We analyze deep near-IR adaptive optics imaging (taken with NAOS/CONICA on the VLT) 1 as well as new proper motion data of the nuclear star cluster of the Milky Way. The surface density distribution of faint (H≤ 20, K s ≤ 19) stars peaks within 0.2 ′′ of the black hole candidate SgrA ⋆ . The radial density distribution of this stellar 'cusp' follows a power law of exponent α ∼ 1.3 − 1.4. The K-band luminosity function of the overall nuclear stellar cluster (within 9 ′′ of SgrA ⋆ ) resembles that of the large scale, Galactic bulge, but shows an excess of stars at K s ≤ 14. It fits population synthesis models of an old, metal rich stellar population with a contribution from young, early and late-type stars at the bright end. In contrast, the cusp within ≤ 1.5 ′′ of SgrA ⋆ appears to have a featureless luminosity function, suggesting that old, low mass horizontal branch/red clump stars are lacking. Likewise there appear to be fewer late type giants. The innermost cusp also contains a group of moderately bright, early type stars that are tightly bound to the black hole. We interpret these results as evidence that the stellar properties change significantly from the outer cluster (≥ a few arcsecs) to the dense innermost region around the black hole.We find that most of the massive early type stars at distances 1-10" from SgrA ⋆ are located in two rotating and geometrically thin disks. These disks are inclined at large angles and counter-rotate with respect to each other. Their stellar content is essentially the same, indicating that they formed at the same time. We conclude that of the possible formation scenarios for these massive stars the most probable one is that 5-8 million years ago two clouds fell into the center, collided, were shock compressed and then formed two rotating (accretion) disks orbiting the central black hole. For the OB-stars in the central arcsecond, on the other hand, a stellar merger model is the most appealing explanation. These stars may thus be 'super-blue-stragglers', formed and 'rejuvenated' through mergers of lower mass stars in the very dense (≥ 10 8 M ⊙ pc −3 ) environment of the cusp. The 'collider model' also accounts for the lack of giants within the central few arcseconds.The star closest to SgrA ⋆ in 2002, S2, exhibits a 3.8 µm excess. We propose that the mid-IR emission either comes from the accretion flow around the black hole itself, or from dust in the accretion flow that is heated by the ultra-violet emission of S2.1 Based on observations obtained at the European Southern Observatory, Chile

Many galaxies are thought to have supermassive black holes at their centres-more than a million times the mass of the Sun. Measurements of stellar velocities and the discovery of variable X-ray emission have provided strong evidence in favour of such a black hole at the centre of the Milky Way, but have hitherto been unable to rule out conclusively the presence of alternative concentrations of mass. Here we report ten years of high-resolution astrometric imaging that allows us to trace two-thirds of the orbit of the star currently closest to the compact radio source (and massive black-hole candidate) Sagittarius A*. The observations, which include both pericentre and apocentre passages, show that the star is on a bound, highly elliptical keplerian orbit around Sgr A*, with an orbital period of 15.2 years and a pericentre distance of only 17 light hours. The orbit with the best fit to the observations requires a central point mass of (3.7 +/- 1.5) x 10(6) solar masses (M(*)). The data no longer allow for a central mass composed of a dense cluster of dark stellar objects or a ball of massive, degenerate fermions.

We surveyed all stars in Taurus (3h 45m < α < 4h 15m, 15° < δ < 35°) for multiplicity which are contained in the Herbig-Bell catalogue of young stars and have a 2 micron brightness of K ≤ 9.5 mag. This sample consists of 106 stellar systems (single or multiple), of which 43 are double or multiple according to the criteria of our survey, i.e. with separations of ≈0″.2 ≤ d ≤ 10″. Of these, 23 binaries are new detections found in this survey. The resulting degree of multiplicity, 43/106 = 41±6%, is higher than found for main-sequence stars. Provided that the period distribution is the same for young stars as on the main sequence, our result implies that the vast majority of stars are born in binary or multiple systems.

Mass is the most fundamental parameter of a star, yet it is also one of the most difficult to measure directly. In general, astronomers estimate stellar masses by determining the luminosity and using the 'mass-luminosity' relationship, but this relationship has never been accurately calibrated for young, low-mass stars and brown dwarfs. Masses for these low-mass objects are therefore constrained only by theoretical models. A new high-contrast adaptive optics camera enabled the discovery of a young (50 million years) companion only 0.156 arcseconds (2.3 au) from the more luminous (> 120 times brighter) star AB Doradus A. Here we report a dynamical determination of the mass of the newly resolved low-mass companion AB Dor C, whose mass is 0.090 +/- 0.005 solar masses. Given its measured 1-2-micrometre luminosity, we have found that the standard mass-luminosity relations overestimate the near-infrared luminosity of such objects by about a factor of approximately 2.5 at young ages. The young, cool objects hitherto thought to be substellar in mass are therefore about twice as massive, which means that the frequency of brown dwarfs and planetary mass objects in young stellar clusters has been overestimated.

We present a comprehensive data description for Ks-band measurements of Sgr A*. We characterize the statistical properties of the variability of Sgr A* in the near-infrared, which we find to be consistent with a single-state process forming a power-law distribution of the flux density. We discover a linear rms-flux relation for the flux-density range up to 12 mJy on a timescale of 24 minutes. This and the power-law flux density distribution implies a phenomenological, formally non-linear statistical variability model with which we can simulate the observed variability and extrapolate its behavior to higher flux levels and longer timescales. We present reasons why data with our cadence cannot be used to decide on the question whether the power spectral density of the underlying random process shows more structure at timescales between 25 min and 100 min compared to what is expected from a red noise random process.