Wide-Field Survey of Globular Clusters in M31. II. Kinematics of the Globular Cluster System

We present a kinematic analysis of the globular cluster(GC) system in M31. Using the photometric and spectroscopic database of 504 GCs, we have investigated the kinematics of the M31 GC system. We find that the all GC system shows strong rotation, with rotation amplitude of v_rot~190km/s, and that a weak rotation persists even for the outermost samples at |Y|>5kpc. The rotation-corrected velocity dispersion for the GC system is estimated to be sigma_{p,r}~130km/s, and it increases from sigma_{p,r}~120km/s at |Y|<1kpc to sigma_{p,r}~150km/s at |Y|>5kpc. These results are very similar to those for the metal-poor GCs. This shows that there is a dynamically hot halo in M31 that is rotating but primarily pressure-supported. We have identified 50 "friendless" GCs, and they appear to rotate around the major axis of M31. For the subsamples of metal-poor and metal-rich GCs, we have found that the metal-rich GCs are more centrally concentrated than the metal-poor GCs, and both subsamples show strong rotation. For the subsamples of bright and faint GCs, it is found that the rotation for the faint GCs is stronger than that for the bright GCs. We have identified 56 GCs and GC candidates with X-ray detection. It is found that the majority of X-ray emitting GCs follow the disk rotation, and that the redder, more metal-rich, and brighter GCs are more likely to be detected as X-ray emitting GCs. We have derived a rotation curve of M31 using the GCs at |Y|<0.6kpc. We have estimated the dynamical mass of M31 using `Projected Mass Estimator(PME)' and `Tracer Mass Estimator(TME)'. We finally discuss the implication of these results and compare the kinematics of GCs with that of planetary nebulae in M31.