We present a Mass-Luminosity Relation (MLR) for red dwarfs spanning a range of masses from 0.62 M to the end of the stellar main sequence at 0.08 M . The relation is based on 47 stars for which dynamical masses have been determined, primarily using astrometric data from Fine Guidance Sensors (FGS) 3 and 1r, white-light interferometers on the Hubble Space Telescope (HST), and radial velocity data from McDonald Observatory. For our HST/FGS sample of 15 binaries, component mass errors range from 0.4% to 4.0% with a median error of 1.8%. With these and masses from other sources, we construct a V -band MLR for the lower main sequence with 47 stars, and a K-band MLR with 45 stars with fit residuals half of those of the V -band.We use GJ 831 AB as an example, obtaining an absolute trigonometric parallax, π abs = 125.3 ± 0.3 milliseconds of arc, with orbital elements yielding M A = 0.270 ± 0.004M and M B = 0.145 ± 0.002M . The mass precision rivals that derived for eclipsing binaries.

The empirical mass-luminosity relation at is presented for stars with masses 0.08È0.20 based M V M _ upon new observations made with Fine Guidance Sensor 3 on the Hubble Space T elescope. The targets are nearby, red dwarf multiple systems in which the magnitude di †erences are typically measured tô 0.1 mag or better. The values are generated using the best available parallaxes and are also accu-M V rate to^0.1 mag, because the errors in the magnitude di †erences are the dominant error source. In several cases this is the Ðrst time the observed sub-arcsecond multiples have been resolved at optical wavelengths. The mass-luminosity relation deÐned by these data reaches to and provides a M V \ 18.5 powerful empirical test for discriminating the lowest mass stars from high-mass brown dwarfs at wavelengths shorter than 1 km.

We report on Keck Interferometer observations of the double-lined binary (B) component of the quadruple pre-main sequence (PMS) system HD 98800. With these interferometric observations combined with astrometric measurements made by the Hubble Space Telescope Fine Guidance Sensors (FGS), and published radial velocity observations we have estimated preliminary visual and physical orbits of the HD 98800 B subsystem. Our orbit model calls for an inclination of 66.8 ± 3.2 deg, and allows us to infer the masses and luminosities of the individual components. In particular we find component masses of 0.699 ± 0.064 and 0.582 ± 0.051 M ⊙ for the Ba (primary) and Bb (secondary) components respectively.Spectral energy distribution (SED) modeling of the B subsystem suggests that the B circumstellar material is a source of extinction along the line of sight to the B components. This seems to corroborate a conjecture by Tokovinin that the B subsystem is viewed through circumbinary material, but it raises important questions about the morphology of that circumbinary material.Our modeling of the subsystem component SEDs finds temperatures and luminosities in agreement with previous studies, and coupled with the component mass estimates allows for comparison with PMS models in the low-mass regime with few empirical constraints. Solar abundance models seem to under-predict the inferred component temperatures and luminosities, while assuming slightly sub-solar abundances bring the models and observations into better agreement. The present preliminary orbit does not yet place significant constraints on existing pre-main sequence stellar models, but prospects for additional observations improving the orbit model and component parameters are very good.

We present an absolute parallax and relative proper motion for the fundamental distance scale calibrator, δ Cep. We obtain these with astrometric data from FGS 3, a white-light interferometer on HST. Utilizing spectrophotometric estimates of the absolute parallaxes of our astrometric reference stars and constraining δ Cep and reference star HD 213307 to belong to the same association (Cep OB6, de Zeeuw et al. 1999), we find π abs = 3.66 ± 0.15 mas. The larger than typical astrometric residuals for the nearby astrometric reference star HD 213307 are found to satisfy Keplerian motion with P = 1.07 ± 0.02 years, a perturbation and period that could be due to a F0V companion ∼ 7 mas distant from and ∼4 magnitudes fainter than the primary. Spectral classifications and VRIJHKT 2 M and DDO51 photometry of the astrometric reference frame surrounding δ Cep indicate that field extinction is high and variable along this line of sight. However the extinction suffered by the reference star nearest (in angular separation and distance) to δ Cep, HD 213307, is lower and nearly the same as for δ Cep. Correcting for color differences, we find = 0.23 ± 0.03 for δ Cep, hence, an absolute magnitude M V = −3.47 ± 0.10. Adopting an average V magnitude, = 15.03 ± 0.03, for Cepheids with log P = 0.73 in the LMC from Udalski et al. (1999), we find a V-band distance modulus for the LMC, m-M = 18.50 ± 0.13 or, 18.58 ± 0.15, where the latter value results from a highly uncertain metallicity correction (Freedman et al. 2001). These agree with our previous RR Lyr HST parallax-based determination of the distance modulus of the LMC.

We present an absolute parallax and relative proper motion for the fundamental distance scale calibrator, RR Lyr. We obtain these with astrometric data from FGS 3, a white-light interferometer on HST. We find π abs = 3.82 ± 0.2 mas. Spectral classifications and VRIJHKT 2 M and DDO51 photometry of the astrometric reference frame surrounding RR Lyr indicate that field extinction is low along this line of sight. We estimate = 0.07 ± 0.03 for these reference stars. The extinction suffered by RR Lyr becomes one of the dominant contributors to the uncertainty in its absolute magnitude. Adopting the average field absorption, =0.07 ± 0.03, we obtain M RR V = 0.61 −0.11 +0.10 . This provides a distance modulus for the LMC, m-M=18.38 -18.53 −0.11 +0.10 with the average extinction-corrected magnitude of RR Lyr variables in the LMC, , remaining a significant uncertainty. We compare this result to more than 80 other determinations of the distance modulus of the LMC.