Observations are probably still too scarce to reveal the magnitude and distribution of extragalactic magnetic fields. Kronberg (1994) has extensively reviewed all available observations based on synchrotron radiation and its Faraday rotation, and has proposed several properties of which the following are outstanding:
a) Typical values of intergalactic magnetic field strengths are in the range 1-3 G. These are larger than previously thought, so their influence on a large variety of phenomena must be revised.
b) These values are found nearly everywhere and are noticeably independent of the density of the zone observed. They are found in cluster cores, in clusters and in regions between clusters (e.g. between Coma and A1367, Kim et al. 1989). The first measure of magnetic fields in the intracluster medium was reported by Vallée et al. (1987), finding 2G in A 2319. Feretti et al. (1999) have obtained a field strength of between 5 and 10 G in Abell 119. In superclusters, high value strengths have been reported (e.g. Vallée, 1990, finding about 2 G in the Virgo Supercluster for the ordered component of the field).
Kronberg has speculated about a ubiquitous magnetic field. In some particular objects, such as radiosources, magnetic fields can be much higher, but this 1G background field seems to be ubiquitous. The value of 3G is particularly interesting since then the magnetic energy density equals that of CMB, thus suggesting an equipartition of both energies. Note that both energy densities decrease as R-4 (being R the cosmic scale factor), but this equipartition, if it exists at all, cannot be primordial, as argued below. However, magnetic fields have never been reported in the large-scale 100 Mpc sized voids. Only Vallée (1991) has searched for an excess rotation measure in the Bootes Void, estimating that the magnetic field strength was less than 0.1G.
c) Magnetic fields of this magnitude were also present in quasar absorption line clouds, usually interpreted as pregalactic systems. Therefore pregalactic clouds were magnetized as much as present galaxies. Field strengths of this order have also been measured at redshifts 0.395 and 0.461 (Kronberg, Perry and Zukowski, 1992; Perley and Taylor, 1991).
We favour another global picture that is fully compatible with observations but slightly different to Kronberg's view of ubiquitous 1G field strength. This global picture is also based on our own theoretical work, which will be commented later. We assume that magnetic fields vanish, or have very small strengths in the large-scale voids, in agreement with Vallée (1991), i.e. in most of the volume of the Universe, but are much higher in the filaments of matter (100 Mpc long, 10 Mpc thick) characterizing the large scale structure. Therefore, magnetic fields would be neither ubiquitous nor in energy equipartition with the CMB, but, in any case, they are high, about 1-3G, in the medium surrounding nearly all galaxies. The medium around galaxies should have 1 G strengths because this is the value at the particular sites where galaxies lie.
More recently, the review by Eilek (1999) confirms the existence of G field strength in clusters (even in cluster halos), being much higher at the centre ( B || 50G in M87, for instance).