The Hipparcos Catalogue characteristics are described in full detail in the documentation accompanying the catalogue. A series of papers in Astronomy and Astrophysics summarise the key attributes of the catalogue (Perryman et al. 1997a, Lindegren et al. 1997, Kovalevsky et al. 1997, van Leeuwen et al. 1997a). The Tycho Catalogue is also described in details in the documentation and a summary presented in Høg et al. (1997).
The Hipparcos Catalogue includes 118218 preselected entries brighter than
V
12.4, 117955 with associated astrometry, 118204 with associated
photometry. Observations were made from 1989.85 to 1993.21, with a mean epoch
of close to J1991.25, adopted as the catalogue epoch.
Standard errors are functions of magnitude and ecliptic latitude.
Median standard errors for stars brighter than Hp=9 mag
are given in Table 1.
Systematic parallax errors are estimated to be smaller than 0.1 mas from various
investigations (Arenou et al. 1997). The coincidence with the adopted reference
system (ICRS) is estimated to be within
0.6 mas about all 3 axes, and the
deviation from inertial in the range
0.25 mas/yr, also about all
3 axes. (Kovalevsky et al. 1997).

Table 1: Median astrometric standard errors for stars brighter than
Hp=9 at epoch J1991.25
Among the 118218 entries, 23882 were classified as double or multiple
systems or components of such systems (Lindegren et al. 1997). Detailed
solutions for these entries are given in the five parts of the Double and
Multiple Systems Annex (DMSA), as decribed in Table 2. The DMSA annex
provides a fairly complete census of binaries among the Hipparcos stars with
4 and
arcsec (depending on
). The
accuracy of the astrometric parameters of double and multiple entries may be
slightly degraded with that of comparably bright single stars, depending
on the component separation and magnitude difference.

Table 2: Statistics of solved or suspected double/multiple systems contained in
the Double and Multiple Systems Annex (DMSA)
In addition to the astrometric parameters, the Hipparcos data reductions produced high-precision, fully calibrated, photometry in the Hp broad-band filter (spectral range: 375-750 nm). The median photometric precision on Hp for stars brighter than Hp=9 mag is 0.0015 mag. The mean number of photometric observations per star is 110 (van Leeuwen et al. 1997a). The standard errors of one observation (transit error) and of the median magnitude, as a function of Hp are given in Table 3. As a result of this high accuracy, investigations for variability were made possible for all programme stars. The results are summarised in Table 4.

Table 3: The average errors, in magnitude, per field transit (second
line), and on the median Hp magnitude (thirrd line), as a function of Hp

Table 4: Statistics of variability investigations. `New' = newly detected as
variable by Hipparcos