Central velocity dispersions, Mg_2 line indices and radial velocities for 220 E and S0 galaxies are derived on the basis of intermediate resolution spectroscopy. Galaxies in the following clusters have been observed: Abell 194, Abell 539, Abell 3381, Abell 3574, S639, S753, Doradus, HydraI (Abell 1060) and Grm 15. For 151 of the galaxies, the velocity dispersion has not previously been measured. 134 of the Mg_2 determinations are for galaxies with no previous measurement. The spectra cover either 500 or 1000A, centred on the magnesium triplet at 5177A. The observations were obtained with the Boller & Chivens spectrograph at the ESO 1.5-m telescope and with the OPTOPUS, a multi-object fibre-fed B&C spectrograph, at the ESO 3.6-m telescope. The data are part of our ongoing study of the large-scale motions in the Universe and the physical background for the Fundamental Plane. The Fourier fitting method was used to derive the velocity dispersions and radial velocities. The velocity dispersions have been corrected for the effect of the size of the aperture. The correction was established on the basis of velocity dispersion profiles available in the literature. A comparison with results from Davies et al. shows that the derived central velocity dispersions have an rms error of 0.036 in logsigma. There is no offset relative to the velocity dispersions from Davies et al. The offset relative to data from Lucey & Carter is -0.017+/-0.011 in logsigma, with our velocity dispersions being the smallest. The velocity dispersions derived from the B&C and the OPTOPUS observations, as well as the velocity dispersions published by Davies et al., Dressler, Lucey & Carter and Lucey et al., can be brought on a system consistent within 3 per cent. The Mg_2 line indices have been corrected for the size of the apertures, transformed to the Lick system, and corrected for the effect of the velocity dispersion. From comparison with data from Davies et al. and from Faber, we find that the rms error of Mg_2 is 0.013. Comparisons of the radial velocities with data from the literature show that our determinations are accurate to within ~=35 km s^-1. The accuracies reached for these observations are adequate for the study of the large-scale motions in the Universe and for investigations of the Fundamental Plane.