Based on observations obtained on Cerro Paranal, Chile, in ESO programs 65.L-0144 and 71.C-0206C.

.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.

While the measurement of the separation between the two object is very precise on the ISAAC aquisition image obtained in July 2003, we could not measure the position angle PA with high precision, see Table 1, because there are no other objects detected on the (short exposure) aquisition image, with whose positions we could have measured the actual detector orientation; hence, we have to rely on the header information, which we know to be less precise from previous experience; also, being just an aquisition image obtained in a service mode night, we do not have other images obtained at the same detector orientation of other fields, which could otherwise be used for such a calibration; hence, the large uncertainty in the third PA values in Table 1, while the separation is more precise being independant of the detector orientation.

.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.

... dwarfs

Because the object is close to the main sequence, it is much closer to true dwarfs than to objects intermediate between dwarfs and giants, so that we use the dwarf scale instead of Luhman's intermediate scale; for intermediate objects, the temperature would be $\sim$ 2635 K.

.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.

... dimension

In radial velocity searches, one can detect all companions, unless the orbital plane is perpendicular to the line-of-sight.

.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.