We re-analyse the velocity-dispersion profile \sigma ( r ) at radii r > 10 kpc in the Galactic stellar halo , recently derived by Battaglia et al .
( 1 ) , who concluded that , for a constant velocity anisotropy of the tracers , these data rule out a flat circular-speed curve for the Milky Way .
However , we demonstrate that if one makes the natural assumption that the tracer density is truncated at r _ { \mathrm { t } } \ga 160 kpc and falls off significantly more steeply than r ^ { -3.5 } at r \ga 80 kpc , then these data are consistent with a flat circular-speed curve and a constant velocity anisotropy comparable to that observed for halo stars in the Solar neighbourhood .
We also consider a more detailed mass model with an exponential stellar disc and an extended non-isothermal dark-matter halo .
In this two-component model , the Milky Way ’ s virial radius and mass are r _ { \mathrm { vir } } \simeq 200 kpc and M _ { \mathrm { vir } } \simeq 1.5 \times 10 ^ { 12 } { \mathrm { M } _ { \odot } } .
Still assuming the tracers ’ velocity anisotropy to be constant ( at \beta \simeq 0.5 ) we again find good agreement with the observed \sigma ( r ) , so long as the tracer density is truncated near the virial radius .
These data by themselves do not allow to differentiate between different dark-halo or total-mass models for the Milky Way , nor between different velocity-anisotropy profiles for the tracers .