The optical astrometry observations of latitude/universal time variations made with 48 instruments at 31 observatories are used to determine the Earth orientation parameters (EOP) since the beginning of the century. The Hipparcos Catalogue is used to bring more than four million individual observations, made in the interval 1899.7-1992.0, into the International Celestial Reference System. The Earth orientation parameters (polar motion, celestial pole offsets and, since 1956.0, also universal time UT1) are determined at 5-day intervals, with average uncertainties ranging from 8 mas (in the eighties) to about 40 mas (in the forties). Making use of very long series of ground-based observations, the solution also leads to the improvement of proper motions of about ten per cent of the observed Hipparcos stars, with precision of ±0.2 — 0.5 mas/yr. In addition, 474 auxiliary parameters, describing the rheological properties of the Earth and seasonal deviations of the observations at contributing observatories, are found. The new solution provides the EOP series suitable for further analyses, e.g., for studying long-periodic polar motion, length-of-day changes or precession/nutation.

Obervations of the positions of Europa (J2) and Titan (S6) by Hipparcos, and Ganymede (J3) and Callisto (J4) by Tycho are analysed to give checks on the latest JPL ephemerides of the planets Jupiter and Saturn.

The observed positions of the satellites are compared with DExxx, using the G5 theory of the Galilean satellites and D.B. Taylor’s theory ofTitan to calculate their offsets from the barycentres of the two systems.

Each Tycho observation can be resolved into RA and Dec because the slitsare inclined to the scan direction. Hipparcos, on the other hand, which hasa uni-directional scan, requires a minimum of two (preferably orthogonal) scans to resolve the offsets into RA and Dec. We have subdivided the Hipparcos observations in time and solved for corrections to DExxx at several epochs over the period of the mission.

The Hipparcos observations of J2, and Tycho of J3 and J4, put tight constraints on the orbit of Jupiter, and agree closely with the series of ground-based observations made by the Carlsberg Meridian Telescope. In particular, they help to resolve a problem in the declination of Jupiter associated with larger than expected errors in the tracking of Voyager II and radio positions from the VLA.

The theory of stellar interiors is generally considered as well settled, and many branches of astrophysics rely on its results. However, the true situation is not so bright. Despite great successes at its beginning, in the sixties, it has to deal with many physical processes, generally poorly understood, which have large influences on the results and the derived quantities, like for instance ages.

After a rapid description of these difficulties, I will recall the majorquestions of the confrontation between theory and observations, and how HIPPARCOS can and will help. As an illustration, a few preliminary results will be presented, and future prospects foreseen. But, this short review is byno means exhaustive; there are many more results dealing directly or indirectly with stellar structure theory which have already been presented at theVenice Symposium (Perryman et al, 1997a), and many more, I am not aware of, will certainly appear soon.

An extensive work on the luminosity calibration of the HR diagram has been performed using the LM method based on the Maximum Likelihood estimation (Luri et al. 1996). The method uses all available information: Hipparcos parallax and proper motion data as well as radial velocities and some measured spectroscopic or photometric parameters related to luminosity. On the other hand, the method takes into account the censorship of the sample and the errors of the data. Finally, it is able to identify and separate in a given sample, groups with different luminosity, kinematical or spatial characteristics providing not only the corresponding luminosity calibration relationship, but also improved individual absolute magnitude estimates.

In order to asses how reliable the luminosity calibrations based on the MK spectral classification are, the LM method has been applied to a sample of B to K stars with MK classification selected from the Hipparcos survey. All known binaries, variables and spectroscopically peculiar stars were removed from the sample. The final sample contains 22054 stars. Our results show that the relation between absolute magnitude and luminosity class has a large intrinsic dispersion. As a consequence, the assignation to a star of a mean absolute magnitude given by a MK classification calibration is a rough procedure. It is preferable to use the individual absolute magnitudes estimated using all the available information for each star, like the estimates provided by the LM method (Gómez et al. 1997).

Early release data were obtained under the 1982 announcement of opportunity for about two dozen stars belonging to seven classes whose basic properties or evolutionary status were uncertain at the time. A good many remain so. Many of the stars were too faint and far for parallaxes to be statistically significant, though many of the proper motions are. Two papers presenting the data have been accepted by archival journals (Trimble and Kundu 1997, 1998). Both bear the same title as this Abstract The first deals with the R Coronae Borealis variables, the second with the other six classes (FK Comae Stars, CVs, runaway T Tauri stars, and so forth).

As a preparation to the HIPPARCOS mission, a large observing programme on NLTT stars (propermotion > 0.18 ″/yr) was started in Genevaphotometry. The original programme consists of 10047 stars brighter than mR = 11.5, or mR = 12.5 if of colour class m. Among them, 7813 targets could be included in the HIPPARCOS programme, selected according to their observability and internal priorities in favour of large parallaxe stars (photometric distances < 100 pc) and high-velocity stars. The bulk of new nearby, halo, mild-metal poor and SMR stars in the HIP Catalogue originates from this proposal (N° 139). No less than 208 new nearby stars with π ≥ 40 mas were discovered south of δ +10°, the closest has π(HIP)= 182 mas. Radial velocities were obtained with CORAVEL at OHP and ESO. Most aspects of the early evolution of the Galaxy may be addressed with this sample. Here we discuss, as examples, the ages of the thick disk and of the galactic bulge.

A major result of the Hipparcos mission is the availability of accurate parallaxes for a much-increased sample of nearby subdwarfs. On the basis of those data, both Reid (1997-R97) and Gratton et al (1997-G97) have used main-sequence fitting techniques to obtain new estimates of globular cluster distances, deriving significantly larger moduli (by 0.15 to 0.3 mag) for extreme, metal-poor clusters such as M92 and M15. These re-calibrations are not solely the result of the revised parallax data - both D’Antona, Caloi & Mazzitelli (1997 - DCM) and Reid show that similar results can be derived from pre-Hipparcos subdwarf data if the calibrations are not tied exclusively to HD 103095. The longer distance scale leads to a brighter inferred turnoff, and, combined with the latest models (DCM), ages of only 11 to 13 Gyrs - substantially younger than the 16±2 Gyrs proposed by Bolte & Hogan (1992).

Recently, Pont et al (1997) have suggested that these revised distances reflect systematic bias in the calibration, notably underestimation of subdwarf abundances. Their distance modulus for M92 ( (m-M)0=14.68) is closer to the Bolte/Hogan value (14.65 mag) than those derived byG97 (14.82 mag) or Reid (14.93 mag). While we believe that Pont et al overestimate the bias, particularly, where high-resolution spectroscopic analyses are concerned (G97, Reid, preprint), there is no question that significant uncertainties remain in the cluster distance calibration. Colours,rather than luminosities, are particularly vulnerable, and a mismatch of 0.01 in (B-V) translates to δ(m — M)0 ~ 0.05 mag.

The parallax data gathered by Hipparcos for field subdwarfs allow a muchmore precise determination of globular cluster distances by direct sequencefitting than was previously possible. We determined the distance and age ofthe old, representative globular cluster M92 from a set of more than 500 subdwarf candidates with Hipparcos parallaxes. Precise [Fe/H] values were derived using the equivalent width of the CORAVEL cross-correlation function.

Our best estimate of the distance of M92 is (m — M)v = 14.67 ± 0.08 (including binaries with a statistical correction) or (m — M)v = 14.74 ± 0.08 (classic treatment, i.e. excluding binaries). The agreement of the cluster sequence with the position of extreme metal-poor field subdwarfs is remarkable [figure, left]. The distance found is slightly higher than previously thought. The corresponding ages, derived by comparing the luminosity of the turnoff and subgiant-branchstars with up-to-date evolution models [figure, right], are 14±1.2 Gyr or 13±1.2 Gyr respectively, implying a minimum age of 13 Gyr for the Universe.

Other authors have claimed that larger distances, and smaller ages, resulted from Hipparcos subdwarf parallaxes. Although there are some differences in the data sets used by each author, the main difference resides in the treatment of systematic and selection biases. We have examined at some length the biases affecting the determination of the mean luminosity of a set of subdwarfs selected in the usual way. By means of Monte Carlo simulations, we show that selection biases act in a direction opposite to the classic Lutz-Kelker bias affecting parallax data, and that they can be dominant. The biases introduced are of the order of 0.1 mag.

It is pointed out that, now that the whole Hipparcos catalogue is available, we shall be able to greatly reduce the systematic biases by refining the fitting procedure. Firstly, by imposing no [Fe/H] limits and fitting the subdwarf locus for all [Fe/H] values at the same time. Secondly, by imposing no parallax error limit and fitting the data in parallax space instead of magnitude space (comparing observed parallaxes directly with parallaxes predicted from the models).

Wavelengths of stellar spectral lines depend not only on the star’s motion. Until recently, accurate studies of shifts not caused by radial motion were feasible only for the Sun. Solar lineshifts are interpreted as gravitational redshift (636 m/s) and convective blueshifts (~ 400 m/s; caused by velocity-brightness correlations). In other stars, such effects may be greater (Dravins & Nordlund 1990). Accurate astrometric radial velocities are now available from Hipparcos (Dravins et al. 1997a; 1997b), permitting studies of such shifts also in some other stars.

For such stars in the open clusters of Hyades, Ursa Major and Coma Berenices, a spectroscopic program is in progress, analyzing wavelength shifts in groups of lines with different strengths, excitation potentials, etc., using the ELODIE high-precision radial-velocity instrument (Baranne et al. 1996) at Haute-Provence Observatory.

Space astrometry now permits accurate determinations of stellar radial motion, without using spectroscopy. Although the feasibility of deducing astrometric radial velocities from geometric projection effects was realized already by Schlesinger (1917), only with Hipparcos has it become practical. Such a program has now been carried out for the moving clusters of Ursa Major, Hyades, and Coma Berenices. Realized inaccuracies reach about 300 m/s (Dravins et al. 1997).

Discrepancies between astrometric and spectroscopic radial velocities reveal effects (other than stellar motion) that affect wavelength positions of spectral lines. Such are caused by stellar surface convection, and by gravitational redshifts. A parallel program (Gullberg & Dravins 1997) is analyzing high-precision spectroscopic radial velocities for different spectral lines in these stars, using the ELODIE radial-velocity instrument atHaute-Provence.

After having studied in great detail the observational HR diagram (log Teff, Mbol) composed by 40 main sequence stars of the Hyades (Perryman et al.,1997, A&A., in press), we have tried to apply the same method to the observational main sequences of the three next nearest open clusters: Coma Berenices, the Pleiades, and Praesepe. This method consists in comparing the observational main sequence of the clusters with a grid of theoretical ZAMSs. The stars composing the observational main sequences had to have reliable absolute bolometric magnitudes, coming all from individual Hipparcos parallaxes, precise bolometric corrections, effective temperatures and metal abundances from high resolution detailed spectroscopic analyses. If we assume, following the work by Fernandez et al. (1996, A&A,311,127), that the mixing-lenth parameter is solar, the position of a theoretical ZAMS, in the (log Teff, Mbol) plane, computed with given input physics, only depends on two free parameters: the He content Y by mass, and the metallicity Z by mass. If effective temperature and metallicity of the constituting stars of the 4 clusters are previously known by means of detailed analyses, one can deduce their helium abundances by means of an appropriate grid of theoretical ZAMS’s. The comparison between the empirical (log Teff, Mbol) main sequence of the Hyades and the computed ZAMS corresponding to the observed metallicity Z of the Hyades (Z= 0.0240 ± 0.0085) gives a He abundance for the Hyades, Y= 0.26 ± 0.02. Our interpretation, concerning the observational position of the main sequence of the three nearest clusters after the Hyades, is still under way and appears to be greatly more difficult than for the Hyades. For the moment we can say that:

1. ‒ The 15 dwarfs analysed in detailed in Coma have a solar metallicity: [Fe/H] = -0.05 ± 0.06. However, their observational main sequence fit better with the Hyades ZAMS.

2. ‒ The mean metallicity of 13 Pleiades dwarfs analysed in detail is solar. A metal deficient and He normal ZAMS would fit better. But, a warning for absorption in the Pleiades has to be recalled.

3. ‒ The upper main sequence of Praesepe, (the more distant cluster: 180 pc) composed by 11 stars, analysed in detail, is the one which has the best fit with the Hyades ZAMS. The deduced ‘turnoff age’ of the cluster is slightly higher than that of the Hyades: 0.8 Gyr instead of 0.63 Gyr.

The ‘1996 Edition’ of the Catalogue of [Fe/H] determinations by Cayrel de Strobel et al. (1997, A&A S 124,1) and two recent papers by Castro et al. (1997, AJ Vol.114, N.1) and by Feltzing and Gustafsson (A&A in press) have made possible to increase in the theoretical HR diagram (log Teff, Mbol the number of SMR stars.

The SMR Stars of this new enlarged sample had to have reliable absolute magnitudes, coming all from Hipparcos parallaxes, precise bolometric corrections, effective temperatures and metal abundances from high resolution detailed spectroscopic analyses.

With the help of an appropriate grid of isochrones computed by Lebreton (1997, Perryman et al. A&A, in press), ‘turn-off ages’could then be attributed to the slightly evolved stars (subgiants) of the sample. The (log Teff, Mbol) diagram constituted by the new sample of SMR stars, shows that the conclusions in a former paper by Cayrel de Strobel (1987, A&AJ 8,141) remain valid: the SMR stars areold stars in spite of their higher than solar metallicity. The result, that metal-rich stars were in the mean old stars, was interpreted in the 1987 paper as due to a more chemical uniformity of the nowadays interstellar medium of the Galaxy with respect of the older much more active interstellar medium.

We have analysed the standards of the MK system in the B0-F5 spectral region with the help of Hipparcos parallaxes, using only stars for which the error on the absolute magnitude is ≤ 0.3 mag. The sample stars (about one hundred) were scrutinized for companions and for interstellar extinction. We find that the main sequence is a wide band and that, although in general giants and dwarfs have different absolute magnitudes, the separation between luminosity class V and III is not clear. We conclude that there is no strict relation between luminosity class and absolute magnitude. The relation is only a statistical one and has a large intrinsic dispersion.

We have analysed similarly the system of standards defined by Garrison and Gray (1994) separating low and high rotational velocity standards. We find similar effects as in the original MK system.

The absolute magnitudes of δ Scuti stars derived from parallaxes measured by the Hipparcos satellite were compared with the previous estimates based on photometric uvby² indices, and significant differences were found which are related to photometric effects of metallicity and rotational velocity. A reliable calibration of Mv in terms of the photometric indices shall include an estimate of these effects. It is important also to take into account the possible presence of unresolved close companions in order to fully exploit the accuracy of Mv of nearby stars derived from the trigonometric parallaxes.

The Mv of few bright SX Phe stars support the period-luminosity relation obtained with ground based observations of globular clusters, while it does not seem to confirm the empirical dependence of this relation on the metallicity.

Some high amplitude δ Scuti stars with intermediate or normal metallicity and small and uncertain parallax have apparently a very low luminosity. Simulations of Mv determinations from observed parallaxes based on the discussion of observational errors by Lutz and Kelker (1973) have shown that the low luminosity could be an effect related to these errors.

Distances and absolute magnitudes have been obtained from an early distribution of Hipparcos data for a number of well-known late-type variables. These data were requested in 1981 in order to ensure that these important stars were not excluded from the Input Catalogue on account of their variability or very red color.

Measurements were requested for 20 Mira variables, and 14 of these were successfully observed. Of these, the nearest to the Sun were found to be R Leo (at 102 pc), χ Cyg (106), R Cas (107), R Car (128), T Cep (210),and R Aql (211), with formal errors ranging from ±12 to ±57 pc. Mira itself was not considered because the presence of a companion causes an 11-month wobble in its photocenter which is likely to lead to a spurious parallax. The absolute visual magnitudes of Miras at mean maximum range approximately from -2.5 to +2.0 with a clear dependence on spectral type (cooler stars being fainter). This trend is reversed for absolute magnitudes in the near infrared.

Results are also given for 10 small-amplitude variables of late type. The nearest of these are L2 Pup (M5 III, 61 pc) and R Dor (M8e, 62pc). The symbiotic star CH Cyg lies at a rather uncertain distance of 268 ± 65 pc.

Data were obtained for 23 carbon stars, including the brightest representatives of each of several subgroups. The nearest of these was found to be the SRb variable U Hya, at 162 ± 20 pc. The stars 19 Psc, Y CVn, W Ori, and HD 201626 were all found to be somewhat more than 200 pc away.

Alpha Orionis (Betelgeuse) is the brightest, and presumably the nearest, of the M-type supergiants. It is also one of the least reddened. These features make it important for the calibration of the absolute magnitudes and intrinsic colors of M supergiants. However, its distance has been difficultto establish, and there has been considerable debate concerning its reddening.

The Hipparcos parallax for α Ori is 7.63 ± 1.64 mas, corresponding to a distance of 131 ± 30 pc and a true distance modulus (m — M)o of 5.59 ± 0.48 mag. If we adopt <V> = 0.50 and E(B - V) = 0.05, the mean absolute visual magnitude becomes <Mv> = -5.24, a result consistent with a luminosity classification of lab.

The CN strength of α Ori measured in the near infrared also indicates luminosity class Iab. Furthermore, the star’s absolute magnitude at 1.04 μm, for the Hipparcos distance, is -8.34, in agreement with the M2 Iab supergiants in h &. χ Per for a distance modulus of (m — M)o = 12.0.

From the Hipparcos proper-motion components and distance and a published radial velocity, we derive a space motion of 27.8 km s-1 with components U = +19.7, V = -7.3, and W = -1.5 km s-1 with respect to the Sun. This motion is not very different from the local standard of rest. It seems unlikely that μ Ori, at a distance only one-third that of the Trapezium, is physically associated with the Orion Complex.

At 131 pc, the observed optical angular diameter of μ Ori, 0.055 arcsec, translates to a linear diameter of 7.2 AU. Centered on the Sun, itsphotosphere would be in the asteroid belt. Its angular diameter measured bythe Hubble Space Telescope in the UV (corresponding to the size of the chromosphere) is a factor of 2 larger. Surface features detected interferometrically are believed to be ~ 10% of the diameter in extent, corresponding to 0.7 AU or 108 km.

The new Hipparcos parallaxes and photometry are used to determine individual masses and absolute (bolometric) magnitudes for the components of nearby visual binaries with good to very good orbits. The impact on the mass-luminosity relation (MLR) in the range 0 < MBol< +7.5 mag is then evaluated.

We selected 335 visual binaries within 50 pc (σπ/π < 10%) for a full error analysis of their orbits by computing the covariance matrix of the orbital elements with Pourbaix’(1994) algorithm. Using ΔHp and π, we estimated fractional and component masses as well as absolute magnitudes with theirrespective errors: 52 binary systems have relative mass errors smaller than 15%. Lutz and Kelker (1973) corrections have been applied to both datatypes. A new relation BC(Hp) as a function of Teff was obtained for the conversion to bolometric magnitudes. A doubly weighted linear regression model was applied next (Babu and Feigelson, 1996): we derived a ”mean” slope 3.82 ± 0.07 and zero point 4.94 ± 0.03 for the MLR, assuming a linear relationship.

Conclusions: a) the improvement of the data on masses by Hipparcos is largely quantitative; b) not all systems agree: small fluctuations from a ”mean” MLR are found as expected from evolutionary or abundance effects; c) the break in the slope of the MLR near MBol = +7 cannot be assessed due to a lack of low-mass binaries in our sample.

The Hipparcos astrometric satellite observed 993 known eclipsing binaries. Prom 1048 eclipsing binaries of the Brancewicz and Dworak ’s catalogue (1980), there are only 368 observed by Hipparcos, as the others are too faint.

On the basis of these common stars, we have undertaken the comparison ofthe satellite parallaxes with the photometric ones. We obtain a good agreement between the two systems with a mean dispersion of 5 mas. The greatest dispersion is for the goup of eclipsing binaries of the EW type. The eclipsing binaries of this group have a mass-luminosity relation different from those of the main sequence stars.

An inadequate knowledge of the component spectra of an eclipsing binary seems to be one of the most important cause of discrepancies. For some brightest stars, the new Hipparcos trigonometric parallaxes differ very significantly from those previously known from ground based measurements. It seems that the presence of a close visual components causes statistically an underestimation of photometric or overestimation of Hipparcos parallaxes.

172 Hipparcos eclipsing binaries have a close visual component, generally less than 10 arcseconds apart. The list of these stars for various stellar groups and a brief analysis are presented.

We use the Hipparcos survey to derive an improved model of the local galactic structure. The availability of parallaxes for all the stars permits direct determination of stellar distributions, eliminating the basic indeterminacy of classical methods based on star counts. Hipparcos gives for the first time a truly three-dimensional view of the solar vicinity, and a complete, homogeneous and highly accurate set of magnitudes and colours. This means that new techniques can be applied in the treatment of the data which place strong constraints on a model that tries to describe the local Galactic structure. Here we investigate how well a static model of low complexitycan describe the Hipparcos observations.

The interpretation of the Hipparcos data is complicated by various observational errors and selection effects that are hard to treat correctly. We do not try to correct the data, but instead use a model and subject this model to the same observational errors and selection effects. A model catalogue is created that can be compared with the observed catalogue directly in the observational domain, thereby eliminating the effects from various biases.

Many features in the HR diagram are for the first time seen in field stars thanks to Hipparcos, such as the slanted red giant clump, previously seen in rich old open clusters such as Berkeley 18. This and other features ofthe observed HR diagram are well reproduced by the model thanks to the rather detailed modelling of the joint Mv/B — V distribution. Actually, separate distributions were derived for the three different components, disk, thick disk and halo, using the kinematic characteristics of the components to discriminate between them.

The Hipparcos catalogue provides an accurate and extensive sampling of the solar neighbourhood HR diagram. The morphology of this diagram depends on selection criteria of the catalogue such as the limiting magnitude, angular separation and on the characteristics of the stellar populations near the sun (space density, metallicity, star formation rate, etc). Since the Hipparcos data are so accurate, one needs to model precisely the different selection bias and, at the same time, parametrize models of the galactic stellar populations with sufficient flexibility that as much information as possible can be grasped from the catalogue.

Comparisons between our model and the Hipparcos catalogue will be presented elsewhere. Since the quantity of information contained in the Hipparcoscatalogue is so important, models ought to be complex, and external contraints, obtained prior to any general comparison with the model, are welcome.

A major factor that influences the distribution of the stars in the HR diagram is the metallicity. For the late type stars, the metallicity distribution can be best studied by re-analysing a volume-limited sample of stars from the catalogue.