A complete description of all aspects of the preparation of the Hipparcos Input Catalogue can be found in the proceedings of two Colloquia on `Scientific Aspects of the Input Catalogue Preparation' (Aussois, June 1985, ESA-SP 234, Turon & Perryman eds; and Sitges, January 1988, Torra & Turon eds), in ESA-SP 1111, Volume II (The Hipparcos Mission: Pre-Launch Status, Perryman & Turon 1989), and in a series of papers in Astronomy & Astrophysics covering star selection (Turon et al. 1992a), astrometric aspects (Jahreiß et al. 1992), photometric aspects (Grenon et al. 1992), solar system objects (Bec-Borsenberger 1992a), long-period variable stars (Mennessier et al. 1992), the adopted galactic extinction model (Arenou et al. 1992), and the VLBI link to an extragalactic frame (Lestrade et al. 1992). The performances of the Input Catalogue, as tested by the first results from the satellite, are presented by Crifo et al. (1992) and Turon et al. (1992b). All aspects of the link of the Hipparcos reference frame to an extragalactic reference frame are described by Argue (1989) and Jahreiß et al. (1992). The data on double and multiple systems will be contained in the `Catalogue of the Components of Double and Multiple Stars' (CCDM, Dommanget et al., in preparation). Annex 1 is a subset of a preminary version of this catalogue. The improved orbital elements of the relevant minor planets, along with predicted ephemerides for the years 1991 and 1992, have been given by Bec-Borsenberger (1990, 1992b).
The 118000 stars of the Hipparcos Input Catalogue were selected from some 214000 distinct candidates, contained in the 214 observation programmes listed at the end of this introduction. The scientific proposals deal with a wide variety of astrophysical and astronomical subjects: solar system, stellar physics and evolution, galactic structure, kinematics, dynamics and evolution, cosmic distance scale, and reference systems. With the endorsement of the ESA Hipparcos Scientific Selection Committee, the INCA Consortium supplemented the proposed list of stars with others considered to be unrepresented in the original compilation. Most importantly, this included a basic list of bright stars, essentially complete to a well-defined limiting magnitude, and hence referred to as the `survey'.
The original set of proposed stars was, not surprisingly, poorly suited
to the observing capabilities of the Hipparcos experiment: the sky
distribution of the 214000 proposed stars
showed a strong
concentration towards the galactic plane, and especially towards the
galactic centre, and their 
-magnitude distribution showed a predominance
of faint stars. As a result of the work of the INCA Consortium in merging,
optimising, and supplementing the original list, the final Hipparcos Input
Catalogue of about 118000 stars can broadly be considered as being
composed of two parts:
If no spectral type was available, the break was taken at 
mag. The original star selection for this survey was made from the
SIMBAD data base (Egret et al. 1991).
Although the inclusion of the bright stars was partly motivated by reasons related to the satellite operations (attitude determination) and data reductions, the precise choice of stars was not in itself significant for these purposes. The only two basic requirements were to retain a reasonably regular stellar density of bright stars all over the sky, and to ensure that a large sub-sample of these stars would have sufficiently accurate a priori positions. Stars were therefore chosen such that any further statistical uses of the catalogue, for example in the domains of galactic physics, would be enhanced by this choice. The above survey definition satisfies all three requirements, while at the same time reducing the proportion of red giant stars in favour of A and F stars. The latter, being typically closer, will get a correspondingly more accurate determination of their distances and space motions.
As a result, the 118000 selected stars, just 55 per cent of those
originally proposed, actually represent 94 per cent of the `priority 1'
stars. Their distribution on the sky is much smoother than
that of proposed stars, even though more (bright) stars are selected in
the galactic plane than outside. Their -magnitude distribution
shows a maximum between 7 and 10 mag, and very few stars
fainter than 12 mag (these stars require a very large fraction of the
available observing time). The final star selection (Turon 1989) was
made following an iterative process, using numerical simulations of the
mission (Crézé et al. 1989). These simulations progressively took
into account all the astrometric and photometric data collected or
newly obtained within the INCA Consortium, as well as all data coming
from parallel work on double, multiple and variable stars. All these
data are collected in a dedicated data base, the INCA data base,
originally constructed as a sub-base of SIMBAD, with the data and
cross-identifications from SIMBAD (Arenou & Morin 1988; Turon et al.
1991). The updating and interrogation software of this data base was an
important asset in storing and managing all the data necessary for the
construction of the Input Catalogue.
In the context of the Hipparcos Input Catalogue, and especially in discussions of the Input Catalogue's statistical properties, the word `star' is frequently used when catalogue `entry' would be more appropriate. An `entry' refers to a satellite target: it may be a single star, a component of a double or multiple system, or a unique joint entry for a close double or multiple system. This is a specific feature of the Hipparcos Input Catalogue, reflecting the satellite's detection system. It should be stressed again that the Input Catalogue was compiled specifically for the Hipparcos programme. Its stellar and data contents reflect these fundamental requirements.