This chapter will focus on those aspects of polarimetric instruments that are peculiar to certain wavelength regions. The concepts discussed in previous chapters will be used freely. Non-polarimetric wavelength-peculiar concepts will generally be taken for granted, but a few are essential and must be recapitulated briefly.

Optical/infrared systems

Optical polarimetric instrumentation has a long history of development. Early polarimeters had errors at the level of a few tenths of a per cent at best, and polarization signals were small, so that polarimetry was very much a specialist craft. B. Lyot was the first to obtain very high accuracy by devising a modulator and using it on the Sun. For stars, the signals were generally so small that photon shot noise was appreciable, and there was little incentive to design sophisticated systems of unavoidably smaller throughput.

The situation has changed drastically within the last decade or two. Larger telescopes are available, CCD detectors now offer thousands of parallel channels of potentially very good accuracy, and improved modulators of high transmission have been devised. The higher signal levels have meant that greater resolution (spectral, temporal, spatial) can be used, and this has had the effect of increasing the degree of polarization provided by nature (less smearing of polarizations from neighbouring resolution elements); the end result is that (i) many more situations within astronomy can usefully be tackled by polarimetry without exceptional cost in telescope time and (ii) ‘common-user’ polarimetry is becoming available in the optical/near-infrared wavelength region (the ‘CCD domain’).