Broadening of visible Kr i, Kr ii, and Xe ii lines is measured behind reflected shock waves in a diaphragm shock tube. Revised estimates of broadening by competing mechanisms reduce the Stark parameters for seven lines of Xe ii measured earlier in the shock tube. Stark broadening for 30 lines are obtained, including 16 not previously reported. Results are compared with findings elsewhere, and systematic broadening trends in the homologous ion sequence are examined.

For optical homodyning, the matrix representation of a lossless beam splitter belongs to the SU(2) group of unimodular second-order unitary matrices. The connection between this group and the rotation group in three dimensions permits the field density operators at the input and output ports of the beam splitter to be related by means of well-known angular-momentum transformations. This, in turn, provides the joint output photon-number distribution, which may be written as a Fourier series in the relative phase shift imparted by the beam splitter, for a general joint state at its inputs. The series collapses to a single term if one of the input fields is diagonal in the number-state representation. If the inputs to both ports are further restricted to be pure number states, the joint, as well as the marginal photon-number distributions, turn out to be directly proportional to the square of Jacobi polynomials in the beam-splitter transmittance. These photon-number probabilities are invariant to a set of physical and time-reversal symmetries. When one of the input photon-number states is the vacuum, the beam splitter simply deletes photons from the other port in Bernoulli fashion, as if they were classical particles. The output photon number is then described by the binomial distribution. If the inputs at the two ports are different number states, neither of which is the vacuum, the photon-number distribution is expressible in terms of summed and weighted products of the results for photomixing with the vacuum. If the inputs at the two ports are identical number states, and a beam splitter of transmittance τ=(1/2 is used, the photon-number distribution assumes a simple but interesting form. It vanishes for odd photon numbers, indicating that the photons assemble in pairs at each output port. Finally, it is shown that homodyning quantum fluctuations can be reduced by using a balanced photomixer for arbitrary input states.

Excellent squeezing in intracavity second-harmonic generation has been predicted to occur on cavity resonance in the output intensity fluctuations. Cavity detunings cause laser phase noise to couple in and reduce the squeezing observable. Here we consider the effects of laser phase fluctuations on the output-squeezing spectrum. Laser phase noise is modeled as an Ornstein-Uhlenbeck (colored-noise) Gaussian stochastic process and its effects are compared with the white-noise limit. This indicates that the white-noise model may qualitatively overestimate the deleterious effects of laser fluctuations on sideband squeezing. We compare our results with the recently reported experiment of Pereira et al. [Phys. Rev. A 38, 4931 (1988)] and present an analysis of the empty cavity for comparison.

A quantum-mechanical theory is developed to treat the interaction of a multilevel atom with cavity fields of arbitrary detunings. The Hilbert space spanned by the energy eigenvectors is divided into subspaces specified by eigenvalues of the total excitation number, which is a constant of motion. Since the total Hamiltonian does not connect states in different subspaces, it can be diagonalized in each subspace independently. The time evolutions of the level occupation probabilities and the mean photon number are investigated numerically, and their variations with the atomic level number and the initial photon number are discussed. Their relation with the field squeezing is also discussed.

A study is made of the operation of the electron cyclotron resonance maser (ECRM), when subjected to an external rf signal. The signal is introduced both via direct injection through a coupling hole in the oscillator and by modulating the electron beam in separate cavities, upstream of the oscillator. Experiments using both one and two ‘‘prebunching’’ cavities are reported. It is experimentally found that the gyromonotron, a specific embodiment of the ECRM, can be phase locked by premodulating the electron beam at drive power levels more than two orders of magnitude below that predicted by Adler’s widely applicable theory for single-cavity oscillators. A perturbation theory is used to predict the phase-locking frequency band for a gyromonotron with any number of prebunching cavities. The predictions of this theory agree with the experimental results for two- and three-cavity phase locking. An investigation is made into the general amplitude and frequency response of the ECRM to an applied external signal. Experimentally, three distinct regions of qualitatively different ECRM behavior are noted: soft excitation, which is free, self-excited oscillation; hard excitation, where the oscillation requires an external impulse for startup; and amplifier, in which the output power level and frequency are linearly related to the drive in the small-signal regime.

An improved adiabatic elimination technique, previously applied to resonant-laser equations for a small population decay rate, is extended to the general detuned case. The initial five-dimensional system is approximated with a set of only two differential equations. This allows an analytical investigation of the bifurcation properties of the second laser threshold. The final two-dimensional model represents a substantial improvement in the correct description of many lasers, from ${\mathrm{CO}}_2$ to Nd:YAG (YAG denotes yttrium aluminum garnet), from semiconductor to NMR lasers.

A proposed experiment is analyzed theoretically. In the proposed experiment two coherent pump waves fall on two identical nonlinear crystals, down-converted signal and idler beams from the two crystals are mixed by two beam splitters, and the coincidence counting rate for photons leaving the beam splitters is measured. We show that this counting rate depends on the phase difference between the two coherent pump waves, and results from the interference of the vacuum with the down-converted photons. The experiment could be used to look for locality violations along the lines recently proposed by Grangier, Potasek, and Yurke [Phys. Rev. A 38, 3132 (1988)], but without the need for a coherent reference beam for homodyning.

It is possible to obtain gain from a free-electron laser without spatially bunching the electrons, and several mechanisms are analyzed for accomplishing this. The effect can be particularly important for low-quality beams, both for providing gain and increasing efficiency.

We show that when a coherent light beam is sent through a passive bistable device, the output beam may exhibit a substantial reduction of photon noise below the shot-noise level, within some frequency range of the noise spectrum. This effect is not limited to the case of a high-finesse cavity and is observable as soon as a bistability turning point is approached. Linear losses have the unexpected consequence of improving the noise-reduction factor at low frequencies.

We present a detailed theoretical treatment of the response of a two-level atom to a fully-amplitude-modulated, narrow-band optical field. Specifically, we calculate the resonance behavior of modulated fluorescence and susceptibility components from the scalar continued-fraction solution of the optical Bloch equations. In the limit of vanishing damping the resonances are generally doubly branched: One set of branches reproduces the resonances of the time-averaged fluorescence, while the second set is different for each harmonic component. However, each branch is uniquely associated with an odd multiple quantum resonance. In analogy with magnetic resonance work, we derive analytical expressions for the generalized Bloch-Siegert shifts of the additional resonances. We discuss also the possibility of experimental verification of the theory.

Analytic aspects of the photon-counting distributions in quantum optics are investigated for the case where the underlying field includes a mixture of harmonic signal and thermal noise. A Lorentzian spectral profile for the thermal light is used to find simple iterative relations for the factorial cumulant moments. Explicit analytic expressions for the moments are then given to eighth order. Approximate interpolation formulas, which display proper limiting behavior at small and large correlation length, are also investigated.

Exact solutions in the form of infinite series are well known for the Jaynes-Cummings model describing the interaction of a two-level atom with a single-mode field prepared in a coherent state. While these solutions have led to the discovery of many interesting properties of the Jaynes-Cummings model as a function of time, the nature of the solutions is such that it is not a simple matter to make statements of a general nature about the properties of the system. In this paper it is shown that by taking infinite time averages of the series expressions for the atomic inversion and inversion correlation, squeezing, and antibunching, simple closed-form expressions are obtained from which general properties of the Jaynes-Cummings model can be readily inferred. In particular, new conditions are obtained under which the field is significantly squeezed on average, and further information is obtained on the long-time behavior of the revivals in the atomic inversion.

Essential mechanisms limiting the achievable pulse duration in synchronously pumped, hybrid-mode-locked, and passively mode-locked cw subpicosecond dye lasers are analyzed. The action of spontaneously emitted light is shown to be the detrimental factor in synchronous mode locking. The influence of this factor can be suppressed by a saturable absorber in the hybrid-mode-locked lasers. In this case as well as in the case of passive mode locking the combined action of group-velocity dispersion and Kerr-effect-type nonlinearity, though serving as an additional shortening mechanism, causes fluctuations of the pulse parameters and determines the limit of the steady-state regime.

Magnetic dipole transitions within the ground-state configuration of Si-like Ni, Cu, Zn, Ge, and Se have been observed. Seven lines have been newly identified. Observations are made on the Texas Experimental Tokamak by laser-ablation injection of each of these elements into the plasma. The radiative-branching-ratio technique, utilizing lines originating from the same upper level, 3$p^2$ ${}_{}{}^1{}_{}{}^{}D_2^{}{}_{}{}^{}$, is used for the radiometric calibration of a 1-m normal-incidence spectrometer. This calibration is in good agreement with the absolute radiometric calibration obtained by using an argon miniarc.

Spectra of highly ionized krypton in the wavelength range 40–335 Å have been observed in the Joint European Torus tokamak plasmas, into which small amounts of Kr were admitted. Among the observed lines are all the strong resonance lines within the n=2 shell from fluorinelike to lithiumlike krypton, as well as all the stronger magnetic dipole lines that are expected to occur in these spectra. The wavelengths of the krypton lines were determined using known wavelengths of lines of elements intrinsic in the plasma, mostly carbon and nickel lines, as standards. In most cases of stronger lines, the accuracy is estimated to ±0.025 Å, or better.

In this paper normal stresses are included in the calculation of flow profiles in tube flow of a non-Newtonian fluid. The velocity profile obtained is shown to reduce to the well-known Hagen-Poiseuille parabolic velocity profile as the fluid is made Newtonian. The effects of normal stresses become increasingly significant as the pressure difference increases and the fluid becomes more non-Newtonian. Inclusion of normal stresses gives rise to a critical parameter directly depending on the pressure difference beyond which the flow variables behave singularly and there appears an immobile boundary layer sticking with the tube wall. Thus the flow looks like a plug flow that leaves a boundary layer behind, while the velocity of the mobile pencil of fluid becomes divergent. It is argued that this divergence can be removed by cutting off the domain where the singularity resides, on the grounds that removal of some of the assumptions made for the present study would remove the singular behavior and the cutoff procedure is in some way equivalent to the removal of such assumptions.

Simple analytical expressions for equilibration times of nonrelativistic monoenergetic electrons in plasmas are evaluated in the ‘‘weak’’-beam approximation when the density of the monoenergetic electrons is much smaller than the plasma density. The equilibration time is defined as the time needed by the beam of monoenergetic electrons to lose most of its energy as a result of collisions with plasma particles having a Maxwellian energy distribution. The process of the energy equilibration is treated as a statistical superposition of both elastic (electron-electron, electron-ion, and electron–neutral-particle) and inelastic (electron–neutral-particle) collisions in the plasma. The possibility of collisionless equilibration is also discussed. Comparison of the equilibration times with the Spitzer relaxation times indicates that the former times are more appropriate for an estimate of the energy loss of the ‘‘weak’’ electron beams in highly ionized plasmas. The approach of this work can be generalized in a straightforward way to beam-plasma and beam-gas systems with ionic and neutral-particle beams.

It is shown that for an axially periodic column of magnetofluid driven by an applied axial electric field, the total rate of energy dissipation (Ohmic plus viscous) can be lowered by permitting a helical component with vortical flow in the solution. The principle of minimum energy-dissipation rate suggests that this partially helical state will be preferred to the axisymmetric one that exists for the same parameters. The result is consistent with the repeated appearance of such partially-helical states in several fully three-dimensional numerical computations and is not inconsistent with the data from some confinement experiments.

The statistics of many-arm star polymers in good solvents are analyzed by using the magnetic analogy in the n→0 limit (n is the number of spin components) and the renormalization-group ε expansion (ε=4-d, and d is the spatial dimensionality). The configuration-number exponents, i.e., γ(f) for star polymers with f arms, are derived exactly to O(${\epsilon }^2$). In order to obtain a suitable result for a large number f of arms, I made a resummation of terms that become dominant for large f in each order of ε and derived a self-consistent equation for an effective single-chain propagator. This equation is solved at criticality, yielding the configuration-number exponent behaving as γ(f)∼-$f^{d/\left(d\mathrm{-}1\right)}$ for d<4 and f≫1, which is similar to the recent result of Witten and Pincus [Macromol. 19, 2509 (1986)]. Moreover, I discuss another treatment (the cone approximation) assuming that each arm moves freely inside each different circular cone and also derive a similar result of γ(f).

The linear and nonlinear hydrodynamics of the two-dimensional lattice-gas automata (LGA) are discussed. The physics of the LGA is found to be richer than previously expected. Together with sound and shear waves (characteristic of simple fluids) there are three new hydrodynamic modes. The conserved quantities corresponding to the latter arise from a feature of the microscopic definition of the LGA; i.e., the particles of the microscopic gas occupy the sites of a regular lattice and can only hop from one site to its nearest neighbors. The presence of these new conserved quantities has unexpected results on the macroscopic behavior of the fluid. In fact, there is a nonlinear coupling between the two classes of modes, and while the new conserved densities are merely convected by the momentum density, the current of the latter contains terms that depend only on the new modes. Thus the presence of a finite amount of the new conserved densities produces flow patterns that are not solutions of the Navier-Stokes equation. Although only the two-dimensional hexagonal-lattice gas is discussed, the arguments described here apply with equal force to currently proposed three-dimensional models.