The Space Telescope (ST) is one of the most exciting projects presently planned in space astronomy. It will make a considerable contribution to astronomical research throughout the rest of this century. The high resolution 2.4 m telescope will be placed in orbit probably in 1985 by the Space Shuttle and will provide an astronomical capability unattainable by the ground-based telescopes.

The European Space Agency decided to participate in this NASA programme in 1976. ESA’s part in the programme includes the production of a major subsystem (the solar arrays and associated mechanisms), the building of the Faint Object Camera, and future participation in the operational activities and in the running of the Space Telescope Science Institute.

The Hipparcos satellite is described as well as the scanning process of the sky by its optical design, as a consequence of the satellite’s motion. From these data, the possibilities of the mission is deduced and applied in the case of the binaries.

Discussion is made of what may be expected for different double star categories including spectroscopic, astrometric, close visual and wide pairs. Information are given on how proposals should be made for lists of binaries that should be observed.

The number of new couples available for the Space Telescope can be estimated through the study of prior surveys and observational selection.

Surveys realized at the beginning of this century, as well as actual surveys being recalled, show that couples remain to be discovered. A method is developed showing the importance of instrumental selection by the difference of magnitude between the components, this way can be estimated the number of couples inaccessible by earth observations due to differences of magnitudes.

Limiting ourselves to the magnitude 10, we show that the Space Telescope will permit to discoverer mostly close couples with a small difference of magnitude and a short period (P 20 y). They represent several thousands of stars.

The problem under consideration is that of determining the seven elements (the vector ) characterizing the relative three-dimensional motion of the components of a binary star, using observations of the two-dimensional projection on the sky of the relative positions of those components as a function of time. Sooner or later a differential correction process is employed; therefore, this correction procedure will be reviewed first.

The basic approach is to derive a first estimate of the orbital parameters, form residuals from each of the observations, and assume these residuals represent a normally distributed random error, plus systematic errors due to an incorrect initial orbit. Expressing the residuals as functions of the orbital elements, these functions can be expanded in a Taylor series about the initial orbit and truncated after the first-order terms, giving the following approximate linear relation:

The special problems presented to the computer of orbits by radial-velocity observations of visual binaries are discussed under three heads: (i) problems caused by the small range of velocity variation, (ii) problems caused by the long periods, (iii) inconsistency between visual and spectroscopic observations. It is pointed out that radial-velocity observations, even when they are insufficient for independent determinations of orbital elements, impose constraints on possible values of those elements which may in fact be helpful to the computer of visual orbits. In particular, as is illustrated by reference to ADS 8189, even a few observations of the radial-velocity are sufficient to destroy the mutual independence of the elements e, and ω.

The use of the Fourier transform in the computation of binary star orbits has many advantages over traditional (configuration space) analysis. The method can be applied to visual, spectroscopic and astrometric binaries, as well as to visual binaries where only the separation or position angle is known. While this technique requires the use of a digital computer, the algorithms for implementation of the solutions for the various types of binary star motion are quite similar. A description of the implementation of the astrometric binary solution will be presented along with a discussion of the various estimators for the significance of the derived orbital elements obtained in the course of the solution. If time permits, a summary of some Monte Carlo investigations of the effects of observational noise and temporal sampling on the derived orbital elements will be presented.

A complete overview on the programmes that are used at the Royal Observatory of Belgium, for conducting the computation of the orbit of a visual binary based on three fundamental points and the apparent areal constant is given. They include the necessary investigations of all incompatibilities that may be encountered during the computation, for instance - when resolving the Thiele-van den Bos equations – the recognition of the nature of the solution (elliptic, parabolic or hyperbolic) and the identification of trivial solutions. Some particular computations are made when necessary, for guidance to the best choice of the three fundamental points and the areal constant.

The Thiele-Innes elements show a discontinuity for e = 1. Other elements are proposed having similar expressions and same advantages but offering the interest of being valid in all cases of eccentricity.

Described is the plan of a study just underway to assess the progress so far in the astrometric search for low mass companions. The range of periods which would be discernible, based on the obsprvations’ distribution in time, and the range in amplitude of the detectable orbits, based on careful analysis of the accuracy of the observations, can be used to determine the confidence level of any perturbations as well as to specify what possible companions and orbits for a given star have been ruled out. This study, when completed for several nearby stars, should yield the first comprehensive progress report for the search for astrometric companions.