Search

Background and aims Endovenous laser ablation (EVLA) has been well-reported as a minimally invasive method to deal with varices of the lower extremities. The lasers used fall into two categories: pigment, i.e., hemoglobin-specific lasers in the visible and near-infrared (near-IR) wavebands and longer wavelength mid-infrared lasers where the chromophore is water. The fiber used to deliver the laser energy is also important, and not enough attention has been paid to this element of EVLA. The present study was therefore designed to compare EVLA delivered through two specific fiber types coupled with a near-IR laser wavelength where water was the major chromophore. Materials and methods A laser diode system at the wavelength of 1470 nm was used as the laser energy source near a peak in the water absorption spectrum. Laser energy was delivered with two specific types of optical fiber, a Radial™ fiber and a Radial 2ring™ fiber (CeramOptec, Germany), and EVLA was evaluated using a computer simulation model taking light transport into account based on the Monte Carlo method and temperature distribution with the heat conduction equation. Results and conclusions It was confirmed from both the simulation model and a previously published ex vivo experiment that carbonization and sticking during EVLA caused by excess temperature rise can be minimized by using the Radial 2ring fiber compared with the Radial fiber, coupled with the 1470 nm wavelength. In the future, lasers with different wavelengths or optical fibers with differing irradiation modes may appear as candidate systems for EVLA. It is important to evaluate safety and efficacy carefully using the methods in the present study before moving to in vivo indications in human subjects.

The flare catalogue of the Yohkoh mission is compiled and linked to this article as an electronic supplement. For showing flare characteristics over wide energy range concisely, we provide the images of Hard X-ray Telescope (HXT) and the Soft X-ray Telescope (SXT), and the spectra of Hard X-ray Spectrometer (HXS) and Gamma-Ray Spectrometer (GRS) with the Wide Band Spectrometer (WBS) time profiles. The energy versus pulse height (PH) data channels in HXS and GRS are re-calibrated by using the data of the whole mission period. Secular gain changes are recognized in HXS, and the characteristics of power-law flare spectra simultaneously observed by HXT and HXS confirms the trend. The GRS gains are different for the flare observations during the previous maximum and for the current maximum. The total of 33 γ -ray events are observed, and for 12 of them γ-ray flare spectra are obtained.

Solar neutrons have been detected using the neutron monitor located at Mt. Chacaltaya, Bolivia, in association with a large solar flare on November 24, 2000. This is the first detection of solar neutrons by the neutron monitor that have been reported so far in solar cycle 23. The statistical significance of the detection is 5.5 sigma. In this flare, the intense emission of hard X-rays and gamma-rays has been observed by the Yohkoh Hard X-ray Telescope (HXT) and Gamma Ray Spectrometer (GRS), respectively. The production time of solar neutrons is better correlated with those of hard X-rays and gamma-rays than with the production time of soft X-rays. The observations of the solar neutrons on the ground have been limited to solar flares with soft X-ray class greater than X8 in former solar cycles. In this cycle, however, neutrons were detected associated with an X2.3 solar flare on November 24, 2000. This is the first report of the detection of solar neutrons on the ground associated with a solar flare with its X-ray class smaller than X8.

The Yohkoh spacecraft observed a large flare (Ha importance 3B and GOES (Geosynchronous Operational Environmental Satellite) class X6.1) on 27 October 1991. The flare showed strong gamma-ray emission above 1 MeV. The gamma-ray spectrum, which extends to 10 MeV/nuc, consists of electron bremsstrahlung continuum and several gamma-ray lines. The gamma-ray observation indicates that electrons and protons were accelerated to >1 MeV and >10 MeV, respectively, during the maximum phase of the flare. Assuming second-order Fermi stochastic acceleration, we derive the Bessel function-type proton spectrum of αT = 0.029 from a ratio of neutron capture line to nuclear deexcitation line fluences and estimate several parameters describing the acceleration process. In a case of first-order Fermi shock acceleration, the proton spectrum can not be derived from the gamma-my lines. If the shock compression ratio (ratio of upstream to downstream bulk plasma flow velocities) and the spectral characteristic energy are taken to be 1.8 and 100 MeV, respectively, the shock acceleration gives the proton spectrum similar to that obtained from second-order Fermi stochastic acceleration. It is not possible to determine the acceleration mechanism from the gamma-ray lines alone. Moreover, the Yohkoh hard X-ray spectrometer observed 7Be (429 keV) and 7Li (478 keV) lines resulting from 4He+ 4He reactions from two flares on 27 October 1991 and 15 November 1991 (Hα importance 3B and GOES class X1.0). From the 7Be and 7Li line profiles we suggest that the angular distribution of accelerated alpha particles (4He nuclei) is peaked in a direction tangential to the photosphere for the 27 October flare, whereas it is peaked in the downward direction for the 15 November flare.

The gamma-ray spectrometer on Yohkoh has detected the positron annihilation line at 511 keV produced during the 1991 November 15 flare (X1.0/3B). The 511 keV line fluence, integrated over the time interval of 22:37:50-22:38:14 UT, is (6.7 ± 2.2) photons cm−2 . The time profile of 511 keV line exhibited long decay time compared with the electron bremsstrahlung and prompt gamma-ray line components. From the analysis of time profile of the 511 keV line, we come to the following conclusions: (1) the main source of positrons is deexcitation of 16O *6.052 by e+ – e- pair emisson. (2) β+ −emitting nuclei of 31S, 29P, 27Si, 26mAl, 25Al,23Mg, 19Ne, and 21Na are also important sources of positrons in the decay phase. (3) The density of the positron annihilation region in the photosphere is 1016 cm−3 . (4) Most likely interpretation of the time profile is that at least 50% of positrons annihilate in coronal flare loops with a density of 1012−1013 cm−3 and with a temperature of 106 −3 × 106 K.Subject headings: nuclear reactions, nucleosynthesis, abundances — Sun: flares — Sun: X-rays, gamma-rays

The SOLAR-A spacecraft has spectroscopic capabilities in a wide energy band from soft X-rays to gamma-rays. The Wide Band Spectrometer (WBS), consisting of three kinds of spectrometers, soft X-ray spectrometer (SXS), hard X-ray spectrometer (HXS) and gamma-ray spectrometer (GRS), is installed on SOLAR-A to investigate plasma heating, high-energy particle acceleration, and interaction processes. SXS has two proportional counters and each counter provides 128-channel pulse height data in the 2–30 keV range every 2 s and 2-channel pulse count data every 0.25 s. HXS has a Nal scintillation detector and provides 32-channel pulse height data in the 20–400 keV range every 1 s and 2-channel pulse count data every 0.125 s. GRS has two identical BGO scintillation detectors and each detector provides 128-channel pulse height data in the 0.2–10 MeV range every 4 s and 4-channel pulse count data (0.2–0.7, 0.7–4, 4–7, and 7–10 MeV) every 0.25–0.5 s. In addition, each of the BGO scintillation detectors provides 16-channel pulse height data in the 8–100 MeV range every 4 s and 2-channel pulse count data (8–30 and 30–100 MeV) every 0.5 s. The SXS observations enable one to study the thermal evolution of flare plasma by obtaining time series of electron temperatures and emission measures of hot plasma; the HXS observations enable one to study the electron acceleration and heating mechanisms by obtaining time series of the electron spectrum; and the GRS observations enable one to study the high-energy electron and ion acceleration and interaction processes by obtaining time series of electron and ion spectra.

Gamma-ray emitting solar flares observed with Yohkoh were analyzed from a statistical viewpoint. The four-band hard X-ray (15--95 keV) photometric data, taken with the Hard X-ray Telescope onboard Yohkoh, were utilized in combination with the spectro-photometric gamma-ray (0.2--30 MeV) data obtained with the Gamma-Ray Spectrometer. The GOES class was also incorporated. Out of 2788 X-ray flares observed from 1991 October to 2001 December, 178 events with strong hard X-ray emission were selected. Among them, 40 flares were further found to show significant gamma-ray emission. A fractal dimension analysis and multi-band color--color plots of the 40 flares suggest that their soft X-ray to MeV gamma-ray spectral energy distributions involve at least four independent parameters. These are: (1) the overall flare size; (2) the relative intensities of the thermal vs. non-thermal signals; (3) the gamma-ray to hard X-ray intensity ratio; and (4) the hard X-ray spectral slope. These results are examined for possible selection effects. Also, the meanings of the third parameter are briefly considered., Comment: 23 pages, 7 figures, PASJ accepted

A Yohkoh hard X-ray spectrometer observed gamma-ray line features around 400 keV from a flare (X1.0/3B) at 22:35 UT on 15 November 1991. The time profile of hard X-ray count rate consisted of three significant peaks. The line features were detected at the third peak (22:37:46–22:37:50 UT) where strong hard X-rays and gamma-rays were emitted. A continuum-subtracted count spectrum of the hard X-ray spectrometer is shown in Fig. 1. The line features are interpreted in terms of the complex of Li (478 keV) and Be (429 keV) nuclear deexcitation lines which are produced by He(α, p) Li and He(α, n) Be reactions, respectively. Since the two line profiles are sensitive to the angular distribution of the interacting a particles, they can provide an essential key to the directivity of accelerated ions [1]. Although the two lines were not resolved due to low energy resolution of the NaI detector, we can derive the best fitting incident Li and Be line profiles using a hypothesis-testing approach. The derived center energies of the Li and Be lines are 431 ± 32 and 401 ± 32 keV, respectively, and the derived width of each line (FWHM) is 15 keV. The Yohkoh observation suggests that the two lines were redshifted by about 7%. Murphy et al.[2] calculated the Li and Be line profiles in detail using a transport model. They took account both magnetic mirroring and MHD scattering of accelerated α particles in a flaring loop. Their calculation for a disk-centered flare such as the flare on 15 November indicated that the Li and Be lines are redshifted when α particles stream down to the chromosphere. The present result is consistent with a redshift of 7%, suggesting the possibility of downward beaming of accelerated α particles. Downward beaming could be caused by strong pitch-angle scattering with MHD turbulence in the flaring loop.

Two gamma‐ray flares were observed with a gamma‐ray spectrometer and two X‐ray imagers aboard YOHKOH. The gamma‐ray flare on 27 October 1991 showed strong narrow deexcitation lines and two bright hard X‐ray sources. We discuss a gamma‐ray production site from comparison the gamma‐ray (4.5–6.8 MeV) time profile with the hard X‐ray (53–93 keV) time profile of each hard X‐ray bright source. The gamma‐ray lines are thought to be produced at one of the two hard X‐ray sources. The gamma‐ray flare on 3 December 1991 showed strong continuum extending about 8 MeV without detectable gamma‐ray line. It implies that electrons were preferentially accelerated to relativistic energies. The rapid enhancement in hard X‐ray (53–93 keV) emission was observed nearly simultaneously with the large change in flaring loop structures. It suggests that the electrons were efficiently accelerated by strong electric field induced by the rapid variation in the magnetic field.

Yohkoh observed three impulsive gamma-ray flares with different spectral characteristics: a narrow line flare on 27 October 1991, a broad line flare on 15 November 1991 and a continuum flare on 3 December 1991. The narrow gamma-ray lines are produced by accelerated proton interactions with heavy nuclei in the solar atmosphere, the broad gamma-ray lines by accelerated heavy nuclei interactions with protons in the solar atmosphere (reverse interactions) and the continuum by relativistic electron bremsstrahlung. The gamma-ray count spectrum of the flare (X6.1/3B) on 27 October 1991 is shown in Fig. 1. Several strong narrow lines (neutron capture line and deexcitation lines) and broad lines are apparent. These lines result from accelerated particles of the coronal abundances [1]. Fermi acceleration is most probable, because it is most effective in the early stage of the acceleration and the accelerated particle abundances are similar to the source abundances. The gamma-ray count spectrum of the flare (X1.5/3B) on 15 November 1991 is shown in Fig. 2. The spectrum shows broad C and 0 lines and a complex of broad Fe, Mg, Ne and Si lines without detectable narrow lines. Since the acceleration rate of the shock acceleration is proportional to the ratio of square root of mass to charge, heavy nuclei are accelerated in a different manner than protons. Assuming a two-stage acceleration process, Fermi acceleration (first stage) plus shock acceleration (second stage), we can explain the observed high variabilities of the ratio of heavy nuclei to proton abundances (the ratio varies from 0.3 to 0.01)[3]. The gamma-ray the electron bremsstrahlung continuum. This type of specturm is called an electron-dominated event [3]. Since betatron acceleration is more effective for electrins than ions, this spectrum suggests the possibility that betatron acceleration was dominant in this flare.

We present gamma-ray observations of the 1991 October 27, November 15, and December 3 flares with the Yohkoh gamma-ray and hard X-ray spectrometers. The October 27 flare (X6.1 /3B) shows significant gamma-ray line emission, and the temporal evolution of the gamma-ray line-to-bremsstrahlung flux ratio indicates that protons and electrons were complicatedly accelerated during flare. The December 3 flare (X2.0/2B) shows a strong bremsstrahlung continuum extending to 10 MeV and indicates that electrons were preferentially accelerated to 10 MeV. A line feature at 420 keV was observed from the November 15 flare (X1.0/3B). This is most likely due to a compound of redshifted nuclear deexcitation lines of Be (429 keV) and Li (478 keV) resulting from He-He reactions. The spectral feature indicates that the accelerated He nuclei suffer strong pitch-angle scattering in the corona and form a downward-peaked distribution. Furthermore, the November 15 flare exhibits evidence of positron annihilation line at 511 keV. The positron production processes and the electron density of the annihilation region are discussed on the basis of the temporal characteristics of 511 keV line emission.Subject headings: line: identification — Sun: flares — Sun: X-rays, gamma-rays

We discuss X-ray spectra and soft X-ray images of an M1.9 flare that occurred on 1992 January 5 near 13.3 UT. These data were obtained with instrumentation on the Japanese Yohkoh spacecraft. They cover the entire rise phase of the flare. To supplement these data we have ground-based magnetograms and Halpha spectroheliograms. We calculate the electron temperature and emission measure of the flare as a function of time during the early rise phase using X-ray spectral line intensities and line ratios. Using spectral line widths, line profile asymmetries, and wavelength shifts due to the Doppler effect, we calculate the dynamical properties of the flare. The time development of the morphology of the flare, as revealed by the soft X-ray images and the Halpha spectroheliograms, and the physical quantities inferred from the X-ray spectra, are compared with chromospheric evaporation models. There is an enhancement of blueshifted emission that is closely correlated with the hard X-ray bursts. Heating of one loop in the flare is consistent with a conduction-evaporation model, but heating is found in several structures that do not appear to be physically associated with each other. No standard evaporation model can adequately explain all of the observations.

A gamma-ray spectrometer aboard the solar flare obsevation satellite HINOTORI launched on 21 Feb. 1981 is described. The gamma-ray spectrometer consists of a phoswich type scintillator of CsI(Tl) crystal surrounded by a plastic scintillator, which can discriminate between gamma-rays and charged particles. The CsI(Tl) scintillator, whose size is 62 cm2 in area and 5.1 cm in height. covers the energy range 0.21–6.67 MeV. The energy resolution is 0.1E 1 2 (MeV) fwhm and the full energy peak efficiency is 0.40 at 0.662 MeV. The primary data output is a full resolution 128 channel pulse height spectrum every 2 s. Since the launch HINOTORI recorded about forty solar gamma-ray flares and four cosmic gamma-ray bursts, and several interesting results were reported. The typical results of the large solar flare on 27 April 1981 and the cosmic gamma-ray burst on 21 July 1981 are presented in this paper. These observed data could provide important clues for advancing our understanding of high enegry phenomena in solar flares and cosmic gamma-ray bursts.

A balloon-borne cosmic gamma-ray burst detection system has been designed for observation of small gamma-ray burst at a rate of better than one event per day and for determination of source locations. The detection system consists of four inclined large-area high pressure xenon counters. The source location is determined from counting differences between four xenon counters. The directional accuracy of the burst source depends on the gamma-ray burst size, the inclination of the four xenon counters and the elevation angle of the burst source. The relation between the burst detection frequency and the detector area is estimated for this design of the detection system.

The response of a 7.6 cm diameter × 7.6 cm long bismuth germanate (BGO) scintillator to gamma rays from 0.511 to 17.23 MeV has been measured. The full energy peak efficiency of the BGO scintillator is 0.90 at 0.511 MeV, 0.40 at 3.08 MeV and 0.33 at 4.95 MeV. The total peak (full energy + single escape + double escape) efficiency is 0.45 at 10.76 MeV and 0.40 at 17.23 MeV. The energy resolution (fwhm) as a function of gamma-ray energy is approximated by 3.89E0.48 (keV) for energies below 5 MeV. This is about twice that of a NaI scintillator of the same size. The present results indicate that the BGO spectrometer has attractive properties for solar gamma-ray observations.

A balloon-borne high-resolution Ge(Li) spectrometer for observation of electron cyclotron line emissions in the hard X-ray spectra of neutron stars is described. The spectrometer consists of four planar type Ge(Li) crystals and operates in the energy range of 20–200 keV. The Ge(Li) crystals are housed in vacuum cryostats and cooled by liquid nitrogen. A total effective area of the spectrometer is 120 cm 2 and an energy resolution is 2 keV fwhm at 60 keV. A 2 mm thick lead honeycomb collimator provides background rejection and collimation, defining a field of view 20° × 20° fwhm. The sensitivity of the spectrometer to line emissions from cosmic point sources is 10 −3 photons cm −2 s −1 over the energy range. The Ge(Li) spectrometer is scheduled for balloon-flight in September 1979. Electron cyclotron line emissions from Her X−1 and Crab Nebula are to be searched.

The Hinotori gamma-ray spectrometer observed time histories of gamma-ray line and electron bremsstrahlung X-ray emissions from four solar flares. The time histories of impulsive flares require that both nuclei and electrons are rapidly accelerated to energies >10 MeV/nuc and >1 MeV respectively in a time scale of

Solar gamma-ray lines and protons were simultaneously observed for six flares on 1 April, 4 April, 27 April, 13 May 1981, 1 Feb., and 6 June, 1982 by Hinotori and Himawari satellites. The flare list is presented in Table I and the time histories of gamma-rays and protons are shown in Figs. 1 to 6. The relation-ship between the gamma-ray line fluences and peak proton fluxes for these flares is shown in Fig. 7. It does not reveal an apparent correlation between them. The present results imply that the protons producing gamma-ray lines in the flare region and protons observed near the Earth do not always belong to the same population, and favor the downward streaming model for the gamma-ray line production.

The Hinotori satellite observed an unusual time history of hard X-ray and gamma-ray emissions from a limb flare on 4 April, 1981. The detailed comparison of the time history reveals that the gamma-ray emission precedes the hard X-ray emission by about 10 s. Following three possibilities are suggested to explain the unusual time history: (1) The ion acceleration precedes the electron acceleration, (2) the ions and electrons are simultaneously accelerated, but the nuclear reactions responsible for gamma-ray emission precede the electron bremsstrahlung responsible for hard X-ray emission, and (3) the gamma-rays and hard X-rays are simultaneously emitted, but the hard X-rays are strongly attenuated from the limb effect.

Solar neutron fluences from large gamma-ray flares are deduced from accelerated proton spectra and numbers derived from the gamma-ray observations. The deduced solar neutron fluences range from 1 to 200 neutrons cm -2 . The present result indicates a possibility that high sensitivity ground-based neutron monitors can detect solar neutron events, just as detected by the Jungfraujoch and Rome neutron monitors.

Solar flare photon energy spectra in the 20 keV to 7 MeV range are derived from the 1981 Apr. 1, Apr. 4, Apr. 27 and May 13 flares. The flares were observed with a hard X-ray and a gamma-ray spectrometers on board the Hinotori satellite. The results reveal that the shape of hard X-ray spectrum for the gamma-ray flares does not much vary from flare to flare, and the spectra harden apparently at energy about 400 keV. Effects of nuclear line emission on the continuum and of higher energy electron bremsstrahlung are considered to explain the spectral hardening.

Energy spectra of accelerated protons at flare sites are derived by comparing the observed ratios of the 4.44 MeV gamma-ray line fluence to the 2.22 MeV gamma-ray line fluence with the calculated ratios. The proton spectra derived for the 1981 Apr. 1, 1981 May 13, 1981 Oct. 15 and 1982 June 6 disk flares do not much vary from flare to flare, and are in agreement with average proton spectra observed in interplanetary space within the errors. It implies that both the protons which produce gamma-ray lines at the Sun and those which escape from the Sun are probably accelerated by the same mechanism.

The electronically gated in-flight energy calibration is applied to the X-ray and gamma-ray spectrometers on a SOLAR-A satellite. The NaI X-ray spectrometer covering the 20–400 keV energy band is calibrated by an 241 Am radioactive source which decays by the simultaneous emission of 60 keV X-rays and 5.48 MeV alpha-particles. The X-ray calibration spectrum is accumulated in coincidence with an event tag pulse generated by the simultaneous detection of an X-ray and an alpha-particle. The BGO gamma-ray spectrometer covering the 0.2–10 MeV energy band is calibrated by a 60 Co radioactive source which decays by the simultaneous emission of 1.17 and 1.33 MeV gamma-rays and a beta-ray (maximum energy is 313 keV). The gamma-ray calibration spectrum is accumulated in a manner similar to the X-ray spectrometer. Since the present method enables to select the calibration pulse without the disadvantage of introducing extra pulses, it is suitable for a space experiment where external conditions and background counting rates can significantly change.

Large-area silicon semiconductor detectors with an effective area of 20 cm2 have been fabricated to observe primary cosmic ray isotopes from hydrogen to oxygen with a high resolution. As a preliminary experiment the energy resolution of detectors was investigated for α particles and internal conversion electrons. A detector telescope consisting of ΔE and E detectors was constructed and the mass resolution of the detector telescope was investigated by measuring long-range 11H, 21H and 31H nuclei emitted from the radioactive 25298Cf source. The expected result was obtained from this preliminary experiment and it seems possible to separate masses of primary cosmic ray isotopes up to oxygen with a high resolution.

Possible solar neutron emission associated with five SMM gamma-ray events on 7 June 1980, 21 June 1980, 6 November 1980, 26 November 1982 and 25 April 1984 was found from analysis of 10-minute records of the ground-based neutron monitor at Tokyo. Of these the two events on 21 June 1980 and 25 April 1984 have been already known as neutron events. The time histories of the neutron monitor count rate are compared with those of the gamma-ray count rate and the possibility of energetic neutron emission at the flare site is discussed.

The large-area lithium-ion-drifted silicon semiconductor detector of 70 mm diameter and 1 mm thickness is fabricated by the improved lithium-ion drift method. The electric characteristics of the reverse-bias versus reverse-current and the reverse-bias versus capacitance are investigated. The energy spectra of α-particles from 23994Pu, 24196Am and 24496Cm and internal conversion electrons from 20783Bi are measured. The energy resolutions are 90 keV (fwhm) for α-particles and 80 keV (fwhm) for electrons, respectively. The characteristics of the pulse height versus reverse-bias and the pulse height uniformity over the entire area of the detector are investigated. The results show that the pulse height is saturated above 150 V and that the complete uniformity is obtained.

During February, 1981 and June, 1982 the gamma-ray and the hard X-ray spectrometers on the Hinotori satellite observed four gamma-ray bursts on 28 February, 21 July, 1981, 26 February and 13 March, 1982. These gamma-ray bursts were simultaneously observed by other satellites. The time histories and energy spectra are shown for these gamma-ray bursts, and the burst sizes (erg cm−2) are estimated. Two possible source locations for the burst of 21 July, 1981 are roughly determined from arrival time delays between two pairs of satellites, PVO-Hinotori and ISEE-3-Hinotori. The weak gamma-ray line peak structure around 1.8 MeV was observed for the burst of 13 March, 1982. The line could be interpreted in terms of gravitationally redshifted neutron capture line at 2.22 MeV.

The Wide Band Spectrometer (WBS) for the solar flare satellite SOLAR-A scheduled for launch in 1991 consists of three kinds of detectors to observe the wide band spectrum from soft X-rays to gamma-rays. The soft X-ray spectrometer (gas proportional counter), hard X-ray spectrometer (NaI scintillation counter and gamma-ray spectrometer (two BGO scintillation counters) cover the 2–30 keV, 20–400 keV and 0.2–100 MeV bands, respectively. Further, the WBS contains a radiation belt monitor consisting of a NaI scintillation counter which is capable of detecting cosmic gamma-ray bursts. The details of these detectors, electronics systems and data collection are described.

The Hinotori satellite observed eight solar gamma-ray line flares between April, 1981 and June, 1982. Ratios of the fluence of 2.22 MeV gamma-ray line restulting from neutron capture on proton to the fluence of 4.44 MeV gamma-ray line resulting from the nuclear deexcitation of C are shown as a function of the heliolongitude of flare. The ratios are roughly constant for disk flares, but decreased for limb flares. The present result suggests that the particle acceleration produces energy spectra that do not vary much from flare to flare. The limb darkening of the ratios can be interpreted in terms of the attenuation of the 2.22 MeV gamma-ray line by Compton scattering in the photosphere.

Numbers of solar protons accelerated at flare sites are derived from gamma-ray observations of the C deexcitation line at 4.44 MeV and the neutron capture line at 2.22 MeV from the 1981 Apr. 1, 1981 May 13, 1981 Oct. 15 and 1982 June 6 flares. The derived numbers of protons accelerated to energies above 10 MeV are of the order of 10 32 . The number of protons producing gamma-ray lines seems to be much higher than the number of escaping protons from the Sun. The present result suggest that most of protons stream down and produce gamma-ray lines, and only a small fraction of protons escapes into interplanetary space.

Time variations of the hard X-ray spectrum in solar flares are observed by the hard X-ray spectrometer (HXM) aboard the Hinotori satellite. With a new presentation of the dynamic spectrum we have studied the differences between impulsive and gradual hard X-ray bursts. In the impulsive events a “bent” spectrum up to some hundred keV persists at least until the main peak. In the gradual events, on the other hand, a power-law spectrum augmented by a low-energy excess is dominant.

Some interesting results on gamma-ray line emission and its time profiles observed by Hinotori are presented. Possible explanations of gamma-ray line and hard X-ray emissions for the impulsive and gradual flares are discussed. Relationship between the gamma-ray line emission and acceleration and escape of the solar particles is also studied.

SINCE the γ-ray burst was discovered in 1973, approximately 50 events have been observed using artificial satellites1,2. In addition, several bursts of smaller size have been found using balloon-borne detectors3–6 with large sensitive areas. No burst has yet been located on the celestial sphere, with an adequate precision to associate it with an astronomical object. To determine the precise position of a γ-ray burst which had not been predicted to occur, the detector must have a wide field of view and the capability of precise location of the source. A rotating cross-modulation-collimator (RCMC) proposed7 as a device to fulfill these apparently conflicting requirements was used in the series of balloon observations reported here. A small γ-ray burst was found during ∼ 150 h of observations and its celestial position was determined with a precision of ∼ 0.3°.

Galactic y-ray lines have been investigated with a balloon-borne Ge(Li) spectrometer. The galactic plane was surveyed over the longitudinal interval 40�_210�, but upper limits only were obtained for the intensities of y-ray lines from positron annihilation at 0�511 MeV and the nuclear decays of 2�Ne� (1� 634 MeV), 24Mg* (1�369 MeV), 28Si* (1� 779 MeV) and S6Pe� (0�847 MeV). These upper limits are of the order of 10-3 photons cm-2 S-l sct, which are consistent with measurements for other lines observed so far. A possible y-ray line feature at 400 keV, with a flux of (7 �4� 5 �4) x 10-3 photons cm-2 s-t, was detected when the Crab Nebula was in the field of view. This line seems to be identical with one reported by Leventhal et al. (1977), who suggested that it may originate from a gravitational redshift of the 0�511 MeV annihilation line produced near the surface of the neutron star.