A classical coset Hamiltonian is introduced for the system of one electron in multi-sites. By this Hamiltonian, the dynamical behaviour of the electronic motion can be readily simulated. The simulation reproduces the retardation of the electron density decay in a lattice with site energies randomly distributed - an analogy with Anderson localization. This algorithm is also applied to reproduce the Hammett equation which relates the reaction rate with the property of the substitutions in the organic chemical reactions. The advantages and shortcomings of this algorithm, as contrasted with traditional quantum methods such as the molecular orbital theory, are also discussed.

We present two theorems on calculating the relative entropy of entanglement. Theorem 1 is an extension of Vedral and Plenio's theorem (Phys. Rev. A 57 (1998) 1619) for pure states, which is useful for calculating the relative entropy of entanglement for all pure states as well as for a class of mixed states. Theorem 2 gives the relative entropy of entanglement for any bipartite state whose tripartite purification has two separable reduced bipartite states.

The quantum fluctuations of the charge and current in a non-dissipative mesoscopic circuit with coupled inductors and capacitors are studied for the squeezed vacuum state. The influence of the phase angle upon the quantum fluctuations is discussed in particular. Our results indicate that when the circuit parameters remain constant and the phases θ of the two circuits are equal, the squeezing of the charge or its conjugate variable increases. When the difference of the two phases equals π, the squeezing will deteriorate. Therefore, if we want to decrease the quantum noise, we should not only control the amplitude r, but also control the phase θ carefully.

We propose that a slowly-moving standing wave can be used to manipulate the motion of an atomic wave packet in a hollow optical fibre. A cloud of cold atoms can follow the motion of such a slow wave. The evolution of an initial Gaussian atomic wave packet in the moving standing wave is calculated by numerically solving the Schrödinger equation. Our calculation shows that by using a sufficiently strong field and by adjusting the velocity of the standing wave pattern, the motion of the atomic cloud can be controlled in a hollow fibre.

A fast melting release method for the free-fall equivalence principle test using laser interferometry is discussed. The primary experiment result shows that the uncertainty of the differential release time could be controlled at the level of 1ms by this release system, which satisfies the requirement of the expected experimental precision.

We study the escape problem of a bistable sawtooth system
driven by correlated multiplicative and additive white noises. Exact analytic representation of the relative escape rate is obtained. Some novel phenomena are found from the calculation result;“suppression platform”appears in the relative escape rate when the noise correlation coefficient λ is positive. Moreover, for small positive values of λ, the relative escape rate exhibits “suppression platform” and then “resonant activation” with the increasing noise strength ratio R.

In our previous work [Phys. Rev. Lett. 80 (1998) 696] we investigated a crisis-induced transition from temporal chaos to spatiotemporal chaos (STC) due to a collision between the unstable orbit of the carrier saddle steady wave (SSW) and the attractor of its perturbation wave (PW). In this letter, we find that at the crisis one PW mode-phase experiences a state transition. The variations of PW mode-phases are governed by the linear dispersion as well as two competitive effects arising from the system nonlinearity: (1) interaction between the SSW and the PW (SP); (2) self-nonlinearity of the PW (PP). With the increasing importance of system nonlinearity we find that before the crisis for all PW modes the SP effects dominate and the mode-phases vibrate chaotically; while after the crisis for one PW mode the PP effect surpasses that of SP, in which case the phase transits to a state of combination of chaotic whirling and vibrating, which is different from the case of weak nonlinearity. In the meantime, the PW partial wave is free from the trapping of its carrier, and can be driven or damped chaotically depending on the phase difference relative to the carrier SSW, causing the STC motion after the crisis.

The reconstruction process of the Si(111) √3 x√3 R30°-Ag surface is studied by using a scanning tunnelling microscope at 78K. By applying a strong interaction between the tip and the surface, a tip-induced reconstruction corresponding to the mergence of two Si(111) √3 x√3 R30°-Ag domains is observed. Based on the inequivalent trimers (IET) model, this reconstruction process is attributed to a transition between the clockwise and counterclockwise IET domains. With this transition, the honeycomb-chained-trimer Si(111) )√3 x√3 R30°-Ag anti-phase boundary disappears and changes to the IET structure.

By incorporating the flavour-dependent quark--antiquark annihilation amplitude into the mass-squared matrix describing the mixing of the isoscalar states of a meson nonet, the new version of Schwinger's nonet mass formula, which holds with a high accuracy for the 0^{-+}, 1^{--}, 2^{++}, 2^{-+} and 3^{--} nonets, is derived and the mixing angle of the isoscalar octet--singlet for these nonets is obtained. In particular, the mixing angle of the isoscalar octet--singlet for the pseudoscalar nonet is determined to take the value of -12.92°, which is in agreement with the value range from -13° to -17° deduced from a rather exhaustive and up-to-date analysis of data. It is also pointed out that the omission of the flavour-dependent qq annihilation effect might be a factor resulting in the invalidity of Schwinger’s original nonet mass formula for the pseudoscalar nonet.

The glueball spectrum is studied using an improved gluonic action on asymmetric lattices in the pure SU(3) lattice gauge theory. The smallest spatial lattice spacing is about 0.08fm which makes the extrapolation to the continuum limit more reliable. Converting our lattice results to physical units using the scale set by the static quark potential, We obtain the following results for the glueball masses: M_{G}(0^{++})=1730(90) MeV for the scalar glueball and M_{G}(2^{++})=2400(95) MeV for the tensor glueball.

We investigate the effect of intrinsic sea quark-antiquark pairs on the magnetic moments of the SU(3) octet baryons. The intrinsic quark-antiquark pairs are introduced by a model of energetically-favoured non-perturbative baryon-meson fluctuations. It is shown that the baryon-meson fluctuation model provides a mechanism of SU(3) symmetry breaking in the baryon sea and is related to the violations of Gell-Mann-Okubo and Coleman-Glashow sum rules for the baryon magnetic moments. Such a model is also consistent with its explanation of the violations of Gottfried and Ellis-Jaffe sum rules in deep inelastic scattering.

The rigid-rotor with a q-deformed moment of inertia is introduced to describe the nuclear rotational spectra. With the representations of quantum algebra, the normal deformed and superdeformed bands are naturally differentiated by softness. The yrast normal deformed bands in rare earth and actinium regions, as well as the yrast superdeformed bands in A-190 and 150 regions are investigated. The calculated results agree with experimental data qualitatively well, and the values of the parameters are physically reasonable. This indicates that the fixed deformation, the stretching effect and the many body statistics effect are three possible dominant factors to govern nuclear rotational bands.

For the first time, the analytical continuation in the coupling constant method has been combined with the relativistic mean field theory to study the unbound states in spherical nuclei. The 1d_{3/2} neutron state in ^{16}O and the 2d_{5/2} and 1g_{9/2} neutron states in ^{48}Ca are taken as examples. The calculated energies and widths are compared with available data.

The intermediate mass dileptons from the quark phase, secondary hadronic processes and background sources have been studied based on a relativistic hydrodynamic model. Due to the effect of the phase boundary on the evolution of the system the contribution from the quark phase is much more important than that from secondary processes and is even comparable with that from background sources. This leads to a spectrum without the obvious humps of the hadronic phase contribution and the yield increasing with the incident energy of colliding nuclei. It is shown that the enhancement of dileptons is a signature for the formation of the quark--gluon plasma.

Analytic expressions have been derived of the alternating current (ac) Zeeman and ac Stark effect in an atomic beam magnetic resonance method using Ramsey separated oscillating fields. An interesting feature which will affect the normal Ramsey pattern is that an interference fringe is superimposed on the dispersion lineshapes of the normal ac Zeeman or ac Stark effect. We point out that this new character of ac Zeeman (ac Stark) effect generally exists in all kinds of Ramsey method, for example, in the optical Ramsey atomic interferometer and atomic beam frequency standard. An important implication is that, particularly in an atomic beam frequency standard using Ramsey method, this effect has an influence on the evaluation of the second-order Doppler frequency shift.

Large two-photon absorption cross section of a novel compound
trans-1,3,5-tri-(4-N,N-dimethylaminobenzylstyryl)-acetylacetone is measured to be σ_{2}=8.4 x 10^{-20}cm^{4}/GW or σ^{'}_{2}=1.594 x 10^{-47}cm^{4}/(photon.s) by using a 1064 nm Nd:YAG laser with 35 ps pulse width based on the intensity-dependent transmissivity measurement. The frequency upconversion fluorescence is observed by two-photon laser pumping. This material is an ideal candidate for optical limiting and upconversion fluorescence.

Experimental evidence of abnormally deep penetration in some botanical targets by low-energy ion beams is presented. The energy spectra of 818 keV He^{+} ions penetrating a 70μm thick seed coat of maize, fruit peel of grape and of tomato all have a common feature. The leading edges of these broad spectra indicate that some of the penetrating ions pass through the thick targets easily and only lose a small fraction of their initial incident energy. Rutherford backscattering spectrometry and electron microprobe measurements are used to determine the argon concentration in multilayer samples of the seed coat of maize implanted by 200 keV Ar^{+} ions. The results show that about 10% of the Ar^{+} ions can penetrate deeper than ~ 100μm in these samples.

Laser absorption is measured in Xe, Ar and He clusters irradiated by 5 mJ laser pulses in 150 fs. The measurements show that the absorption efficiency strongly depends on the backing pressure of the gas jets, the laser pulse duration and the atomic number Z of the working gas. The laser absorption for Xe clusters is found to be as high as 45% at a backing pressure of 20 atm and a laser intensity of 1 x 10^{15} W/cm^{2}. Significant numbers of ions with energies up to 100 keV are detected from Xe cluster explosion. The variation trend of the average ion energy at different backing pressures is consistent with that of the laser energy absorption.

A two-dimensional fractional Brown motion (fBm) fractal model is presented, which is suitable for describing the natural rough surface. Using the Kirchhoff approximation, the expressions of the electromagnetic scattering field and the normalized radar cross section from this fractal surface are obtained. Numerical results of the normalized radar cross section as a function of the fractal dimension, characteristic length and incidence frequency are given. Our fBm model is also compared with classical models of Gaussian and exponential correlation.

We propose a method for the generation of self-affine fractal random surfaces, in which we use Fourier transform and its inversion in the algorithm. The light scattering of surfaces of this kind is simulated at different incident angles of illumination. The variation of the full width at half maximum (FWHM) of the intensity profile versus the perpendicular component k_{⊥} of the wave-vector shows clearly the characteristics of the surfaces parameters. The simulation demonstrates how the value of FWHM at k^{2}_{⊥} w^{2} ≤ 1 region and the slope of ln w_{p} - ln k_{⊥} curve at k^{2}_{⊥} w^{2} ≥ 1 region are used, respectively, to extract the lateral correlation length ξ and the roughness exponent α.

As a continuation of the previous paper ‘Collapse and revival in the damped Jaynes--Cummings model’ [Chin. Phys. Lett. 16 (1999) 895], at present the problem is solved under the condition that the atomic spontaneous emission is included. In the case of no damping, c=0, γ_{1} ≠ 0, γ_{2} ≠ 0, the half atomic inversion operator <σ_{Z}> possesses an analytic solution, whereas in the general case c ≠ 0, the problem is reduced to the numerical evaluation of first-order nonlinear differential equations. The final results show that with the increase of γ_{1} and γ_{2}, the kinetic quantities converge rapidly to stationary solutions.

The effects of atomic coherence on the single-mode two-photon micromaser injected with slow V-type three-state atoms are studied for the first time. It is shown that the atomic coherence can modify the atomic emission probability. The effects of the atomic centre-of-mass momentum, the cavity length and other parameters are also studied.

Tunable radiation with a linewidth of 0.014 nm was obtained in Rhodamine B, Rhodamine 6G, Perylene orange and Pyrromethene 567-doped GPTMS solid-state dye materials using intracavity grating dispersive oscillation. The solid dye samples were specially designed in the way in which they are gelling, drying and ageing between two anti-reflection coated disks, thus without any mechanical and optical processing. The conversion efficiency of 4.6% and the tunable range of 40 nm from 589--629 nm have been achieved in Rhodamine B-doped samples. The broadband lasing of 4.4 nm and 54% slope efficiency were also demonstrated in the experiment.

Q-switched pulses at 1.064μm with a peak power of 5.02 kW and a pulse width of 2.8 ns were obtained which were pumped by a 1 W laser diode on the Nd:YVO_{4} microchip at the 1 kHz repetition rate. These values were achieved by combining the techniques of acousto-optic Q-switching and electro-optic pulse-transmission-mode Q-switching. The temporal characteristics of the pulses were analysed numerically. The experimental results are shown to be in good agreement with theoretical predictions.

A theoretical model of the self sum-frequency-mixing (SFM) laser generated by a single crystal is proposed, in which spatial distribution of the pump and circulating fundamental lasers with arbitrary beam waists are taken into account. The model is then applied to two kinds of crystals of current interest, Nd:YAI_{3}(BO_{3})_{4} and Nd:Ca_{4}GdO(BO_{3})_{3}. Numerical analyses of the self-SFM laser properties predict and confirm some experimental results. The model proposed is not limited to self-SFM lasers and may be applied to general analyses of the fundamental or nonlinear laser generation with Gaussian beams.

The theory of the screening-photovoltaic solitons is improved in biased photorefractive-photovoltaic crystals. When the photovoltaic effect is negligible, the screening-photovoltaic solitons are the screening ones, and their space-charge field is the space-charge field of the screening solitons. If the external field is absent, the screening-photovoltaic solitons are the photovoltaic ones on the open- and closed-circuit conditions, and their space-charge field is of the photovoltaic solitons. We also show theoretically that the screening and the photovoltaic solitons on the open- and closed-circuit conditions may be studies together as the screening-photovoltaic solitons.

Within Born--Oppenheimer approximation, by using the classical trajectory theory, a description for the high-order harmonic generation of the hydrogen molecular ion interacting with ultrashort laser pulses has been presented. The Coulomb singularities have been remedied by the regularization. The action-angle variables have been used to generate the initial inversion symmetry microcanonical distribution. Within a proper intensity range, a harmonic plateau with only odd harmonics appears. For a larger intensity, because of the existence of chaos, the harmonic spectra become noisier. For a large enough intensity, the ionization takes place and the harmonics disappear. So the chaos causes the noises, the ionization suppresses the harmonic generation, and the onset of the ionization follows the onset of chaos.

Starting from our definition of apodization profile functions, we discover the optimal profiles and the characteristics of defined apodization functions in sidelobe suppression. It is shown by numerical analysis that the optimal grating length is 45 mm to compensate for dispersion induced by 100 km fibres and the smoothness of the ripples in time-delay characteristics is related to the defined parameters.

The optical waveguide was formed on an LiNbO_{3} substrate by 2.6 MeV nickel ions implantation to the dose of 9 x 10^{14} ions/cm^{2}. Five dark modes were observed by the prism coupling technique. The refractive index profile was obtained by using the reflectivity calculation method. A large index decrease was found in the guiding region and in the optical barrier, which is somewhat different from that of the LiNbO_{3} waveguide formed by the MeV He^{+} ions. The position of the optical barrier is deeper than that of the damage peak calculated by TRIM'90 (Transport of Ions in Matter) code. The crystal lattice damage in the guiding region caused by the Ni^{+} ion implantation was analysed by the Rutherford backscattering/channelling technique.

We demonstrate a type of 2 x 2 multimode interference 3 dB coupler based on silicon-on-insulator. The fabrication tolerance was investigated by the effective index method and the guide mode method. The devices with different lengths were fabricated and near-field output images were obtained. Tolerances to width, length and etch depth are 2, 200 and 2μm, respectively. The devices show a uniform power distribution.

The effects of the background ions on the selection of the discharge path in an air gap have been studied with two different methods. The lightning impulse air discharge experiment is conducted using an independent ion generator, while the air discharge experiment uses a lightning impulse superimposed on a dc high voltage used to produce background ions. The influence of different background ions on the leader development, and thus on the discharge path, is observed. Consistent results have been obtained with the two methods. The probability for the discharge path passing through the negative ion space is much higher than that for the passing through the positive ion space. The mechanism of the effects of background ions is analysed based on the electron avalanche and the electric field.

The dispersion relation of the lower hybrid wave in the tokamak current drive regime (ω > 2ω_{LH}, where ω_{LH} is the lower hybrid frequency) is analysed. Toroidicity effects can shift the parallel refractive index upward or downward. The downward shift will lead to a mode conversion from the slow wave to the fast wave. It is indicated that this is the main mechanism limiting the effectiveness of tokamak lower hybrid current drive. Accordingly, a sufficient condition is obtained for the wave penetration, corresponding to a critical density, which is proportional to f^{4/3}B^{2/3}(A)^{4/3}, with f being the wave frequency, B the magnetic field strength and A the aspect ratio. Quantitative agreement with experiments is found.

High spatial resolution measurements of the electrostatic Reynolds stress, radial electric field and poloidal phase velocity of fluctuations in the edge region of the HT-6M tokamak are carried out. The Reynolds stress shows a radial gradient in proximity to the poloidal velocity shear. A comparison of the profiles between the Reynolds stress gradient and the poloidal velocity damping reveals some similarity in their magnitude and radial structure. These facts suggest that the turbulence-induced Reynolds stress may play a significant role in generating the poloidal flow in the plasma edge region.

Time and space resolved measurements of the radial electric field (E_{r}) have been conducted during the electrode biasing experiments on the KT-5C tokamak. The suppression of the turbulent transport with the change of E_{r} induced by the biased electrode is observed. It is found that the poloidal flow contributes to the main part of the E_{r}, and the change of the poloidal flow has a lead of about 20μs to the formation of E_{r}. These observations suggest that a radial current, responding to an induced voltage on the electrode, drives a poloidal flow which in turn drives the radial electric field.

A thin film of GaN with the thickness of 1.0μm was grown on α-Al_{2}O_{3} substrate by metal organic chemical vapour disposition and then a thick GaN film with thickness of 12μm was grown in the halide vapour phase epitaxy system. Some macro-pyramids appeared on the surface of the sample. The macro-pyramids made the surface of the GaN film rough, which was harmful to the devices made by GaN materials. These defects changed the distribution of carrier concentration and affected the optical properties of GaN. The step height of the pyramids was about 30-40 nm measured by atomic force microscopy. A simple model was proposed to explain the macro-pyramid phenomenon compared with the growth spiral. The growth of the macro-pyramid was relative to the physical conditions in the reaction zone. Both increasing growth temperature and low pressure may reduce the pyramid size.

A Gunn active layer is used as an X electron probe to detect the X tunnelling current in the GaAs--AlAs heterostructure, from which a new heterostructure intervalley transferred electron (HITE) device is obtained. In the 8 mm band, the highest pulse output power of these diodes is 2.65 W and the highest conversion efficiency is 18%. The dc and rf performance of the HITE devices was simulated by the band mixing resonant tunnelling theory and Monte Carlo transport simulation. The HITE effect has transformed the transit-time dipole-layer mode in the Gunn diode into a relaxation oscillation mode in the HITE device. From the comparison of simulated results to the measured data, the HITE effect is demonstrated straightforwardly.

Using the method of few-body physics, we obtain the relation between the correlation energies of the low-lying states of the positively charged exciton X^{+} and the radius of quantum dots. We also calculate the binding energies of the ground state of a positively charged exciton as a function of the electron-to-hole mass ratio for a few values of the strength of the confinement. We find that there exists a critical mass ratio σ^{c}, such that if σ > σ^{c} (σ < σ^{c}) the X^{+} configuration is stable (unstable), and σ^{c} increases as the strength of the confinement increases.

Using fractional Fokker--Planck equation and Langevin equation's Monte Carlo simulation, we find that when current density is less than critical current density, at the very beginning the motion for vortices is mainly within the pinning well with the typical relaxation time T_{r}. When time t is longer than T_{r}, the vortex glass will distort between two adjacent pining wells and will lead to memory effect which is lacking in the Mott insulator and pancake vortices. The inertial effect is not dominating as compared with the formation and growth of easy-flow channels. The random quenched disorder will lead to Mittag--Leffler relaxation.

The giant magneto-impedance (GMI) effect has been investigated in sandwiched FeCuCrVSiB films annealed at 300°C for 1.5 h. The frequency and field dependences of the GMI have been observed in the frequency range from 50 kHz to 13 MHz. The GMI ratio increases at first with increasing frequency, and reaches its maximum value of 136% at a very low characteristic frequency of about 4 MHz, and then decreases with further increasing frequency. These superior properties are related to the special structure of the sandwiched films.

The Schlieren technique coupling with a differential interference microscope was applied to visualize the KNbO_{3} melt motion in a loop-shaped Pt wire heater. The natural convection in KNbO_{3} melt was traced by observing the movement of the tiny KNbO_{3} crystals (～10μm) and the stream velocities of these tracer crystals were measured. In theoretical analysis, the Navier--Stokes equation was solved as a stable field. The general solution for this system of the differential equation was expressed by an approximate power series of azimuth and radius vector. The expression was substituted in the differential equation; a non-trivial solution was obtained exactly. The velocity distribution in the vertical section was obtained which is in qualitative agreement with the experimental result.

The lanthanum strontium aluminium tantalum oxide (LSAT) crystal is grown by means of a floating or pulling method. Its optical transmittance, refractive indices in visible and near-infrared areas and its dispersion curve are reported. The transparence range of LSAT is from 0.45 to 4.2μm and its refractive index is 2.0244.

The upconversion luminescence of several YbEr co-doped Zr--Ba--La--Al--Na (ZBLAN) glass samples (Er(0.5)Yb(3): ZBLAN, Er(0.5)Yb(1):ZBLAN and Er(0.5):ZBLAN) has been studied. A new kind of upconversion cooperative radiation fluorescence, which comes from coupled cluster states of two Yb^{3+} ions has been observed. This is significant in that it may lead to a new path to achieve blue upconversion luminescence, which has promising upconversion applications.

Electrical and optical properties of InGaN/AlGaN double heterostructure blue light-emitting diodes were investigated. Measurement of the forward bias current--voltage behaviour of the device demonstrated a departure from the Shockley model of a p--n diode, and it was observed that the dominant mechanism of carrier transport across the junction is associated with carrier tunnelling. Electroluminescence experiments indicated that there was a main emission band around 2.80 eV and a relatively weaker peak at 3.2 eV. A significant blueshift of the optical emission band was observed, which was consistent with the tunnelling character of electrical characteristics. Furthermore, the degradation in I-V characteristics and the low resistance ohmic short of the device were observed.

Nano-crystalline diamond films are successfully deposited on silicon substrates via the hot filament chemical vapour deposition process using a CH_{4}/H_{2}/Ar gas mixture. The as-grown films are analysed by using field emission scanning electron microscopy, micro-Raman spectroscopy and x-ray diffraction. These results show that the films consist of nano-diamond grains with sizes ranging from 10 to 100 nm, and argon is an important element in the formation of nano-crystalline diamonds.

The ionization kinetics of sodium diluted in argon is studied in a shock tube, in which the test gas mixture is ionized by a reflected shock wave and subsequently quenched by a strong rarefaction wave. A Langmuir electrostatic probe is used to monitor the variation of the ion number density at the reflection shock wave region. The working state of the probe is in the near free fall region and a correction for reduction of the probe current due to elastic scattering in the probe sheath is introduced. At the temperature range of 800 to 2600 K and in the ambience of argon gas, the three-body recombination rate coefficient of the sodium ion with electron is determined: 3.43 x 10^{-14}T^{-3.77}cm^{6}.s^{-1}.

We first demonstrate how to quantify the information conveyed in temporal firing patterns of neurons. We then show that neurons are more sensitive to signals with frequencies within the range of 20-70 Hz in weak signal detection. Such frequency sensitivity is characterized by both the output signal-to-noise ratio and the information measures.

The generation of spontaneous synchronized rhythm and its role are studied by using a globally coupled excitable stochastic neuronal network. When the coupling strength exceeds a critical value, the neurons with a suitable noise in the network exhibit a strong tendency to synchronize and display a spontaneous rhythm. The coherence of the network can be enhanced by a suitable noise and a coupling of the network. The spontaneous rhythm enhances the ability of the network in processing weak periodic signals.

Metal ions are essential to the structure and physiological functions of bacteriorhodopsin. Experimental evidence suggests the existence of specific cation binding to the negatively charged groups of Asp85 and Asp212 via an electrostatic interaction. However, only using electrostatic force is not enough to explain the role of the metal cations because the carboxylate of Asp85 is well known to be protonated in the M intermediate. Considering the presence of some aromatic amino acid residues in the vicinity of the retinal pocket, the existence of cation-π interactions between the metal cation and aromatic amino acid residues is suggested. Obviously, introduction of this kind of interaction is conducive to understanding the effects of the metal cations and aromatic amino acid residues inside the protein on the structural stability and proton pumping of bacteriorhodopsin.

We extend the research by Lubow and Pringle of axisymmetric waves in accretion disks to the case where self-gravity of disks should be considered. We derive and analyse the dispersion relations with the effect of self-gravity. Results show that self-gravity extends the forbidden region of the wave propagation: for high frequency p-modes, self-gravity makes the wavelength shorter and the group velocity larger; for low frequency g-modes, the effect is opposite.

The transport of solar flare particles in the corona is studied. Considering the problems in terms of the characteristics of a sunspot group producing solar cosmic rays and solar flare processes, we find that formation of the fast propagation process is associated with annihilation of sunspots in the group with magnetic multipolarity. The slower propagation process depends on magnetic irregularities in the corona, and the evolution of the transport is related to the flare processes. Equations for the coronal transport are proposed and their initial and boundary conditions are given. The predicted results agree with the main observational features.

A mechanism is presented for generating high speed particles in Herbig--Haro flow coming from accretion disks associated with
protostars. The disks are threaded with weak magnetic field lines, in which the magnetorotational instability results in magnetohydrodynamic turbulence. Then the turbulent waves accelerate the thermal particles out of the accretion disks to a few hundred kilometres per second, forming the high speed particles of optical jets in star-forming regions.

We improve the brick-wall model to take only the contribution of a thin film near the event horizon into account. This improvement not only gives us a satisfactory result, but also avoids some drawbacks in the original brick-wall method such as the little mass approximation, neglecting logarithm term, and taking the term L^{3} as the contribution of the vacuum surrounding a black hole. It is found that there is an intrinsic relation between the event horizon and the entropy. The event horizon is the characteristic of a black hole, so the entropy calculating of a black hole is also naturally related to its horizon.