Canopus B: A Candidate Common Proper Motion Companion to the Second Brightest Star
At present, the Washington Double Star catalog does not list any known companions to the 2nd brightest star in the sky - Canopus (Alpha Carinae). I found what appears to be a common proper motion companion to Canopus at wide separation, but comfortably within the projected tidal radius for Canopus (~2.9 pc) if the two stars are codistant. The projected separation for Canopus and the candidate companion "B" at epoch 2000.0 is 1.16 degrees (4169.7 arcseconds) at position angle PA = 181.18 degrees (measured north through east). The star lies at J2000 position 06:23:47.40 -53:51:13.2 and has X-ray counterparts in the ROSAT All-Sky Survey Faint Source Catalog (1RXS J062347.1-535111) and in pointed ROSAT observation catalogs (2RXP J062346.0-535113, 1WGA J0623.7-5351). The ratio of the star's X-ray flux to its bolometric flux is consistent with "saturated" X-ray emission (observed log(Lx/Lbol) = -2.8) indicating a very magnetically active star, which would be unsurprising if the star is coeval with Canopus (~27 Myr). Also, if the star is codistant with Canopus, then its estimated X-ray luminosity based off of the ROSAT WGACat data (log(Lx/erg/s) = 29.45) is similar to that of ~10^7 yr-old M dwarfs (see Fig. 14 of Stelzer et al. 2013), again commensurate with the age of Canopus. If we assume that the companion is coeval with the Beta Pictoris Moving Group (~20 Myr), then its color (V-Ks=5.48) and magnitude (V=15.09, Ks=9.58) would be consistent with a photometric distance of ~52 pc. However, Canopus has a trigonometric parallax distance of 95+-5 pc (revised Hipparcos parallax 10.55+-0.56 mas), so "Canopus B" would appear to be somewhat too bright (~1 mag) if it were a single ~M4.5 pre-main sequence star at the same distance and age as Canopus. It is, of course, possible that "Canopus B" is itself a multiple star, and the unresolved multiplicity is causing it to be somewhat overluminous for its age. However, it is also possible that the age of Canopus itself is in error, or the revised Hipparcos parallax is too small.
While very young (~20 Myr), the Canopus system does not appear to kinematically belong to any known young associations or clusters. Using the revised Hipparcos astrometry from van Leeuwen (2007) and the radial velocity from Gontcharov (2006), I estimate a Galactic velocity of U, V, W = -11.7, -21.5, 1.2 (+-0.6, 0.6, 0.6) km/s. While its U and V velocity components are most similar to that of the Columba, Carina, Tuc-Hor, and Alessi 13 groups, its W velocity is somewhat anomalous compared to these groups. Hence, the association that spawned the Canopus system remains unidentified and may have dissolved into the Galactic field population at a very young age. According to the XHIP catalog (Andersen & Francis 2012), Canopus is the 2nd most luminous star within 100 pc (~13,200 Lsun), outshone only by Alpha Crucis (Acrux; ~30,300 Lsun).
A few more comments are in order regarding whether or not Canopus or 2MASS J06234738-5351131 (star "B") could be members of the Tuc-Hor group. If "B" is comoving with Canopus, its predicted RV is 20.1 km/s. Unfortunately, this is fairly similar to that predicted for a Tuc group member at the star's position. Using the UVW velocity vector for Tucana from Kraus+ (2014), one would predict an RV for an "ideal" Tuc member at the star's position to be 20.9 km/s (and the group has +-1.1 km/s intrinsic velocity dispersion). However, if the star is a Tucana member, then its proper motion is consistent with a kinematic distance estimate of 67+-6 pc (plx = 14.9+-1.3 mas). If Canopus itself were a Tuc member, then its proper motion is consistent with a kinematic distance of 73+-4 pc (and the RV is nearly spot-on; predicted RV = 21.1 km/s). Hence, while an RV measurement may have trouble distinguishing whether the star is comoving with either Canopus or the Tuc group, a sub-milliarcsecond precision parallax may be able to help. It is worth noting that the star is not near the concentration of known Tucana members (see Kraus et al. 2014 Figure 2) - however given how poorly characterized the Tuc group still is, I doubt anyone would be surprised to find a Tuc member in the direction of Canopus. For Canopus itself to be a Tuc member, its revised Hipparcos parallax would have to be in error by about -3 mas or ~3 sigma (i.e. predicted parallax of 13.66+-0.75 mas vs. observed parallax of 10.55+-0.56 mas parallax measured).
Thanks to Adam Kraus for discussions on the Tucana group.