The Pioneer-6 and Pioneer-7 space probes carried charged-particle telescopes which measure, for the first time, both the direction of arrival and differential energy spectra of protons and alpha particles. The intensity changes, directional distributions and energy spectra of proton fluxes associated with solar activity are investigated. The data were obtained in the beginning of the new solar cycle (no. 20), when it is possible to unambiguously associate proton-flux increases with specific solar active regions. The origin, possibly long-term storage, and propagation of these proton fluxes are investigated. It was observed that enhanced 0·6–13 MeV proton fluxes associated with specific active regions were present over heliographic longitude ranges as great as ~ 180°. These enhanced fluxes exhibit definite onsets and cut-offs which appear to be associated with the magnetic-sector boundaries observed by Ness on Pioneer-6. Discrete flare-produced intensity increases extending in energy to more than 50 MeV are observed, superposed on the enhanced flux. These increases displayed short transit times and short rise times. Both the enhanced and flare-produced fluxes propagate along the spiral interplanetary magnetic field from the Western hemisphere of the Sun. From these observations we are led to a model in which the magnetic fields from the active region are spread out over a longitude range of 100–180° in the solar corona. The existence of strong unidirectional anisotropies in the initial phases of flare-proton events implies that little scattering occurs between the Sun and spacecraft. However, the gradual approach to an isotropic flux at late times indicates that the decay phase is controlled by the interplanetary magnetic field.

The observations of extreme ultraviolet (EUV) emission lines of Feix through Fe XVI made by OSO-I have been applied to a study of the solar corona above active regions. Ultraviolet and radio emission are determined for several levels of activity classified according to the type of sunspot group associated with the active region. Both radio emission and line radiation from Fe XVI, the highest stage of ionization of Fe observed, are observed to increase rapidly with the onset of activity and are most intense over an E spot group early in the lifetime of the active region. As activity diminishes, radiation from Fe XV and Fe XIV becomes relatively more prominent. Preliminary X-ray data from OSO-III obtained during a flare are introduced. These indicate that radiation from the highest stage of iron thus far observed, Fe XXV, reaches a maximum first in an X-ray burst and that maxima in lower stages of ionization follow, with delays from 2 to 15 min.

A comparison between XUV and Ca-K spectroheliograms for 9 dates from 1963 to 1967 showed an excellent correlation between plage intensities in Ca-K and He II 304 Å, except for plages near the limb and a few others. Around the limb all but the highest ionization XUV emission lines form a bright ring, usually weaker over the poles. This is an unresolved combination of the limb-brightened emission from the quiet corona and high chromosphere, and emission extending into the corona above plages located as much as several days from limb passage.

In Fe XV and XVI only the localized coronal emissions are observed; these vary in form and intensity from line to line. The 171–500 Å and white-light coronas, recorded on November 12, 1966, correlate well at low altitudes, but beyond 3′ the XUV corona becomes diffuse and without structure.

The slowly varying component of solar X-rays has been studied through records of the satellite Explorer-30 instrumented by the U.S. Naval Research Laboratory. Correlations with plage areas, and flux at radiofrequencies have been studied. High flux values in the 8–20 Å and, still more, the 1–8 Å bands are related to specific ARs, also characterized by great flare productivity, high intensities on 3 cm, and anomalous magnetic structure.

The coronal X-ray emission has been studied in a series of sunpointed Skylark-rocket flights over the period August 1964 to August 1967. Two types of instrument have been used, namely, an array of pinhole cameras, to study the distribution of X-ray sources across the solar disk, and a set of Bragg crystal spectrometers, to examine the detailed spectrum of the emission.

Six rocket flights in all (Skylarks 301, 302, 303, 306, 307, and 406) have provided solar X-ray photographs, covering the wave-band 12–60 Å. The main conclusions of this work may be summarised as follows:

1. (1) Of over twenty separate calcium-plages areas, all those having an area-intensity product greater than 300 were associated with enhanced coronal X-radiation, this ‘threshold’ being merely a function of X-ray camera sensitivity.

2. (2) The plages studied were of from 1st to 4th rotation, indicating the enduring nature of the local coronal X-ray enhancement.

3. (3) Crude spectral analysis, derived from different filters used in a given camera array, showed an increasing contrast between the active region and disk emissions below 25 Å, indicating the probability of the active region coronal material being rather hotter.

4. (4) Photographs of regions on the disk and on the limb showed a typical coronal active region to produce more than 50% of its total X-ray flux from a region within 1 arc minute cube. Electron densities up to 10 times the normal coronal values are required by the measured X-ray brightness of the larger regions.

5. (5) The major part of the emission above 20 Å arises from the (cooler) background corona. Examination of this coronal disk emission revealed strong limb brightening, polar darkening (particularly in the 1964–65 photographs) and a band of slightly enhanced X-radiation near latitude 35 °N. The latter observation was noted on successive flights, in October 1965 and February 1966, and clearly followed the tilt in aspect angle of the solar disk.

Skylarks 304 and 305 were equipped with uncollimated Bragg crystal spectrometers. The former payload, flown in May 1966, revealed strong line emission down to the scan limit at 11 Å. Emission lines of O VII and VIII, Ne IX and X, Fe XVII and XVIII and NI XIX have been identified. All but the O VII series were found to have dominant components arising from a large active region on the solar disk, and an analysis has shown that the complete active-region spectrum is consistent with an electron temperature of 3 million degrees. Intercomparison of the different emission lines indicated coronal abundances of iron, nickel and neon comparable to those derived earlier by Pottasch from an analysis of the solar UV spectrum.

Provisional analysis of the latest flight (Skylark 305 in August 1967), employing crystals of beryl and EDDT in addition to the KAP flown on Skylark 304, has shown strong emission lines of the helium-like ions Mg XI and Si XIII, with weaker but significant intensity in Lyman-α of Mg XII (8·4 Å) and Si XIV (6·2 Å), the latter emission requiring the presence of active region material at a temperature of near 107°K.

A measurement of the solar X-ray spectrum in the 1–30 KeV interval was made by rocket-borne instruments at 0015 UT on April 5, 1967. These measurements show that the solar X-ray spectrum, which falls off steeply with energy below 4 KeV, becomes less steep above this energy. These data are taken to indicate that energetic, non-thermal, electrons are present or continuously produced in quiescent solar active regions. No optical flare activity was reported for several hours around the time of these observations.

It is well known that Fresnel zone plates act as a lens with a focal length inversely proportional to the wavelength. At Professor Moellenstedt's Institute of Applied Physics in Tübingen, a technique has been developed to manufacture micro-zone plates electronoptically from Buckbee-Mears zone plates. These micro-zone plates have a diameter of 0·5 mm approximately, 38 zones and a focal length of 30 cm for X-rays of nearly 50 Å. Their resolving power is of the order of a few seconds of arc (Einighammer et al., 1966). One finds, however, experimentally as well as theoretically that the sharply defined image is surrounded by a halo having a diameter of a few minutes of arc for the above-mentioned zone plates. Whereas for a point source the intensity of a halo is poor and therefore unimportant, for extended sources it becomes large as a result of the superposition of the contributions from the individual points of the source. This has been demonstrated by the photometer curves of the images of circular sources by Einighammer (1966) (Figure 1). The halo is not noticeable in case of very narrow sources (curves a and b). With increasing diameter of the source the intensity of the halo rapidly increases (curves c and d). For the same source the brightness distribution of the halo agrees practically with the one obtained by a pinhole camera, provided the diameter of the hole and that of the zone plate are equal; these brightness distributions are represented by the dashed lines c′ d′ in Figure 1.

For the interpretation of the physical events occurring at the beginning of a flare it seems to be significant not only to observe the spectra of the short-wavelength X-ray radiation but also its polarization. Solar X-ray spectra below 10 Å during flares observed by Bowen et al. (1964) show in the region of long wavelength a slow fall in their intensity and below 5 Å approximately a rapid fall (Figure 1). This spectral variation cannot be understood on the basis of thermal radiation only. However, it is possible to explain the observed spectra by assuming a superposition of a thermal radiation as well as a non-thermal one. Indeed, one should assume that the fast electrons, during their acceleration, obtain a preferred direction and that only afterwards their velocity distribution becomes a Maxwellian one. The assumption of the existence of a non-Maxwellian distribution of electrons in the beginning of a flare is also supported by the fact that the rise in intensity by more than one order of magnitude takes place in a few minutes. This point has already been emphasised by Bowen et al.

Using an improved isodensitometric technique it has been possible to study in great detail the photometric structure and the evolution of eight flares. A comparison has been made between the evolutive curves and the ones obtained from measurements of solar X-rays and radio fluxes at λλ 3·2 and 21 cm.

A reduction of the flare areas (and of the emitted energies) before the flash phase and a continuous pulsation of the flare have been observed. Further it seems that the flares associated with radio bursts or X-ray events are those which show regions of a sufficiently high intensity, the emitting areas not being a very important parameter.

The correlation in time between the various examined aspects of the AR seems to indicate that the sequence for the beginning of the different phenomena is, in general: optical flare, X-ray events, radio events.

The magnetic model of a proton flare is discussed, in agreement with the known characteristic features of proton-flare development, on the basis of the interaction of an emerging magnetic channel from the photosphere, and its being wedged into the field of an already existing interaction of two systems of magnetic fields of a complex active region with spots. This explains the origin of the two emission filaments following the splitting and of the ‘annular’ emission filament, entering the corona, connected by a fine loop structure with two split filaments in the chromosphere. The acceleration of particles and their ejection will take place in a number of created ‘zero line’ spaces and in spaces where magnetic lines of force are reconnected.

An attempt is made to explain some Hα phenomena in connection with proton flares, by means of interactions of magnetic fields in the chromosphere and in the lower corona.

A model of solar flares is proposed in which the preflare state comprises a bipolar magnetic-field structure associated with a bipolar photospheric magnetic region. At low heights, the magnetic-field lines are closed but, at sufficiently great heights, the lines are drawn out into an open structure comprising a bipolar flux tube containing a ‘neutral sheet’ or ‘sheet pinch’. Such a sheet pinch is probably related to a coronal streamer. The energy stored in the closed-field region is derived from photospheric motion whereas energy stored in the open-field region is derived from the non-thermal energy flux which heats the corona and drives the solar wind.

The flare itself is identified with reconnection of magnetic field by the tearing-mode resistive instability. If the thickness of the sheet pinch is determined by resistive diffusion and a growth time of the bipolar region of order 1 day, the transverse dimension will be about 104 cm. The rise time of the tearing-mode instability is then a few seconds, compatible with the characteristic time of Type-III radio bursts. One can understand that the time-scale of the reconnection process is of order 102–103 sec if reconnection proceeds by the Petscheck mechanism, with the modification that resistive diffusion is replaced by the more rapid Bohm diffusion.

The evolution of a flare, according to this model, appears to fit a number of the observational characteristics of flares.

A solar flare has various aspects: the optical flare is often associated with emissions in the microwave or X-ray regions: this indicates the occurrence of a highly excited plasma, which we call the high-energy flare plasma. The existence of the high-energy flare plasma was first shown by radio observations in the microwave regions (Hachenberg) and later confirmed by X-ray observations in the energy range 102–106 eV.

There are few observations of cases in which a nimbus has developed around a solar flare of major importance. Speculations have been made about the nature of this obscuring cloud. We shall give here some consequences of the assumption that we have to do with scattering by a cloud of electrons, ejected together with the protons during the formation of the flare.

We present a coronal propagation model for fast chromospheric disturbances following flares that were first observed by Moreton and Ramsey, and deduce estimates of average coronal magnetic fields (6 gauss) and disturbance energies (≤ 1029 erg).

More than 70 cases have been observed of energetic solar flare X-ray bursts by large ionization chambers on the OGO satellites in space. The ionization chambers have an energy range between 10 and 50 KeV for X-rays and are also sensitive to solar protons and electrons. A study has been made of the X-ray microwave relationship, and it is found that the total energy released in the form of X-rays between 10 and 50 KeV is approximately proportional to the peak or total energy simultaneously released in the form of microwave emission. For a given burst the rise time, decay time and total duration are similar for the 10–50 KeV X-rays and the 3 to 10 cm radio emission. Roughly exponential decay phases are observed for both emissions with time constants between 1 and 10 min. All 3 or 10 cm radio bursts with peak intensity greater than 80 solar flux units are accompanied by an X-ray burst greater than 3 × 10−7 ergs cm−2 sec−1 peak intensity. The probability of detecting such X-ray events is low unless the radio spectrum extends into the centimetric range of wavelengths. The best correlation between cm-λ and energetic X-rays is observed for the first event in a flare. Subsequent structure and second bursts may not correspond even when the radio emission is rich in the microwave component. The mechanism for the energetic X-rays is shown to be bremsstrahlung probably of fast electrons on a cooler plasma. If the radio emission is assumed to be synchrotron radiation then a relationship is developed between density and magnetic field which meets the observed quantitative results. One finds, on the average, that 5 × 10−54 joules m−2 (CPS)−1 of microwave energy at the Earth are required per electron at the Sun to provide the radio emission for the various events.

A strong correlation between interplanetary solar flare electrons observed by satellite and X-ray bursts is shown to exist. This correlation is weak for solar proton events. One may infer a strong propagation asymmetry for solar flare electrons along the spiral interplanetary magnetic field.

All contributions presented at the Friday morning session concerned results of the Proton Flare Project (PFP), which was organized by the IAU Commission 10 under the sponsorship of the IQSY Committee, from May 1 to September 30, 1966. As explained by Z. Švestka in his introductory talk, this project had four main aims:

1. (1) To observe proton flares shortly after the minimum of the solar activity, when, on the rising part of the solar cycle, some proton flares already appear, but not too many of them, so that the individual proton flare phenomena are fairly isolated. This makes it easy to study all the effects of such a proton flare in the interplanetary space and in the Earth's surroundings, and particularly, it allows a detailed study of the isolated active region, in which the proton flare appears.

2. (2) To get some practical experience in the forecasts of proton flares and verify the reliability and the practical use of them.

3. (3) With the aid of these forecasts, to give to the solar and geophysical observatories and to the launching sites the possibility to get prepared for the coming event. It was hoped that in this way one might get very detailed observations of the proton-flare active region, of the proton flare itself, and of its effects in the interplanetary space and in the Earth's magnetosphere and ionosphere.

The CSI scintillator telescope aboard IMP-III indicated a peak intensity of 6·25 proton/sec cm2 ster for a SPE (Solar Proton Event) commencing during the period 1600 to 1700 UT on August 28, 1966. This SPE appears to have been associated with the importance 2–3 flare occurring at 22N/06S at 1522 UT on this date. The comparison of this event to other large proton events occurring during the period May 25, 1965 to September 30, 1966, is made.

Although the area of some sunspot groups declines suddenly after a flare that produces energetic protons, other groups show a delayed or gradual decline, or continue to grow for several days after the flare. The mean area of sunspot groups that produced flares with polar-cap absorption declines gradually and continuously, beginning within a day after the flare.

Successive flares within the same centre of activity sometimes produce radio events that are remarkably similar. The occurrence of such homologous radio events is commonly restricted to periods of less than 48 hours.

Typical for certain centres of activity is their tendency to give rise to impulsive microwave bursts. Such bursts may occur repeatedly during a period of more than a week. The distributions of the lengths of periods during which series of homologous bursts and of impulsive microwave bursts occur are compared.

We present the first results obtained with the 408 MHz interferometer of Nançay. The typical activity observed at 408 MHz is the ‘noise storm’. Undoubtedly with this apparatus we could distinguish multiple centres, generally double.

The comparison with corresponding optical observations and with Nançay observations at frequency 169 MHz allowed us to associate the centres of noise storm at 408 MHz with eruptive centres lying in ‘Anomalous Active Regions’ and to describe the schematical structure of active zones.