Aligned grains provide one of the easiest ways to study magnetic fields in diffuse gas and molecular clouds. How reliable our conclusions about the inferred magnetic field depends critically on our understanding of the physics of grain alignment. Although grain alignment is a problem of half a century standing recent progress achieved in the field makes us believe that we are approaching the solution of this mystery. I review basic physical processes involved in grain alignment and show why mechanisms that were favored for decades do not look so promising right now. I also discuss why the radiative torque mechanism ignored for more than 20 years looks right now the most powerful means of grain alignment.

Grain alignment theory suggests that grains should be aligned in circumstellar regions and the observational data available supports this conclusion. We discuss the alignment of grains via (1) magnetic relaxation, (2) mechanical processes, and (3) radiative torques. We show that ferromagnetic relaxation is likely to be more important than superparamagnetic relaxation if the dust in circumstellar regions is similar to species recently captured in Earth atmosphere. Outflows and stellar winds provide grain streaming along magnetic field lines and therefore mechanical alignment competes with the ferromagnetic and radiative alignments. We estimate measures of grain alignment in circumstellar regions, comets and interplanetary space and conclude that in many circumstellar regions and in the interplanetary space radiative torques may constitue the major alignment mechanism which aligns grain longer axes perpendicular to the direction of magnetic field. Observations in submillimeter and microwave ranges are suggested as a means of disentangling effects of multiple scattering from those related to aligned grains.

This paper gives a short summary of a mechanism of the enhanced paramagnetic relaxation that acts on rapidly rotating interstellar grains. We show that the Barnett magnetization that arises from grain rotation ensures that paramagnetic absorption happens at its maximum efficiency, i.e. the conditions for paramagnetic resonance are automatically fulfilled. The differences between the predictions of classical Davis-Greenstein relaxation and the process which we refer to as ``resonance relaxation'' are most pronounced for grains rotating faster than 10 GHz. Microwave polarization is likely to be an impediment for cosmic microwave background studies, but can provide a good tool for studying galactic magnetic field.

I review my work with Bruce Draine on dust emissivity at microwave frequencies (3 cm - 3 mm). This emissivity explains the recently detected "anomalous" component of the galactic foreground emission. Both small (a<0.001 micron) and large grains contribute to this emission. Small grains have electric dipole moments and emit while they rotate; the microwave emission of large grains is mostly due to magneto-dipole radiation. Most efficient magneto-dipole emitters are strongly magnetic, e.g. ferrimagnetic or ferromagnetic, materials. The relative role of the two mechanisms can be established through observations of microwave emissivity from dark clouds. New microwave window is a window of opportunity for interstellar studies. Magnetic fields inside dark clouds may be successfully studied via microwave polarization. Microwave emissivity constrains the abundance of strongly magnetic materials. For instance, the available data at 90 GHz indicate that not more than 5% of interstellar Fe is in the form of metallic iron grains or inclusions (e.g., in ``GEMS''). Future missions, e.g. MAP and PLANCK, will bring a wealth of microwave data that can be successfully used to study ISM. Such a study would be appreciated by cosmologists who franticly try to remove all foregrounds from their data.