Kinematic constraints on the stellar and dark matter content of spiral and S0 galaxies

We present mass models of a sample of 14 spiral and 14 S0 galaxies that constrain their stellar and dark matter content. For each galaxy, we derive the stellar mass distribution from near-infrared photometry under the assumptions of axisymmetry and a constant K-S-band stellar mass-to-light ratio (M/L)K-S. To this we add a dark halo assumed to follow a spherically symmetric Navarro, Frenk and White profile and a correlation between concentration and dark mass within the virial radius, M-DM. We solve the Jeans equations for the corresponding potential under the assumption of constant anisotropy in the meridional plane, beta(z). By comparing the predicted second velocity moment to observed long-slit stellar kinematics, we determine the three best-fitting parameters of the model: (M/L)K-S, M-DM and beta(z). These simple axisymmetric Jeans models are able to accurately reproduce the wide range of observed stellar kinematics, which typically extend to approximate to 2-3R(e) or, equivalently, approximate to 0.5-1R(25). Although our sample contains barred galaxies, we argue a posteriori that the assumption of axisymmetry does not significantly bias our results. We find a median stellar mass-to-light ratio at K-S-band of 1.09(M/L)K-S,circle dot with an rms scatter of 0.31. We present preliminary comparisons between this large sample of dynamically determined stellar mass-to-light ratios and the predictions of stellar population models. The stellar population models predict slightly lower mass-to-light ratios than we measure. The mass models contain a median of 15 per cent dark matter by mass within an effective radius R-e (defined here as the semimajor axis of the ellipse containing half the K-S-band light) and 49 per cent within the optical radius R-25. Dark and stellar matter contribute equally to the mass within a sphere of radius 4.1R(e) or 1.0R(25). There is no evidence of any significant difference in the dark matter content of the spirals and S0s in our sample. Models without dark matter are also able to satisfactorily reproduce the observed kinematics in most cases. The improvement when a halo is added is statistically significant, however, and the stellar mass-to-light ratios of mass models with dark haloes match the independent expectations of stellar population models better.