Annu. Rev. Astron. Astrophys. 1996. 34: 155-206 Copyright © 1996 by Annual Reviews. All rights reserved

9. LAST WORDS

We have attempted to draw together various strands contributing to our current understanding of galactic magnetism. We feel that neither dynamo nor fossil theory is at present in a satisfactory state. Nonetheless, we believe that, while the problems with the primordial theory are quite fundamental, ways of resolving the difficulties of the dynamo theory exist, in principle at least. A primordial field may nevertheless be important; for example, it can provide a seed field for a dynamo (see Section 5).

We note the following. Axisymmetric spiral structures and more complicated field structures arise naturally from dynamo models. Pitch angles lie in the correct range. Dynamo models give generically plausible large-scale spatial field structures, which are in some cases quite realistic, and which readily allow the detailed modeling of specific galaxies. Finally, on general grounds, field strengths of order the equipartition value, as observed, seem explicable. These points support our view that a coherent explanation of galactic magnetism will only be achieved via the further development of some form of dynamo theory.

It is now possible to include realistic models of the ISM, including detailed data on the spatial distributions of turbulent velocity and scale, the vertical gradient in the overall galactic rotation, galactic fountains, etc, in dynamo models; however, this remains to be done. A detailed comparison of theory with observations is becoming increasingly, both because the theory is beginning to give results that are sufficiently generic, reliable, and detailed, and because observations have reached the stage where they can seriously constrain many aspects of the theory. Reliable and high-resolution information about the complex magnetic structures found in the disks and halos of spiral galaxies is needed, together with an improved theory of depolarization mechanisms.

ACKNOWLEDGMENTS

The authors acknowledge the hospitality of the Observatory of Helsinki University, where the work was initiated, and of Nordita (Copenhagen), where it was finished. We are grateful to EM Berkhuijsen, who provided the data used in Figure 5 prior to publication. Partial financial support from the NATO grants CRG921273 and CRG1530959 is acknowledged. AS thanks the Mathematics Department of Manchester University and the Max-Planck-Institut für Radioastronomie for their hospitality during his work on the paper. AS and DS acknowledge partial financial support from grants 93-02-3638 and 95-02-03724 from the Russian Foundation for Basic Research and MNP000/300 from the International Science Foundation.