High spatial resolution maps of the optical linear polarisation of a range of galaxies of different morphological types and inclinations are presented. Spiral, barred spiral and irregular galaxies with low inclinations show spiral patterns of polarisation orientation which may be interpreted to indicate that there are large-scale magnetic fields present in the discs of these galaxies. A survey of edge-on galaxies with conspicuous equatorial dust lanes shows remarkable differences in polarisation structure between early and late types. In the Sa galaxies the polarisation in the dust lane is sizeable and parallel to the dust lanes, suggesting that the magnetic field is uniformly distributed in the plane of the galaxy. This polarisation pattern disappears in the Sc galaxies suggesting that the relatively chaotic dust features do not possess a grain aligning mechanism, e.g. a magnetic field, that is uniform or coherent over large distance scales. It is clear that most galaxies do have polarisation patterns that are related to galactic structure and inclination, and further similar polarisation studies should help to unravel the relationship between the generation of polarisation and galactic magnetism.

At the Landessternwarte a new CCD polarimeter was constructed to observe magnetic fields (Davis-Greenstein) of nearby galaxies. This program is performed in collaboration with the MPI für Radioastronomie Bonn. By means of a focal reducer, the instrument records a field of 14.4 × 9.6 arcminutes at the 1.23-m telescope of the MPI für Astronomie on Calar Alto. Hence galaxies with large angular diameter, which have already been observed with the 100-m radio telescope Effelsberg, can be observed.

We observed a sample of nearby galaxies using the CCD-Polarimeter of the Landessternwarte Heidelberg in the R-Band with the 1.23-m telescope of the MPIA at Calar Alto. In most cases polarization was found but the polarization structures cannot be easily related to a single phenomenon. If there is a strong point source, Rayleigh scattering leads to circular patterns of the polarization vectors. If dust grains are aligned in a magnetic field, this so-called Davis-Greenstein-mechanism gives more uniformly distributed E-vectors, which can be distinguished from the Rayleigh scattering. The example of M 82 given here shows both kinds of polarized light. There is evidence of a strong magnetic field component perpendicular to the disk which seems to have its origin in the center of the galaxy. This is consistent with theoretical models as well as with previous observations.

The present status of an ongoing observational program to study the magnetic field structure of the SMC by means of optical polarimetry is presented. Being the SMC more distant than the LMC, the observed optical polarization would mainly originate in the distorted outer layers in the front side of the SMC. Highly sensitive data obtained up to now show a magnetic field aligned with the HI bridge connecting the SMC with the LMC, revealing one of the few detected intergalactic magnetic fields associated with interacting galaxies.

The model of selfpropagating star formation uses local processes (200 pc cell size) in the interstellar medium to simulate the large scale cooperative behaviour of spiral structure in galaxies. The dynamic of the model galaxies is taken into account via the mass distribution and the resulting rotation curve; flat rotation curves are used. The interstellar medium is treated as a multiphase medium with appropriate cooling times and density history. The phases are: molecular gas, cool HI gas, warm intercloud and HII gas and hot coronal fountain gas. A detailed gas reshuffeling between the star forming cells in the plane and outside the galactic plane controls the cell content. Two processes working stochastically are incooperated: the building and the decay of molecular clouds and the star forming events in the molecular clouds.

The flux from the galactic magnetic field alters radically the appropriate description of the equilibrium, collapse and fragmentation of the self-gravitating gas clouds that are the locale of star formation.

General forms of the B-ρ relation are investigated in both the isothermal and the non-isothermal regions. The magnetic flux dissipation either by ambipolar diffusion or by Ohmic dissipation has been restudied. The rates of heating due to the magnetic dissipation processes have been calculated in comparison with the rate of compressional heating.

The formation of cloud cores (or fragments) and their evolution into protostars are the inevitable outcome of the struggle between gravity and magnetic fields, with ambipolar diffusion as the agent employed to weaken gravity's fierce opponent. The very specific and crucial role of magnetic fields in star formation deduced from detailed quantitative calculations is summarized. Criteria for collapse against magnetic and thermal-pressure forces are given. Magnetic braking time scales for both aligned and perpendicular rotators, and ambipolar diffusion time scales in both quasistatically and dynamically contracting cores are presented, and their implications are discussed. The possible role of magnetic fields in the determination of the initial (stellar) mass function (IMF) is beginning to emerge. New calculations on the axisymmetric collapse of clouds due to ambipolar diffusion reveal that the relation Bc ∞ ρc1/2 between the magnetic field strength and the gas density in typical cloud cores holds even in the presence of ambipolar diffusion up to densities ~ 109 cm−3. Small masses, high densities, and strong fields observed in H2O masers are consistent with theoretical calculations.

The angular momentum is one of the major obstacles to the contraction of interstellar clouds. An efficient process of removing the angular momentum from the cloud is via transport along the magnetic field lines to the ambient medium. When the magnetic field is nearly uniform and the direction of the field lines is parallel to the rotation axis, the spin-down time of the cloud is given by σ/2ρVA, where σ is the column density of the cloud along the field lines, and ρ and VA are the density and the Alfvén velocity, respectively, in the ambient medium (Ebert et al. 1960; Mouschovias & Paleologou 1980). However, this is for a cloud with weak gravity. Because a cloud with strong gravity has contracted dragging the field lines, the ambient field is considerably distorted from uniformity. The spin-down time of such a cloud is shorter than given above (Gillis, Mestel & Paris 1974, 1979).

We summarize the results of 3-D MHD calculations for a rotating and self-gravitating interstellar cloud. The angle between the axis of rotation and the magnetic field direction varies between π/2 and 0, the corresponding braking time-scales differ by roughly a factor of 3.

Strong circular polarization of OH masers at 1665 and 1667 MHz lines has been observed towards the molecular cloud cores associated with HII regions. Magnetic field strengths of a few mG are derived from the Zeeman splitting of OH lines. For instance, a magnetic field of about 4 mG in the masing region of W3(OH) has been estimated by OH-line Zeeman splitting (Davies, 1974). VLBI observations show that the OH maser spots project onto or very close to the surface of associated compact HII regions (Reid et al., 1986). The observational evidence demonstrates that the scales of OH maser components surrounding a compact HII region (R ~ 1016 cm) are about 1014 cm in diameter with an amplification pathlength of ~1015 cm. Hence the magnetic fields determined by the Zeeman splitting of OH maser lines appear partly very close to the associated HII region. Elitzur (1979) has theoretically obtained similar results as above.

A small but growing body of observational information now exists regarding magnetic field strengths in molecular regions. Most of these data come from study of the Zeeman effect in 18 cm OH lines. The field is strong enough in many such regions to be dynamically important.

Zeeman splitting in the spectra of molecular masers that arise in the envelopes of newly formed stars (called interstellar masers) offers a method of estimating the magnetic field in these dense regions (106–1011 cm−3). The magnetic field strengths in regions of OH and H2O masers scale as the square root of the density when compared to the density and field of the ambient interstellar medium. The direction of the field in OH masers may help trace the large scale structure of the magnetic field in the Galactic spiral arms.

We have done detailed calculations of the Zeeman effect in the dozen diatomic molecules identified in interstellar clouds.

We present results of recent circular polarization experiments with the MPIfR 100-m telescope, revealing for the first time, the magnetic field strength towards interstellar H2O masers and the dense cores of local dark cloud complexes. Weak Zeeman splittings of a few 10 kHz only in the 22.235 GHz maser transition of the non-paramagnetic H2O molecule imply magnetic field strengths of ~ 50 mG in the dense (n ~ 1010 cm−3) masing layer. With the recently identified CCS radical it became possible to study the magnetic field associated with dense (~ 105 cm−3) dark cloud cores, the potential sites of future star formation. We report the detection of a −110μG field towards TMC-1C, a low-mass core associated with the Taurus Molecular Cloud. From complementary gas density and kinetic temperature probing measurements, we derive approximate equipartition between magnetic, gravitational and thermal energy for this clump.

In the last few years our knowledge of the strengths and directions of magnetic fields in molecular clouds has increased significantly. This paper briefly reviews observations of magnetic field strength and direction, compares such observations with equipartition models, and discusses the implications of these comparisons for the typical internal structure of the field in a molecular cloud.

The periodic structure we [1,2] have previously reported as puzzling in the velocity field and column density of 21 cm self–absorption in the Taurus Complex, can be explained naturally in terms of the gravitational collapse of the HI along the lines of force of the large–scale magnetic field of the region.

The importance of magnetic fields to the evolution of dense interstellar clouds and to the star formation process is now widely appreciated. Troland (this volume) has reviewed work to measure strengths of interstellar magnetic fields by observation of the Zeeman effect in radio frequency spectral lines. The next step is to map magnetic fields with radio synthesis arrays in order to obtain high spatial resolution. This abstract is a preliminary report of VIA results for two clouds - the W 3 region of massive star formation (to be published by Troland, Crutcher, Goss, and Heiles) and S 106, a biconical H II region with a confining molecular disk (to be published by Loushin, Crutcher, and Troland).