Several fields (
) near selected bright, possibly disturbed, galaxies (NGC 5334,
NGC 450, NGC 520, ...) as well as far from them, have been searched for quasar candidates on the basis of
the presence
of an ultraviolet excess (more precisely a U/B excess). This work was initiated several years
ago (Arp & Surdej 1982; Surdej et al. 1982) and status reports may
be found in Gosset et al. (1986) and Swings
et al. (1988). Initially, this project was triggered by the need for further observational evidence
supporting the uniform distribution of quasars, their possible location in superclusters
(Oort 1981; Oort et al. 1981; Oort 1983), or in the vicinity of irregular bright galaxies and/or
nearby the companions thereof
(Arp 1981; Arp 1983; Arp & Duhalde 1985; Arp 1987).
Thanks to their high luminosity, quasars are the landmarks of the far Universe. They constitute one of the
rare tracers of matter at very large distances.
Quasars should outline some characteristics of the Universe, such as the distribution of matter (at least the
luminous one) at mid- to large-scales. If the existence of a clustering up to a typical scale of 
Mpc seems now
estableished (Iovino et al. 1991; Shanks et al. 1987), the question is
still a matter of debate for the largest scales. The signal for clustering in excess of randomness is usually
considered as non-existent at large scales but the detections of three different quasar superstructures
Mpc) have been reported in the literature (Webster 1982;
Crampton et al. 1987, 1989; Clowes &
Campusano 1991a). A fourth possible case is presented in Graham et al.
(1995). Some authors even suggested that all the small-scale clustering be localized in such
superstructures
(Crampton et al. 1989; Mo & Fang 1993;
Clowes & Campusano 1991b). The study of this kind of clustering is of the highest
interest and is a strong observational constraint situated, for quasars, somewhere between the structures
found locally in the different maps of galaxies (as reported, for example, by de
Lapparent et al. 1988 and by Geller & Huchra 1989
and usually corresponding to a redshift z 
0.1) and the region of relevance of the
observations of the cosmic microwave background. The latter is rather consistent with a Universe without
strong pristine structures
(Smoot et al. 1992; see also Partridge 1995). The possible existence of quasar
superstructures is therefore extremely constraining for the different theories of the Universe. Such studies are
however difficult due to the relative rarity of these interesting objects.
Quasars are also interesting because they can be considered as beacons in the Universe, and their luminosity
can be exploited to analyse the intervening absorbing matter lying between the quasars and us. They allowed
the discovery of the Ly
forest clouds which remain one of the most disturbing constituents of the
Universe
(Sargent et al. 1980; Sargent & Boksenberg 1983;
Smette et al. 1992; Borgeest 1993; Smette 1994).
Almost as a by-product, damped Ly
and Mg II systems, usually associated with intervening
galaxies, are now inducing a reappraisal of the dimension of these dynamical systems which are suspected to
be much more extended than previously thought
(Bergeron & Boissé 1991;
Aragón-Salamanca et al. 1994). Comparative studies of the absorbed spectra of a close
pair (be it true or just apparent) of quasars is a tool to derive sizes of the intervening clouds
(see Smette et al. 1992; Smette 1994;
Smette et al. 1995).
Finally, the deep nature of quasars is still something which is not fully understood. Several questions related
to their evolution and to the origin of their high luminosity remain unanswered
(Weedman 1986; Netzer 1989;
Blandford 1989; Krolik 1989;
Begelman 1989; Antonucci 1993).
In order to improve our knowledge of the Universe, of its structure and of its content, as well as of the
physics of quasars, it is evident that we need larger and larger samples in order to statistically increase the
range of observed properties. On one side, this will induce a better knowledge of these interesting objects
and, on the other side, this will increase the probability to detect strange, marginal objects that could shed
new light on the subject. Furthermore, the discovery of a large number of quasars may be a source for new
gravitational lens systems, objects presenting a high cosmological interest (see e.g. the review by
Refsdal & Surdej 1994).
Quasars are easily characterized on the basis of their spectra but a systematic spectrography at even moderate
resolution of every object in a wide region of the sky would be exaggerately telescope-time consuming. A
preliminary selection of quasar candidates in wide fields is thus necessary in order to significantly reduce the
amount of objects to observe spectroscopically. For this reason, we initially performed in a few fields a
selection of quasar candidates on the basis of the U/B excess criterion. However, this kind of survey
contains well-known
caveats, the most famous being the cut-off at z 
2.3 corresponding to the appearance of the
Ly
forest in the U band and of the relevant emission line in the B one. Besides this well-known
effect, other causes of incompleteness exist. An example is the presence of the Mg II emission line
(
in the B band for 
(Wampler & Ponz 1985). For these reasons, it is necessary to cover the same areas with
other survey techniques in order to improve the density of known quasars and to achieve a level of
completeness as high as possible. One field (around NGC 520) has already been covered by an objective-
prism
survey (see Gosset et al. 1990).
Most of the fields are also partly covered by a deep grism survey (see a preliminary report by
Gosset et al. 1991).
We also initiated more sophisticated multicolour 
surveys (see a first brief report by
Moreau et al. 1994). These other techniques are also addressing different populations of
quasars, particularly through the range of accessible redshifts. In order to reach our aims concerning the
density of known objects and the completeness, it is clear that multitechnique surveys are essential to attain a
high level of efficiency at any redshift. Therefore, we decided to search all our selected fields with the largest
possible diversity of techniques (adding, as criteria, the absence of proper motion
, variability, X-ray emission,
etc.). Each of these isolated techniques represents a large amount of work and a detailed account of the
individual results is necessary in order to later investigate and efficiently compare all the techniques which
were used.
In the present paper, we report on the U/B excess technique as applied in a field around the galaxy NGC
450 - a galaxy possibly anomalous
(Arp 1987, Moles et al. 1994) -. This is a rather old survey initiated more
than fifteen years ago. Although we are currently working on the different extensions, we prefer to publish
the U/B survey separately both for practical and historical reasons, thus preserving its homogeneity. In
Sect. 2, we describe the survey itself and the selection of the quasar candidates. Section 3 deals with the
spectroscopic and photometric observations necessary to the project whereas Sect. 4 is devoted to the
astrometric and photometric calibrations. Section 5 details a few results of the present quasar survey and
Sect. 6 contains considerations on the
spatial distribution of the discovered quasars. Section 7 concludes the present paper.