In some recent papers the interdependence of the gas and magnetic-field motions in the solar atmosphere was considered. Some results indicate the occurrence of gas motion along the magnetic-field lines combined with motion of the field line, but sometimes we have to assume an obvious gas motion across the magnetic-field lines. As one of the possible mechanisms explaining this fact the anomalous plasma diffusion may be proposed.

To test the validity of the assumption of a ‘current-free’ magnetic field in the atmosphere above sunspots, we measured all three parameters of the magnetic field in an active region and compared it with the calculated current-free field according to the solution proposed by Schmidt. We found important differences between the calculated and the measured field.

We refer to an earlier paper where the objections to the ‘current-free field’ were that the calculated field has too simple a configuration to account for the fine structure observed in Hα and other solar observations.

Theoretical techniques are developed to study compressible, steady-state, magnetically aligned gas flows in sunspot regions. The flows are adiabatic and occur in a known streamline configuration. The non-linear parabolic partial differential equation describing the flow reduces to an ordinary linear differential equation. The solutions are briefly discussed.

The object of our investigation was to clarify the behavior of the fluctuations in the brightness in photosphere on the scale of the supergranulation 1–2 days before the appearance of sunspots.

In places on the solar surface where longitudinal magnetic fields are detectable using Leighton's photographic technique, spectroheliograms taken in the cores of many Fraunhofer lines show a bright photospheric network similar to, but with finer structure than, the familiar chromospheric network visible on Ca+ K232 spectroheliograms. This paper describes preliminary results of a study of the relation between the photospheric network and its associated magnetic fields.

Assuming the sunspot area to evolve smoothly and the sunspot-group lifetime distribution to be a monotonous decreasing function one can easily conclude that the function N(S), determining the number of sunspot groups having the area S at this moment, must increase monotonously and smoothly with decreasing area S.

An observational study of the fine structure in a sunspot region led to the following results: (a) The photosphere around a sunspot is covered with the so-called Magnetic Knots. These features have a diameter of 800 km and a magnetic field of up to 1400 gauss. Although they coincide with dark, intergranular spaces, they are distinctly different from pores. We estimate some 300 to surround the sunspot. The magnetic field of the smallest pores were found to be 1500 gauss. (b) For the Umbral Dots we find a lifetime of 25 min. Their colour, as derived from simultaneous observations at 4700 Å and 6400 Å, was found to be identical to that of the photosphere. Assuming their brightness to be photospheric we derive a diameter of 160 km.

As a preparation for the observational program of the first flight of the ‘Spectro-Stratoscope’ balloon-borne solar telescope, an attempt was made with the 35-cm aperture domeless Coudé refractor of the Capri Station of Fraunhofer Institute to obtain spectra of penumbral fine structure.

The non-linear, partial differential equations of magnetohydrodynamics are iterated simultaneously by computer to determine the time development of a single sunspot. Axisymmetry and incompressibility are assumed. The initial conditions are (1) zero velocity everywhere, and (2) the magnetic-field distribution of a ring current embedded in the photosphere. The initial magnetic field is then allowed to relax by magnetic diffusion and by the creation of a velocity field. It is shown that (1) Evershed motion outward from the sunspot will develop from reasonable initial parameters, and (2) the growth rate of the magnetic configuration depends on the strength of the initial magnetic field.

1. 1. The fine structure of the proton sunspot group of July 4–8, 1966 was studied on the basis of high-resolution heliograms. The comparison of the orientation between penumbral filaments and the transverse magnetic fields (observed by A.B. Severny and T.T. Tsap) shows that the direction of the filaments coincides in general with that of the magnetic field.

2. 2. Measurements of the magnetic fields of smallest pores (1·5″-2″) showed that the pores are always connected with strong magnetic field (in average 1400 gauss), which is localized at the same small area as the pore.

3. 3. Magnetic fields of faculae are concentrated in small elements with the dimension not exceeding 1·5″-3″. Magnetic-field strength H|| of about 45% of facular granules is within the limits of photographic measuring errors (approximately 25 gauss). For a quarter of all facular granules the strength H|| is from 25–50 gauss; about 30% of facular granules have H|| > 50 gauss, and sometimes there appear faculae with field strength of about 200 gauss. The magnetic-field strength of facular granules, which are found directly above spots, is 10–20 times less than the field strength of spots. This field is 80–210 gauss only.

4. 4. All observational data mentioned above show that the appearance of the fine-structure features in active regions is directly connected with the fine structure of magnetic field of different strength and different orientation. The study of high-resolution heliograms gives additional information about the fine structure of the magnetic field.

During 1963–67 observations of the magnetic fields in sunspots have been obtained at the Oslo Solar Observatory. For the largest spots the detailed distribution of the magnetic-field strength is found. Based on calculations of line profiles made by the author (Kjeldseth Moe, 1967) also the direction of the magnetic field is derived. Observations of the magnetic field of the same spot at several positions on the solar disk give further information regarding the magnetic-field configuration. Our results are in fair agreement with those of Bumba (1962).

According to our program of sunspot proper motion investigations (Kuklin and Syklen, 1966) we study the interdependence of the sunspot proper motions inside the group and the magnetic field of the whole group or active region. This aspect of the dynamics of matter in disturbed regions of the Sun was not considered practically up to the last time.

In the period of May-June 1965, the magnetic-field strengths of twenty sunspots were measured in order to investigate fluctuations within one day. The results of spectrograms, which were taken in the interval of one hour, are given in graphs. The mean error of one value has the size of ± 169 gauss. The graphs show the general tendency in the behaviour of field strength. Fluctuations of magnetic-field strength up to 800 gauss within a few hours were found. The short-time fluctuations shown in the graphs are mostly smaller than the mean error and therefore probably not real.

For more details see Astron. Nachr., 289 (1967), 233.

The model of the magnetic field of sunspots, taking account of fine structure of magnetic field in solar plasma, is considered. Small-scale subgranules with their own field form magnetic filaments in the external current-free field. The filaments are vertical in the umbra, while in the penumbra they run along the surface with sharp bends. In a number of spot umbra the relation between Doppler velocity and the field is established on polarized spectrograms. The π-component splitting in umbra is interpreted as a result of a weak background magnetic-field existence together with a large field of magnetic filaments. Spectrographic definition of the magnetic field in spot umbra is accomplished on the effect of magnetic-lines intensification and directly on spectrograms of low-excitation (Fe I, Ti I) and high-excitation (Fe II) lines. Magnetic field measured in low-excitation lines exceeds twice the field value obtained in high-excitation lines. This result has been considered in the light of the proposed model of sunspot field.

With the Capri magnetograph of the Fraunhofer Institute simultaneous measurements of sightline velocities and longitudinal magnetic fields have been carried out, in order to investigate the dependence of the 300-sec oscillation of sightline velocities on the magnetic-field strengths, and the oscillatory behavior of the latter, if there is any.

With the image of the Sun held fixed, time variations of solar magnetic field were recorded simultaneously in two lines (λ5250 and λ6103) with the aid of double magnetograph (for the description see Severny, 1966, examples of such records are on Figures 1 and 2). The resolution was 2″.5 × 4″.5. The records in both lines λ5250 and λ6103 covering a time interval of about 1 hour show oscillations with the period ∼7·4min and amplitudes 10–20 gauss for moderately active region (full lines on the top of Figure 2). The oscillations of the magnetic field at both levels (the heights of λ 5250 and λ6103 differ by 260 km) are sometimes synchronous but now and then phase shift appears. For quiet regions near the centre of the disk the oscillations with the period ∼ 9·0 min and amplitudes less than 5 gauss are recorded and presented in Figure 1 (full line). These 7–9 min oscillations are probably sometimes damped or interrupted by more short-period fluctuations (with the mean period 80s) of smaller amplitude, but these last oscillations could be due to the fluctuations of seeing.

Test cross-correlation functions between the magnetic field recordings and the sight-line velocity recordings with East and West relative lag for two regions near the centre of the disk have been computed. We have also found the deviations of the absolute value of the magnetic-field strength |H| from the mean absolute value of magnetic-field strength |H̄| for the whole region. The same procedure for the velocity field has been made. The cross-correlation functions for these deviations have also been computed (Figure 1).

A peculiar kind of double-peaked Hα lines in solar flare spectra is described. The most likely explanation to this phenomenon is an absorption in the centre of the line due to mottles and fibrils situated above the flare. Another explanation might be a rapid rotatory mass motion in the filamentary structure of the flare itself. In some other flare spectra a slight inclination of the Hα-emission streak has been found indicating as well that rotatory mass motions may be present in flares.

Spectra, with a radial slit in the K and Hα lines, of a sunspot close to the limb are used to determine chromospheric heights in an active region. Equidensity contours obtained with the Sabattier effect are used to measure the values of h* for different positions in the line core and wings. The spectra have also been utilized to determine similar heights in two very restricted regions of Ca+ emission which show an arched emission structure in K232 with bright continuum emission. At the K-line centre, unit tangential optical depth in the umbra, umbra-penumbra interface, and penumbra are at 1310, 1610, and 2330 km respectively. The corresponding value for the umbra in the Hα-line centre is 1080 km. The arches have a value of h = 3000 km at Δλ = 0 for the K line.

As observational material in this work we used spectroheliograms taken with the Crimean solar tower telescope in K232 and Hα filtergrams taken with the chromospheric telescope in Simeis. The Hα birefringent filter was so adjusted, that by tuning the last polaroid we could take filtergrams in the centre of Hα or combined filtergrams in the two wings at Hα ± 0·5 Å. So the effect of Doppler shifts on image-brightness distribution was diminished. We compared the brightness distribution with that of spectroheliograms.