The systematic differences Astrolabe-FK5 of type
and 
are
derived from an analysis of the zenith distance residuals according
to the star azimuth. The stars are separated into azimuth intervals
which correspond to zones of 17 degrees wide in declination with an
overlap of 8.5 degrees approximately. For the stars in each interval,
the average mean zenith distance residual R.M. and the mean
declination 
are computed. If A is the mean azimuth of the stars in
the interval, R.M.(E) and R.M.(W), the mean residuals for azimuth
intervals at the east and west respectively, and if 
is the adopted
latitude of the astrolabe, it can be shown
(Guinot 1955;
Noël &
Débarbat 1990) that for each pair of azimuth intervals (east and west)
one has:
where A and S are the averages of the azimuth and paralactic angle
respectively of the stars in the azimuth intervals. The results
obtained with expressions (1) and (2) are given in Table 1 (click here), and in
Fig. 1 they are plotted with similar results obtained with the
automatic meridian circle of the U.S. Naval Observatory Station at
Black Bierch, New Zealand
(Corbin 1991). In Table 1 (click here), 
is the mean
declination of the stars comprised in each declination zone designated
by NN, and 
and 
are the number of stars at the corresponding east
and west azimuth intervals respectively, involved in the computation
of the systematic differences Astrolabe-FK5. According to the
principles of the method of equal altitudes
(Débarbat & Guinot 1970),
systematic differences in declination are given only for those zones
where the absolute value of the cosine of the paralactic angle is
greater than 0.3. The systematic differences in right ascension as
well as their mean errors given in Table 1 (click here) and Fig. 1 (click here) are multiplied
by 
.
Figure 1: Systematic differences in right ascension as a function of
declination: Instrument-FK5