outer line of nodes advances in the direction of galactic rotation
     Implication: Warps are not quite in equilibrium at large radii. This points to a link to the environment, or to very long timescales

Warps are primarily observed in HI. Significant warps (misalignments between inner and outer parts of the disk of more than a degree or two) in stellar disks are rarely observed (Reshetnikov & Combes 1999), though recently the warp of the Milky Way has also been seen in the stellar distribution (Alard 2000). This may indicate that warps are a phenomenon which affects only the cold ISM, or that only the very outskirts of galaxy disks, which are only observable in HI, are involved.

The Galaxy warp illustrates the general point that, analogously to spiral structure, warps come in several varieties: `grand-design', nice integral-sign shaped bi-symmetric warps; `irregular' warps which are only visible on one side, or in which one side of the galaxy is more warped than the other; and `feeble' (weak or absent) warps. To determine the relative frequency of these classes is important, as it points the way for attempts at an explanation: should we be looking for mechanisms which naturally produce beautiful integral-sign warps, or for - possibily more chaotic - ways to make irregular ones?

In order to address this question, we have performed a blind survey of edge-on (as judged from optical images) galaxies. These galaxies form part of the WHISP sample, which is basically an HI flux-density limited sample (> 200mJy) selected from the UGC catalogue. We chose galaxies with blue major diameters larger than 2', and inclination class 6 or 7. DSS images of these galaxies were inspected to filter out obviously less-inclined galaxies, to leave us with a sample of galaxies which, as judged from optical images, have inclinations at least ~ 80°. We chose edge-on galaxies for this survey so that the warps can be studied purely morphologically, without the need for interpretation or modelling of the velocity field (e.g., by means of tilted ring models).

Our sample clearly shows that warps occur in all types. About 1/3 of our galaxies show a nice integral-sign warp, about 1/3 are flat, and the final 3rd of the sample are of the irregular (one-sided or asymmetric) types (Table 1).