2004MNRAS.355L..33M

In response to recent criticism by Momany et al., that the recently identified Canis Major (CMa) overdensity could be simply explained by the Galactic warp, we present proof of the existence of a stellar population in the direction of CMa that cannot be explained by known Galactic components. By analysing the radial distribution of counts of M-giant stars in this direction, we show that the Momany et al. warp model overestimates the number of stars in the Northern hemisphere, hence hiding the CMa feature in the South. The use of a better model of the warp has little influence on the morphology of the overdensity and clearly displays an excess of stars grouped at a distance of D= 7.2±0.3 kpc. To lend further support to the existence of a population that does not belong to the Galactic disc, we present radial velocities of M-giant stars in the centre of the CMa structure that were obtained with the 2dF spectrograph at the Anglo-Australian Telescope. The extra population shows a radial velocity of v_r= 109±4 km/s, which is significantly higher than the typical velocity of the disc at the distance of CMa. This population also has a low dispersion (13±4 km/s). The Canis Major overdensity is therefore highly unlikely to be due to the Galactic warp, adding weight to the hypothesis that we are observing a disrupting dwarf galaxy or its remnants. This leads to questions on what part of CMa was previously identified as the Warp and how possibly to disentangle the two structures.