Table of contents

The conservation law for diffusing and reacting tungsten atoms is used to improve thermodynamical calculations on halogen lamps. The assumption of very fast chemical reaction (mass action law) is shown to lead to contradiction with experience.

X rays and electrons are compared as sources of primary ionization. Electron beams are superior in overcoming problems of signal to statistical noise and in providing fast scan rates, but in some cases with threshold energies above 500 eV x ray ionization is likely to give better signal-to-background ratios and superior Auger spectra if long scan times are possible.

By using one setting of the positions of two flat, spiral heating coils it is possible to generate only the even harmonics of a Rijke tube. For another setting it is possible to generate only the third, sixth, ninth harmonics, and so on. The experimentally determined positions are in agreement with those predicted from the Rayleigh criterion.

The problem of optimizing the uniformity of the magnetic field produced in the bore of a toroidal conductor of rectangular cross section is considered. Both axial and radial variations of current distribution in the conducting winding are investigated. Numerical results are presented for a range of coil parameters.

The spectrum of Barkhausen noise has been measured in iron and analysed using signal analysis techniques. The results suggest that the mean duration of the voltage pulses caused by the domain reversals is less than 5 × 10⁻⁵ s and that a domain reversal triggers other domain reversals with a time delay normally of 1 to 2 ms. Delays of up to 10 ms do occur. The results agree with those obtained by Sawada. The technique allows detailed information of the dynamic nature of domain reversals to be obtained quickly and with little sample preparation.

A method of charged particle trajectory tracing, which is essentially an extension of that described by Birtles in 1972 to include time-varying fields, is described. Methods of error estimation and step-length optimization are given which maintain local truncation errors in a prespecified range. The method is at least an order of magnitude more accurate than the Runge-Kutta method described by Chisholme and Stark in 1970, and is shown to require only half the computer time of both that method and the one described by Merson in 1957.

The performance of the Nd³⁺: POCl₃: ZrCl₄ system as a laser amplifier is investigated using mode-locked pulse trains as the input signals. These have energies of 30 to 80 mJ, and they consist of pulses of width approximately 5 × 10⁻¹² s. Gains of 9 have been obtained from a system 155 mm long and 7·5 mm in diameter with input energies of 1 kJ. The high gains obtained make the Nd³⁺: POCl₃: ZrCl₄ system a serious competitor in the field of high-power amplifier development in view of its advantages over other kinds of systems, namely that the active medium can be circulated and is free of damage problems. Beam patterns obtained show that the most serious problem in this kind of amplifier is thermal distortion due to heating by the pumping radiation. Experimental curves showing the gain, input and output characteristics of the amplifier are presented. They are in good agreement with the theoretical curves of Schulz-Dubois and show similar behaviour to the characteristics of Nd-glass amplifiers obtained with different experimental methods.

Low-temperature thermal conductivity measurements have been used to determine the Cr²⁺ concentrations in five ruby laser rods. The results indicate that the variations in laser output energy of the rods cannot be attributed to differences in their Cr²⁺ concentration.

An experimental technique is described from which plasma velocity distributions have been determined. The technique consists of dropping a line of ball bearings of 3, 4 or 5 mm through a horizontal arc and recording the horizontal deflection of the balls by allowing them to strike a carbon paper. Multiple drops through repeated arcs enabled a deflection curve to be compiled. This curve was then subjected to an Abel inversion so that the radial variation of aerodynamic drag on the balls could be determined. By iterative techniques it was then possible to calculate a plasma velocity distribution compatible both with the Reynold's number and the radial deflection profile.

A comparison with Wienecke's results obtained using photographic techniques on a 200 A arc showed good agreement, although the maximum deflection recorded was less than 3 mm with balls of 3 mm diameter. It was not possible to increase the deflection resolution by using smaller balls, since the scatter in striking position even in the absence of an arc increased unacceptably for balls of 2 mm diameter.

An examination of the validity of the assumptions involved and associated errors is included. In particular the disturbing influence of the ball within the plasma is considered to be small. A more serious difficulty, pointed out by Kimura and Kanzawa concerns the difficulty of assigning the effective temperature and Reynold's number which determines the drag coefficient, since the ball and plasma boundary layer are necessarily at different temperatures. It is argued that the error involved in assigning the undisturbed plasma temperature is small.

The overall error in the velocity distributions, which reach to 1500 m s⁻¹ for the 2160 A arc, is considered to be about ±11%.

Measurements of the average electron energy in the positive column of an He - Ne glow discharge are reported. A Langmuir probe second-derivative method was used for discharge currents from 20 to 200 mA with a 4 cm radius tube and a gas pressure of 0·06 Torr. Electron energy distribution functions for He - Ne mixtures were determined and from these the average electron energies calculated. The experimental values obtained are shown to give reasonable agreement with a modified version of the Schottky theory.

A theory of electron cascade is developed which takes explicitly into account the electron energy distribution. For the particular case of a Boltzmann distribution, it is shown that the monoenergetic approximation is valid, except for low flux densities (<20 MW cm⁻²) of the laser beam. It is emphasized that the equilibrium temperature can be calculated easily. When electron-ion inverse bremsstrahlung comes into play, this approximation becomes invalid; there is a pressure below which the cascade cannot develop. Numerical results are presented for model calculations of helium and argon.

Investigations have been made, with an electro-optical streak camera, of the low-current phases of gas discharges (3 cm gap, uniform electric field, and medium over-voltages of some 10% above the DC breakdown voltage). In N₂-CH₄ mixtures (pressure about 100 Torr) the gas discharge develops under these conditions according to the streamer mechanism. Interest was concentrated on the effect of secondary avalanches following the primary avalanche to see if they are necessary for the development of the cathode-directed streamer. It is shown that the cathode-directed streamer can start unsupported by successors. Further, it is demonstrated that an artificial successor very high carrier number accelerates the development.

The relative permittivity prime and loss factor double prime have been determined for methylacrylate, ethylacrylate, methylmethacrylate and ethylmethacrylate at different temperatures and frequencies. The activation energies and relaxation times are estimated for these liquids. It has been observed that absorption maxima for these liquids are in the millimetre region. The dipole moments are also calculated using Onsager's equation and Rao's equation.

It has previously been reported that a spray originates from high-voltage point electrodes in n-hexane. This paper is concerned with more detailed investigation of that spray. Although it is more readily observed if the point electrode is negative, it can also be seen when the point is positive provided that the gap length is adequate. The distribution of spray in the gap of a point-plane system is approximately conical. The apex of the cone is about 60° for a negative point and 120° for a positive point. The spray appears to consist of bubbles which cross the gap with a velocity of the order of metres per second, and which have grown to become about 15 μm in diameter by the time they reach the plane electrode. Their size at the point electrode is not directly measurable but has been found to be an order of magnitude less by light-scattering measurements. The onset of the spray coincides with the appearance of corona at the tip of the point electrode, and the spray onset voltage is a function of the radius of the electrode tip. An account is given of some of the complex processes which seem to govern breakdown processes in insulating liquids.

A study of the interaction of two injection diodes, mounted in a closed-loop liquid system, has shown that a characteristic fluctuation of the injected currents can occur. An attempt is made to explain this phenomenon by considering the electrically induced liquid flow patterns that could be present in the test cell.

An earlier method introduced by Yadav and Parshad for using Bottcher's equation for dielectric correlation between powder and bulk for wide ranges of permittivity and packing fractions has the limitation that dielectric data for a number of appropriate packing fractions are required. In the present work a formula is introduced, based on the consideration of the effective permittivity seen by a central particle in a powder, which determines the correct values of bulk permittivity over a large range when dielectric data for even a single packing fraction are available.

It has been shown that the logarithm of the dielectric constant of KNbO₃ single crystals under a very low biasing field of up to 60 V cm⁻¹ decreases linearly with the field. Here the effect of small measuring signals is predominantly to create additional nucleations; and this has formed the basis of an equation that fits the experimental data well. At a bias of about 1·8 kV cm⁻¹, a peak in the dielectric constant is observed, probably indicating the coercive field value.

A novel experimental technique based upon the hybrid DC thermoelectric-thermomagnetic transformer has been developed to carry out a comparison of the properties of some promising thermomagnetic materials, namely Bi, Bi-Sb, Bi-Ag, InSb-NiSb and Cd₃As₂-NiAs. The factors that determine the performance of various devices have been studied at room temperature in magnetic fields up to 1·5 T. Of the materials examined, InSb-NiSb is the best for use in a heat-flow meter while Cd₃As₂-NiAs is superior for the other applications.

Accurate measurements of the Richardson workfunction ϕR and the effective workfunction ϕE have been made for the thermionic emission from K₂CsSb bialkali and extended red-sensitive trialkali photocathodes. The sources of error in this type of experiment with photomultipliers, and the experimental and analytical techniques required to minimize these, are discussed in detail. It is shown that the true workfunction corresponds more closely with ϕE than with ϕR.

On the basis of the chemical shift of sorbed gases, an NMR method is proposed for the determination of the static susceptibility of solids without resonant nuclei. The experimental results on the system methane-zeolite Linde 5A give susceptibility values for zeolite in good agreement with the Faraday method, the accuracy being 7 × 10⁻⁷ CGSE units with a wide-line spectrometer and 1 × 10⁻⁸ CGSE units with a high-resolution spectrometer. It is shown that the method is applicable even at very low temperatures where other NMR methods could hardly ever be applied.

The network scheme of pulse-height analysis proposed by Mitra and Hall in 1971 can be used in a modified form for the nondispersive analysis of thin films in accord with the electron probe method introduced by Marshall in 1967. The general scheme for n-element analysis and the calibration procedure for the network are presented in this paper. The advantages of the network method for the nondispersive analysis of thick specimens exist equally for the analysis of thin films and the proposed scheme is therefore widely applicable in the field of electron microscopy.

A three-channel network apparatus described by Mitra has been used to analyse a Permalloy thin film in order to demonstrate the validity of the new analytical procedure.

Average spectral irradiance values are given for total daylight, skylight and south sky irradiances as measured during seven months at Pretoria. Special attention is given to values for air mass 1 to enable comparison with data obtained elsewhere. Some results are also given showing the attenuation of daylight irradiance due to smoke and other pollutants.