We discuss some characteristics of the appearance and development of magnetic fields within active regions as well as the large-scale ordering of activity into complexes of activity. It is not possible to separate a study of the evolution of active regions from a study of the model of the activity cycle. Many of the results obtained in the last few years concerning the development of active regions and large-scale activity have not been easily explained within any of the solar activity models. A chronological scheme of the development of a ‘typical’ C- or D-type active region is presented. We point out that the appearance of magnetic flux at the solar surface seems always to be a relatively rapid event, occurring during the course of a day or so. If the region does not receive more magnetic flux to make it a large region or if there is not a resurgence of activity later in its lifetime, the rest of the development is a gradual expansion and mixing of magnetic flux with the surrounding background field pattern.

Solar active regions are considered ‘anomalous’ when they belong to magnetic classes γ,βγ and βf-αf. The study of the solar activity of the region where, later on, these groups are born shows an evident correlation between the presence of an old active center and the complexity of the new active region.

It is found that the complexity is greater if the old active center is younger, and the superposition better. We also observe that the birth of anomalous sunspots groups occurs much more frequently on the western side of the magnetic inversion line of the old center.

When the birth of an active center occurs outside and on the West of the faculae, we observe the weakly anomalous groups βf-αf. The ‘perturbation’ decreases with distance and is extended at least to 10 heliographic degrees of the boundaries of the old faculae.

All the active regions observed on the Sun with the Mount-Wilson magnetograph between August 1959 and December 1962 have been given magnetic classifications in a system similar to the Mount-Wilson sunspot-classification scheme. The flare productivity of regions classified as unipolar, bipolar, and complex bipolar, as well as regions composed of multiple bipolar components has been studied. It has not been necessary to provide a classification corresponding to the γ class of sunspots. Although the relatively poor angular resolution employed in the magnetograms limits somewhat the accuracy of the data, it is clear that both complex bipolar regions and regions with multiple bipolar components produce more than three times the number of flares than the simple bipolar regions produce. The most flare-productive class of regions is the reversed polarity complex classification.

The statistical relation of the spot magnetic-field classification to the classification of the corresponding plage fields has been studied and found to be poor for these data.

The distribution of the magnetic regions in latitude shows that many of the regions with polarities reversed from the usual orientation are confined to the equatorial zone.

The magnetograph of the High Altitude Observatory station at Climax has been used to study changes in the longitudinal component of the magnetic field in a number of active regions. For a description of the instrument see Lee, Rust, and Zirin (1965). By securing simultaneous Hα pictures of the regions we compare the magnetic data with the optical appearance of the plages. From such studies we have selected the results of two active regions to present here.

According to modern theories of the solar cycle, active regions on the Sun are caused by a magnetic disturbance penetrating the solar surface from below. Sunspots, filaments, flares and other conspicuous events in an active region seem to be only secondary phenomena, the basic feature being the magnetic field itself.

Observations of the birth of some proton-flare regions show that they originate in two or more close active centers.

A graphical representation of the 66 solar rotations (Carrington) between January 1, 1962 and December 31, 1966 has been prepared. It includes all centers of activity for which the calcium plage attained an area of at least 1000 millionths of the solar hemisphere and/or intensity 3 (McMath scale). In this study the antecedents, descendents, and neighbors of each region can easily be discerned. The work shows clearly that zones of activity, apparently closely related and much larger than single plages existed for long intervals of time. For example, the significant increases in solar activity in February, May, and October of 1965 occurred in a ‘family’ of calcium plages apparently related through similarities of position and strong radio emission.

The members of ‘families’ of centers of activity are found at systematically changing longitudes. For some ‘families’ the change of longitude appears to be primarily a consequence of differential rotation; for others, the pattern of formation of active centers dominates.

According to the data for 1962–66 a meaningful study of the development of a center of activity may require consideration not only of the past history of the zone of the Sun in which it occurs but also of the zone approximately 180° away on the opposite hemisphere.

A close relation of filament feet to the supergranular structure has been found. Green corona brightness, 1420 MHz flux, and the location of quiescent filaments depend upon the distribution of the photospheric magnetic fields. The development of polar maxima of the green corona and their relation to the following polarity have been studied.

Thirty-eight series of photoelectric recordings of radial velocities in a developed active region have been studied. The recordings were made in the Fei line γ 5250 Å from September 2–7, 1961, when the active region passed across the solar disk from E 30° to W 42°.

The places of local maxima of sunspot activity, i.e. the formation of solar active regions, seem to follow in a definite sequence of tracks over the Sun's surface. These sequences of activity apparently represent distinct continuous streams, whose duration may be of the order of magnitude of 1 year.

Observations of the longitudinal component of the photospheric magnetic fields near sunspots imply that surges and Ellerman bombs occur at neutral points in the magnetic fields (i.e. |B|=0) in the chromosphere. The neutral points appear above satellite sunspots, which are defined as polarity reversals (in B||) near the edges of large-spot penumbrae. A series of magnetograph observations shows a point of satellite polarity vanishing during a period of almost continuous surge activity. The lost magnetic energy is comparable to the energy release evidenced by the surge.

The abnormal evolution of an active region during three solar rotations is studied. The high density of flares during the second and third rotation seems to be caused by the collision of new active centres with existing ones.

The increase of the activity is probably due to the disturbance of the magnetic field which became more complex because of the appearance of new centres near the original one.

The dynamics of the magnetic fields which are imbedded into the non-stationary outer layers of the Sun show many facets of interest to observers and theoreticians alike. In a short review I can only deal with a small number of them and occasionally glance at some others. I hate to call these magnetic fields frozen into a matter which is rather in a boiling state, but the electrical conductivity in these layers is high enough to keep matter and magnetic flux together for rather long times, so that we can discuss the most important questions within the framework of magnetohydrodynamics with infinite conductivity. I will first talk mainly about the layers below the photosphere, where the matter controls the motion of the field, secondly about the intermediate state near the photosphere, where matter and field have comparable energy, and finally about the upper layers where the field controls the material motion.

The strong magnetic fields observed between supergranules indicate that there must be subphotospheric convection in cells with a preferred diameter of about 30000 km. Orthodox mixing length theory assumes that the dimensions of cells are limited by the density scale-height. This is adequate for explaining granules but cannot account for supergranulation. A model is therefore proposed in which cellular motions extend over several scale-heights. In addition to granules and supergranules this model predicts a third characteristic scale of motion, with giant cells around 300000 km in diameter. These cells may produce a pattern of magnetic fields like that suggested by Bumba and Howard for complexes of activity.

It appears that it is necessary to look for oscillatory (i.e. overstable) instabilities as the cause of energy transport in sunspots. Making use of recent calculations on the hydromagnetic stability of thermally unstable layers with open boundary conditions, it is found that oscillatory modes can occur in the interior of sunspots. Some possible consequences of these oscillatory modes are discussed.

1. (1) A scheme of magnetic-field generation in an active region within the subphotospheric convective zone is discussed. Magnetic-field amplification is considered to result from the interaction of the local cyclonic motion and the turbulent motion of the subphotospheric plasma.

2. (2) A model of the magnetoplasma of an active region is considered consisting of a current-free magnetic field, whose lines of force stretch deeply under the photosphere in the form of subgranules (force-free fine-scale plasma elements). The explanation of some peculiarities in the development of an active region with the considered structure is discussed.

3. (3) Some peculiarities of the general structure of the magnetic field and the motion within photosphere and chromosphere obtained by the analysis of observations with the IZMIRAN two-channel magnetograph is discussed.

The convection in a compressible inhomogeneous conducting fluid in the presence of a vertical uniform magnetic field has been studied. It is shown that a new mode of oscillatory convection occurs, which exists in arbitrarily strong magnetic fields. The convective cells are stretched along the magnetic field, their horizontal dimensions are determined by radiative cooling. Criteria for convective instability in a polytropic atmosphere are obtained for various boundary conditions in the case when the Alfvén velocity is higher compared with the velocity of sound.

The role of oscillatory convection in the origin of sunspots and active regions is discussed.