All of the observations listed in Tables 6-9 were obtained using the 1-metre Jacobus
Kapteyn Telescope on the island of La Palma (longitude W 
, latitude N
) using coated CCD detectors at the f/15 Cassegrain focus.
We were allocated one week of telescope time in each of the years 1990, 1991, 1993 and 1994. These generally corresponded to a period around the time of Full Moon near to the date of opposition of Saturn. In the first three years, we were able to make observations throughout the entire week apart from single nights in 1991 and 1993 when poor weather or bad seeing prevailed. Additional observations were obtained during late July and early August 1990 as part of the Isaac Newton Group service programme.
During the 1994 observing run, serious forest fires broke out close to the Observatory in the early evening of August 24th, preventing operation of the telescopes during that night and eventually forcing the evacuation of the entire Observatory for several days. Consequently, observations of Saturn were obtained on only two nights in 1994.
A summary of the observing situation for all four years is given in Table 1.
Table 1: Observing conditions for 1990-94. Entries in italics indicate
service observations which were not part of the main campaign
Two different CCD detectors were used. In 1990 and 1991, we used a GEC
chip with 
pixels, giving an approximate scale of 
per pixel in the focal
plane. The size of the array was 
pixels. In 1993 and 1994, a larger EEV chip was
available, having a 
array of 
pixels. The corresponding fields
of view are listed in Table 1. The larger field of view in 1993 and 1994 enabled us to obtain CCD
frames containing images of up to eight satellites at a time. In particular, we were able to obtain
images of Iapetus, the outermost satellite, in the same CCD frame as one or more of the other
satellites. This is an essential aspect of our observational technique, since the exposure time of
most of our CCD frames is too short (4 seconds) to record images of any astrometric reference
stars and the satellite images must therefore be analysed in the form of relative positions. There
are additional advantages to be gained from treating the data as inter-satellite measures and these
are explained in Sect. 3.3 of Harper & Taylor (1994).
In all four years, we used a Gunn Z filter with a central wavelength of 930 nm and a full width at half maximum of 150 nm. This coincides with methane absorption bands in Saturn's atmosphere and reduces the relative brightness of the planet's globe. However, the spectrum of the rings is not dominated by methane, and the filter therefore has little effect on them. The brightness of the rings made it difficult to measure the position of Mimas reliably on many of the CCD frames since the images of the satellite were superimposed on a background of strong scattered light.
A summary of the number of images of each of the target satellites in each year is given in Table 2.
Table 2: a. Number of images of each satellite