2.4. How well does the galactic dynamo theory conform to the observations?

The answer to this question is mixed; the large scale magnetic field geometry of `well-behaved' spiral galaxies agrees with some, though not all, of the straightforward predictions and model simulations of linear dynamo theory. The affirmative side of the answer is that the predominant field modes do seem to be the simplest ones, as illustrated by recent data shown for M83 and NGC 1068 in figures 1 and 2. This is probably the most significant result, in that it shows that, if the observed large scale fields were amplified by the - dynamo in the disks, the prominent modes are the low order ones, an outcome which is not a priori obvious. However, the angular resolution for many of the galaxies is probably not good enough to confirm definitive agreement with the standard - dynamo theory. Apart from `geometrical' tests, observation-based estimates of magnetic field strength indicate that, if the observed fields were entirely the effect of a galactic dynamo, the (non-linear) amplification must saturate at 2-8 µG, which is when the magnetic energy density _m(= 10^-12(|B| / 5 µ G)² erg cm^-3) becomes comparable with _cr and _t.

Recent images of improved sensitivity and resolution are beginning to elucidate some further details of the magnetic structure near the plane of spiral galaxies. The highest degree of magnetic field ordering generally occurs in the outermost regions of galaxy disks, and the interarm regions. Conversely, field orientation is generally more randomized in the star-forming complexes within spiral arms, and also possibly at smaller galactocentric radii. Specifically, field disordering tends to correlate with locally enhanced H (seen optically) and CO 1-0 ( 2.7 mm) emission (Bajaja et a1 1990), and sometimes enhanced H I ( 21 cm) surface brightness. Where the field ordering is high, the field orientation usually conforms to the galaxy's spiral pattern (figures 1, 2). Most or all of these properties and correlations have been described in the analysis of recent surface polarimetry observations of NGC 6946 (Ehle and Beck 1993), M83 (Sukumar and Allen 1990, Neininger et al 1991), and M51 (Beck et al 1987, Scarrott et al 1987 (optical polarization), and Neininger 1992).

The unprecedented detail of these recent observations has provided some of the first clear evidence for local deviations or `disruptions' of the otherwise ordered azimuthal disk component (Sukumar and Allen 1990, Horellou et al 1992). In M31, the suggested detection of arc-like looping of field lines between tie-down points 2 kpc apart, where the field lines bulge out of the disk, is possibly due to a Parker-Jeans instability (Beck et al 1989). Such phenomena may be the nascent beginnings of magnetic field expulsion. On a simple energetic calculation, where more intense than normal star formation occurs, a `breakout' of field into the local halo and IGM will result. This type of evidence reinforces the notion that such instabilities will `snuff out' the slow field strength build-up by the conventional - dynamo process in an assumed ever-quiescent galaxy disk. This, potentially, puts the conventional - dynamo theory in great difficulty if the dynamo is required to explain the amplification of very weak seed fields in the early universe up to present-day µG levels. Parker's modified - dynamo (section 2.3.6), although not yet incorporated in a detailed numerical model, might overcome this shortcoming, by allowing for a fast dynamo amplification of weaker seed fields, if indeed the protogalactic fields were much weaker.

Another test of dynamo models which require a slow field build-up over cosmic time is to estimate the magnetic field strengths in low-mass and dwarf irregular galaxies, whose lower mass and rotation might be expected to cause field regeneration to a different, lower value than large spiral galaxies. Observations of two such galaxies, the Large and Small Magellanic Clouds (LMC and SMC), indicate (using the assumption of equipartition) interstellar field strengths at the few microgauss level (Loiseau et al 1987 (SMC), Klein et al 1989, 1993 (LMC)). Allowing for uncertainties in the estimates, these field strengths are essentially the same as measured in the Milky Way using pulsars, and in other large spiral galaxies. The similarity of the magnetic field strengths in these low mass, slowly rotating galaxies thus casts doubt on the requirement of a slow-acting - dynamo to regenerate an initially weak primordial field to a few microgauss over several billion years.