THE RESULTS OF PHOTOGRAPHIC OBSERVATIONS OF GALILEAN
     SATELLITES WITH 26-inch REFRACTOR IN 1976-1981 AT PULKOVO.

                                     T.P.Kisseleva

                            (The Main Astronomical Observatory
                             Russian Academy of Science,
                             Pulkovskoe shosse,65-1,
                             St.-Petersburg, 196140
                             Russia)


                     Abstract.

   The photographic observations of Galilean Satellites of Jupiter
have been made in Pulkovo with 26-inch Refractor during the period
1976-1981 and 1984-1993. In this paper the results of the first
period 1976-1981 are analysed when 125 plates have been taken.
   The "scale-trail" technique was used for the astrometric reduc-
tion; each plate was oriented by two trails.
   The results of observations (the jovicentric and mutual coordi-
nates of satellites) have been compared with the ephemerides, cal-
culated by the theory of J.E.Arlot (G5). The internal mean square
errors of one jovicentric coordinates and mutual distances of the
satellites are equal to 0.10". The external errors of the coordi-
nates "satellite-satellite" are equal 0.08" and 0.17" in AR and Decl.
These errors increase with the distances between the satellites
and changes from 0.05" to 0.17" for distances from 100" to 1000".
   The investigations of (O-C) made it possible to determine the
phase corrections of Jupiter, the systematic errors of observations
and reduction, to analyse the orbital phase effects of satellites,
to estimate the precision of ephemerides. The error of the longi-
tude of the 4-th satellite was found equal to 0.3 minute. The pre-
cision of the ephemerides was estimated as 0.10".
   In the cases when there were foundamental stars on the plates
with the Jupiter and its satellites the positions of Jupiter were
determined without measuring the images of the planet using only
the measured positions of satellites and foundamental star. The
precision of this technique is higher then usual photographic ones.
The error of one position of the planet is 0.15". If the method
"scale-trail" is used in this task the positions of reference stars
are not required.