The different methods which have been used, or which may be used in the future, to measure solar magnetic fields are described and discussed. Roughly these can be divided into three groups (a) those which use the influence of the magnetic field on the electromagnetic radiation, (b) those which use the influence of the field on the structure of the solar atmosphere (MHD effects), and (c) those which use theoretical arguments. The former include the Zeeman effect, the Hanle effect, the gyro and synchrotron radiations and the Faraday rotation of radiowaves. The second includes the alignment of details at all levels of the solar atmosphere, and the calcium network, and the third makes use, for example, of the assumption of equipartition between magnetic and kinetic energy density.

The magnetograph is based on a high-resolution filter which serves in place of a spectrograph, except that a reasonably large field of view (one-quarter of the Sun's diameter) can be observed at the one instant. Observations are made by obtaining filtergrams of opposite circular polarizations simultaneously in the wing of a magnetically sensitive line. Exposure times are about 0.3 s, the angular resolution of the magnetic field is about 2 arc s, closest frame repetition rates about 8 s. The filtergrams are processed subsequently by photographic or television subtraction. Semiautomatic photographic and/or TV subtractions yield magnetograms suitable for cinematographic projection though the subtractions are not yet as perfect as those obtained by individual subtraction.

A scanning photoelectric polarimeter of 16 cm aperture that can measure all four Stokes parameters of the visible radiation of the Sun's disk, the Corona, Moon and planets, has been constructed at the Institute for Astronomy and is installed on Mt. Haleakala. It is a two (orthogonal) channel system and uses a rotating λ/4 plate modulator. Photon counting is done by a digital computer that also Fourier analyzes the modulated output of the photomultipliers, and, from the Fourier components, computes the Stokes parameters in real time.

It is demonstrated that certain interrelations between the Stokes parameters are much less dependent on line formation than generally expected. Thus the dependence of these parameters on the magnetic field may be made more explicit and may lead to reliable calibration of polarimetric measurements in terms of magnetic field.

The Aerospace – NASA Videomagnetograph began operation one month ago, two years after components were ordered and construction began. The design grew out of a desire to obtain magnetic fields in real time using an optical filter. The aim was to study and analyze magnetic configurations and changes, quantitatively if possible, with high spatial and temporal resolution and as much sensitivity as possible. This instrument is restricted to the line-of-sight component of the magnetic field and is primarily intended for high resolution studies of selected regions of the sun. The rationale behind our approach is shown in the next section and the design details in the following.

The 40-channel magnetograph is described. Special attention is given to details of the fiber-optic probe which dissects the Fraunhofer line image and provides for a choice of spectral and spatial resolution. Examples of magnetograms taken with the instrument are shown. Maps of quiet areas, made with long integration times to achieve a noise of 0.4 G, yet with 5 arc-sec resolution, indicate the presence everywhere of a minimum background field of average strength 2–3 G.

An oscillating magnetic analyzer (KDP crystal plus Glan-Thompson prism) is coupled to an echelle-interferometer spectrograph. The single slit magnetometer by pressure variations can be made to scan the entire profiles of the circularly and linearly polarized Zeeman components. Freon gas is used as the scanner gas with wavelength displacements of 0.02 Å per 0.1 in. Hg pressure change at the NaD lines. The available scan range is 15 Å in the visual spectral region.

The Sac Peak magnetograph (DZA) has been modified from Evans' original scheme so that it measures the displacement of the right and left hand circularly polarized lines separately. The computer reduction calculates the Zeeman and radial velocity signals. A grating servo system has been added to correct for slow temperature drifts in the spectrograph. A paper-tape reader controls the raster scan and the formatting of data on to magnetic tape.

The influence of some ‘effective’ asymmetry (miscentering) of spectral line on the accuracy of the complete field vector is considered for the measurements with the Crimean magnetograph. This miscentering appears in the device for H⊥ records if the longitudinal field is strong.

For many years solar magnetic fields have been measured by a variety of techniques, all of which exploit the Zeeman splitting of lines in the solar spectrum. One of these techniques (Leighton, 1959) involves a photographic subtraction of two monochromatic images to produce a picture of the Sun in which the line-of-sight component of the solar magnetic field appears as various shades of gray. In a magnetogram made by this method, zero field strength appears as neutral gray, while magnetic fields of one polarity or the other appear as lighter or darker areas, respectively. Figure 1 shows such a magnetogram.

The optics and electronics of a new filter magnetograph will be described. The instrument uses a Zeiss 0.13 Å birefringent filter to isolate magnetic sensitive lines. All four Stokes parameters can be measured. A Westinghouse SEC vidicon WX 30 654 serves as the detector. The data are completely digitized and transmitted in real time into a Univac 1108 computer.

Telescopic phase retardation in connection with the polarizer behind a polarimeter's KDP-crystal strongly influences the Zeeman pattern, since the circular polarized parts of the two σ-components are strengthened and weakened respectively. This influence depends on the hour angle if the solar image rotates with respect to the polarizer (e.g. for Coudé telescopes).

A method of reducing circular parasitical polarization in the antennas of variable profile with the help of a grating of curved wires is reported. The experimental verification of the method shows that the parasitical signal practically vanishes. It was established that using this method, simultaneous observations at three or even five wavelengths can be made.

Some possible developments of magnetic beam absorption devices are described, in view of their very promising possibilities of practical application in the field of solar magnetic field observations.

The theory of line formation in a magnetic field is reviewed. It is shown how the formulations by Unno, Beckers, Stepanov and Rachkovsky are related to each other. The general treatment of true absorption, anomalous dispersion, radiative scattering and level-crossing interference is discussed. Special attention is paid to the inhomogeneous nature of the solar atmosphere. The properties of magnetic filaments are reviewed. It is shown how a filamentary structure will drastically influence the interpretation of magnetograph observations. The treatment of line formation in a turbulent magnetic field is also discussed.

The fine structure of the solar atmosphere will influence different types of magnetographs in entirely different ways. The relation between the observed magnetic field and the resolution of the instrument is discussed for a Babcock-type, Evans-type and a transversal magnetograph. Finally some suggestions for future work in this field are listed.

The scattering of radiation in the presence of weak magnetic fields can give rise to coherence or interference phenomena that will profoundly affect the frequency, geometric, and polarization properties of the scattering event. In this paper we discuss and illustrate some of the features of the coherence phenomena associated with the scattering redistribution for the normal Zeeman triplet. The frequency dependent as well as the frequency independent scattering function is considered in a linear polarization basis. In addition we illustrate some properties of this redistribution function in the Stokes representation. Since the primary purpose of this paper is to demonstrate the nature of some of the properties of the coherence problems, that might be important in the interpretation of magnetic fields from polarization measurements of scattered radiation, it has been necessary in this initial work to neglect several features of the problem which are noted in the paper and are currently under investigation.

The formation of the resonance doublet lines of Li6 and Li7 in sunspots is discussed. It is shown that the magnetic splitting of the lines must be determined according to the (partial) Paschen-Back pattern. As a first approximation to the problem a detailed calculation of the line profile is given for the case of pure absorption and local thermodynamic equilibrium.