Galaxy masses can be estimated by a variety of methods ; each applicable in different circumstances , and each suffering from different systematic uncertainties .
Confirmation of results obtained by one technique with analysis by another is particularly important .
Recent SDSS weak lensing measurements of the projected-mass correlation function reveal a linear relation between galaxy luminosities and the depth of their dark matter halos ( measured on 260 h ^ { -1 } kpc scales ) .
In this work we use an entirely independent dynamical method to confirm these results .
We begin by assembling a sample of 618 relatively isolated host galaxies , surrounded by a total of 1225 substantially fainter satellites .
We observe the mean dynamical effect of these hosts on the motions of their satellites by assembling velocity difference histograms .
Dividing the sample by host properties , we find significant variations in satellite velocity dispersion with host luminosity .
We quantify these variations using a simple dynamical model , measuring M _ { 260 } ^ { dyn } a dynamical mass within 260 h ^ { -1 } kpc .
The appropriateness of this mass reconstruction is checked by conducting a similar analysis within an N-body simulation .
Comparison between the dynamical and lensing mass-to-light scalings shows reasonable agreement , providing some quantitative confirmation for the lensing results .