Both the theory and observations of extragalactic magnetic field strengths and structures have made great progress in the past five years; in this section we give an overview of observational methods and results, and of the `standard' dynamo theory of field amplification in galaxy disks, and compare theory and observation of galaxy-scale fields.
Current research into galactic magnetic fields is aimed at answering the following specific questions and puzzles.
(i) How was the observed large scale organization of galactic magnetic fields achieved?
(ii) How do the large scale patterns of magnetic fields correlate with other aspects of galaxies, such as star formation rate, overall dynamics, molecular gas distribution, strength of outflow, etc?
(iii) To what extent, and by what physical mechanisms do disk galaxies amplify their large scale magnetic fields, and what determines the limiting field strengths achieved over a galaxy's lifetime?
(iv) What were the pre-galactic seed field strengths; were they of order 10-11 G or fainter, or did the galaxy form in the presence of pervasive magnetic fields of order to 10-8 to 10-6 G levels? If the initial field strengths were 10-8 G or weaker, dynamo amplification, an exponential process, is required. However, if microgauss fields already existed at protogalactic epochs, then large scale dynamos may only have influenced the geometry more than the strength of magnetic fields in already formed galaxies.
(v) Were galaxies the prime generators of intergalactic magnetic fields in the first place - through outflow of magnetoplasma over cosmic time - or were significant fields already generated in the pregalactic era?
(vi) Did the degree of ordering and strength of the interstellar magnetic field have a strong influence on the locations and dynamics of starforming regions?
Some important clues are beginning to emerge from recent, detailed observations of nearby galaxies, mostly in the radio. In the following, we review the various observational means available of detecting and measuring the galactic magnetic fields, and summarize what these data reveal. We then summarize the basic dynamo theory of field regeneration, and compare its predictions with current observations. We discuss their successes and shortcomings in reproducing observed galaxy-scale fields. We conclude that a slow buildup of large scale galactic fields over cosmic time from very small initial seed fields now appears less likely to explain observations of galaxy-scale magnetic fields than was first thought.
The Milky Way and other late type galaxies are permeated with (i) cosmic
ray gas, (ii) magnetic fields, which generate polarized synchrotron
radiation (iii) ionized (and neutral) interstellar gas which causes
Faraday rotation in the radio, and (iv) interstellar grains which align
with the interstellar field and induce linear polarization in the
optical starlight by selective absorption or scattering. These four
ingredients make it possible to observationally trace the large (and
small) scale magnetic structure in galaxies. Component (i) (represented
by equation (1.1)) has an intrinsic polarization of up to
75% in a completely
aligned magnetic field. We note that the
observed degree of linear polarization in a 2D projection (what is
observed) does not necessarily reveal the true 3D degree of alignment of
the magnetic field, nor the sign of the local magnetic field
vector. More specifically, observations of polarized synchrotron
emission do not distinguish between a zone of true unidirectional,
aligned field, and one where the magnetic field lines are stretched (or
compressed) in one dimension, but are systematically reversing their
sign (e.g. for a disk seen edge-on - a point elucidated by
Laing (1981)).
Suitable observations of (i), (ii) and (iii) can reveal the sense of the
magnetic field, and also the strength of the line-of-sight field
component (equation (1.3)) if ne, is independently
measureable.
Optical polarization observations relevant to (iv) have made great
strides due to the recent availability of imaging CCD polarimeters (cf
Scarrott 1991).
Optical polarization is a tracer of the interstellar magnetic field
direction, as was first shown by
Davis and Greenstein
(1951).
In the `Davis-Greenstein effect', non-spherical interstellar grains
will align their angular momentum vector with the local magnetic field
direction, thereby causing direction-dependent absorption and scattering
of the incident starlight. Their detailed interpretation requires some
additional knowledge, or assumptions, about the size and scattering
properties of interstellar dust and electrons. Interstellar optical
polarization can be caused by simple scattering. and also by dichroic
extinction i.e. differential transmission for X- and
Y-polarized
photons due to interstellar grains which are aligned with the local
galactic magnetic field
(Scarrott et a1 1986).
Observation of the wavelength dependence of the optical polarization at
a given point in the galaxy can distinguish 'signature'
p(%) - 
curves (Zerkowski curves) due to these different effects -- for which
the grain size-range and composition needs to be known, or modelled. A
notable advantage of optical polarimetry is the higher resolution than
what is possible with current radio techniques. Thus, optical
polarization data can detect small-scale fluctuations in galactic
fields. As with polarimetry of the synchrotron radiation, and in
contrast to Faraday RM imaging, optical polarimetry does not convey
information about the field sign.