Astron. Astrophys. 318, 416-428 (1997)

6. Conclusion

Outer disc cepheids indicate a flat or slightly decreasing rotation curve between R=10 and R=15 kpc, corresponding to km s^-1 for ( kpc, km/s) or km s^-1 for ( kpc, km/s), with an internal error of about 4 km s^-1 and possible systematic effects amounting to approximately 6 km s^-1 when summed quadratically. The problem of the increase at large radii, as seen for instance in the curve by Clemens (1985), that was difficult to understand dynamically and uncharacteristic of galaxies such as the Milky Way, seems to vanish.

The mismatch between the cepheid and HII region rotation curves may indicate either important non-axisymmetric motions in the gas that are smoothed out in the stellar velocity field, or high uncertainties and zero-point shift on the distances of HII regions. These two possibilities can be studied more closely in the future. In the first case, a gradual change from gas kinematics to stellar kinematics would be predicted in the cepheids as a function of their age, ranging from a few Myr to some Myr, as they move from gas to stellar kinematics. This could be observed with a sufficient number of outer disc cepheids. In the second case, new metallicity-dependent models, combined with CCD observations, could yield a more secure HII region distance scale.

In both cases, the cepheid rotation curve is a better indicator of the rotation of the stellar disc, and should be used for kinematical distance determinations. The outer disc velocity field derived from HII regions (see Brandt & Blitz 1993), including non-axisymmetric motions, is thus probably not an accurate indicator of the stellar velocity field.

No evidence for significant non-axisymmetric motions is found in our data. It is possible that either the triaxiality of the Galaxy is not as important as predicted by some models, or that the outer disc cepheids are locally not greatly affected.

Increasing the number of outer disc cepheids studied would obviously be desirable, in order to provide tighter constraints on the north-south asymmetry, the radial motion, and the rotation curve for R=15-16 kpc. Most known classical cepheids in the outer disc have been included in the sample. Completeness arguments alone indicate that a large number remain to be discovered. However the absorption in the plane of the disc, combined with the decreasing density, would make a search for new remote cepheids very ineffective in the visible. The infrared should be more promising, although the effort implied is still considerable. The area to be scanned covers several hundreds of square degrees, and the harvest may not be plentiful. Large infrared surveys such as DENIS may contribute to this effort.

ELODIE spectra obtained for this study can also be used to determine metallicities using cross-correlation techniques. This gives in-situ metallicities in the outer disc. We are now calibrating this method (Pont et al. 1995). A study of cepheid colours, combined with infrared observations, is also under way (Pont&Laney, in prep.), to study with more accuracy the effect of metallicity on intrinsic colours.

© European Southern Observatory (ESO) 1997

Online publication: July 8, 1998
helpdesk.link@springer.de