Two afternoons of three to four hours were set aside for exercises using the WWW facilities listed in these lectures. In the first practical, the students were offered the names of five very extended (~ 20') radio galaxies from the 3C catalogufre, and asked to find out the positions of one or more of them from NED or SIMBAD, to obtain an optical finding chart from one of the various DSS servers, and to plot these with sky coordinates along the margins. The next task was to extract a 1.4GHz radio image from the NVSS survey, and a list of sources from the NVSS catalogue of the same region. A comparison of the two gave an idea of how well (or less well) the catalogued sources (or components) represent the real complex structures of these sources. The students were also asked to look at higher-resolution images of these sources in the ``Atlas of DRAGNs''. A further exercise was to find out where the many names under which these sources were known in NED or SIMBAD come from, by looking up the acronyms in the On-line Dictionary of Nomenclature. The optical object catalogues like APM, APS and COSMOS were then queried for the same regions of sky, which allowed the object classification as star or galaxy to be checked by comparison with the optical charts from the DSS server. Eventually, radio images from the FIRST survey were extracted in order to see to what extent the large-scale structure of the radio galaxies could still be recognised.
In the second practical, the students were given a chance to discover
a new, optically bright radio-loud quasar, a radio galaxy,
a starburst galaxy, or even a radio star! Each of the roughly two dozen
participating students was assigned a region of sky of the size
of a Palomar plate (6.5° × 6.5°) in the zone
08h < RA < 16h, +22°
< 
< +42° (the
region covered by the FIRST 1.4GHz survey at that time).
Each student was asked to extract all bright objects
(10 mag < B < 17 mag) from the USNO A1.0 catalogue (cf. §3 of
my tutorial)
in the region assigned to them. This was done by remote
interrogation of the USNO site, using the command findpmm from the
CDS client software, which had been installed for the school
on the computers at the IAC. The resulting object list (typically 3,000-10,000
objects per student, depending on the Galactic latitude of the assigned field)
was reformatted so as to serve as input for
interrogation of the NVSS source catalogue, which I had installed at the
IAC for the winter school. It had 1.67 million sources in Nov. 1997.
The FORTRAN program NVSSlist, publicly available from NRAO, was used
to search a circle of radius 10" around each optical object in the NVSS
catalogue, limited to 1.4GHz fluxes greater than 10 mJy so as to
assure the positional accuracy of the radio sources.
The students were asked to estimate the chance coincidence rate, and they
found that between 0.5 and 2 matches were to be expected by chance.
Actually each student had between three and 18 ``hits'' and was asked
to concentrate on the optically or radio-brightest objects to
find out whether the identification was correct, whether it was
new, and what was known previously about the object. For promising
candidates, a search in the FIRST 1.4GHz image database was suggested,
as well as an extraction of a DSS image. Some students even managed a
radio-optical overlay, and one of them found that the overlay facility
in SkyView had a bug when the pixel sizes of the overlaid images
(like e.g. NVSS and DSS) was not identical. This was later reported
to, acknowledged and fixed by the SkyView team. Unfortunately the
FIRST image server went ``out of service'' right during the practical.
With the 23 participating students, a sky area of 800 deg2 had been covered, and a cross-identification of altogether ~ 80,000 optical objects with ~ 12,500 radio sources was accomplished. Fifteen students sent me their results, of which only the most spectacular will be mentioned here. The 13.9mag IRAS galaxy NGC 3987 almost filled the 3' × 3' DSS image (which the students were asked to extract) and gave a splendid appearance with its edge-on orientation and strong dust lane. It was found to coincide with a 58 mJy NVSS source extended along the disk of the galaxy, while FIRST clearly shows a strong compact source (AGN?) and weak radio emission along the disk. A subsequent search in NED turned up a few other detailed radio studies ([Burns et al. (1987)], [Condon & Broderick (1986b)], and [Jaffe et al. (1986)]). Comparison of the 609 MHz flux from the latter reference shows that the compact central source has an inverted spectrum (rising with frequency), apparently not noticed before in literature. Another student came across UGC 5146 (Arp129), an interacting pair. While the FIRST image server was unavailable, the FIRST catalogue showed it to be a very complex source, aligned along the connecting line between the pair, and very extended. A third student ``rediscovered'' the well-known Seyfert galaxy NGC 4151 with its 600 mJy nuclear point source. At first sight no bright radio star had been discovered, not surprising given the more thorough searches for radio stars now available (Section 3.6).