Table of contents

A photon correlator instrument has been applied to the measurement of laser Doppler signals. In 1 s a good signal can be obtained at 1 m in backscatter from unseeded air with 10 mW of laser power.

Experiments show that the use of the homeotropic or homogeneous states of a nematic liquid crystal influences its dynamic scattering properties. The fastest rise time for the onset of dynamic scattering when a voltage pulse is applied is obtained when the homogeneous texture is used. In this case a rise time of the order of 5 ms is observed, and the natural decay time of about 100 ms can be reduced to about 25 ms by the use of an AC signal.

Switching behaviour, rather than dielectric breakdown, is believed to occur in the amorphous elements Si and Ge. We have found that the threshold voltage is dependent on the composition of the electrodes. We have suggested that this dependence is related to the appropriate crystallization temperatures of these amorphous semiconductors sandwiched between certain metals. Crystallization occurs during the switching process and probably stabilizes the temperature after thermal runaway initiates a switching event.

The errors in determining the area under an absorption curve from the first-derivative curve by a numerical integration method are calculated for gaussian and lorentzian lineshapes. The errors are greater for lorentzian lines, where a range of integration greater than 110 times the peak-to-peak linewidth is necessary to reduce the error to below 2%. The application of the method to the determination of ESR spin concentrations is discussed.

In observing the energy changes occurring during magnetization of ferromagnetics and ferrimagnetics three effects referred to as the `dip', `low-field rise' and `kink' have been reported at low fields by Bates and his co-workers, but have remained unexplained hitherto. They are investigated in the present paper for a hypothetical specimen by means of the general energy equation previously given by Davis. The `kink' is shown to be due to the large magnetostriction energy change at fields comparable with H_c. The `dip' is due to the long narrow hysteresis loop making changes of magnetization large at small values of external field. The `low-field rise' is due to the remanent magnetization making the energy curve slightly asymmetrical. Three curves are given.

The thermal energy changes occurring in cyclic magnetization of a 4% nickel-silicon alloy (Curie temperature approximately 34 °C) were determined at 20 °C and 30 °C. The total thermal energy change at 30 °C was about a quarter of that at 20 °C. An equation given previously by Davis is used to show that the changes are almost entirely due to the ordinary magnetocaloric effect with small components due to magnetic potential energy and H_e dI. Results are tabulated.

The presence of nitrogen in small quantities in an argon Z-pinch laser is found experimentally to have a marked effect on the laser properties, 1·0 m Torr of nitrogen being sufficient to reduce the peak small-signal gain of the 3511·1 Å line by 0·85 m⁻¹ in a 17 m Torr argon discharge. The reduction in small-signal gain is accompanied by a reduction in the intensity of the 3511·1 Å spontaneous emission but not, apparently, by any change in the plasma parameters. The collision process responsible for the change in the population of the upper laser level has not been identified. The maximum inversion ratio for the 3511·1 Å line is calculated from the impurity-induced population changes and is found to be 2·5±0·6.

The formation of rotating plasmas in hydrogen has been studied with high-speed photographic, electrical and magnetic diagnostic techniques.

The ranges of pressure and magnetic field intensity were wide enough for both a `spoke' and a uniform `disc' mode of plasma production to be investigated. The effects of magnetic field and pressure on the form are discussed.

A mechanism is considered in which plasma motion in the `spoke' mode is controlled by electromagnetically driven shockwaves.

An experimental study has been made of the influence of an axial magnetic field on the axial motion of the anode root of an arc situated inside a tube down which argon was flowing. The motion was studied by both an electrical and a visual technique. It was shown that the arc root continuously traversed the tube axially under the action of the gas flow over a length which, in the absence of a magnetic field, appeared to depend mainly on the relation between the arc diameter and the tube diameter. The velocity of traverse of the anode root was shown to be approximately equal to that of the cold gas in the bypassing layer surrounding the arc. The effect of a magnetic field on the arc was to cause it to rotate rapidly and to reduce the axial length traversed by the arc. The effect of the magnetic field on the traverse length was found to be in general agreement with theory. The heat flux to the walls was shown in all cases to peak at the downstream end of the traversed length. Thus it was possible to achieve good control of the axial heat distribution by the use of the magnetic field.

Measured axial temperature distributions are given for 3·2 mm ϕ spectrographically pure graphite arc electrodes for currents of 10-40 A. It is shown for the anode that the major portion of these profiles can be described by considering radiation losses only. This enables a mean thermal conductivity to be estimated and this is employed in an energy balance equation which yields corresponding anode fall voltages of 24·7-11·6 V for air.

The sensitivity of short electromagnetic flowmeters may be reduced if they are connected to pipes of high electrical conductivity or if they are immersed in large volumes of conducting liquid. The reduction in sensitivity has been calculated for industrial and physiological meters. The magnetic fields are almost uniform, their uniformity being limited only by their finite length. Where the results predict no appreciable change of sensitivity with such fields, there will be no change with most fields produced by coil systems of the same length. Physiological meters of overall length equal to the tube diameter and with much shorter magnetic fields may suffer sensitivity reductions of 1·2 to 3·2%, depending on the type of field. Industrial meters 1·5 diameters long with fields 0·8 diameters long suffer reductions between 0·9 and 2·0%.

A quantitative theory of breakdown between concentric spheres in atmospheric air is developed, in which it is assumed that a conducting corona cloud, produced at the inner sphere, expands until electron avalanche growth is only just possible at its surface. Breakdown of the gap occurs if the corona sphere radius exceeds a fixed fraction of that of the outer sphere. This fraction differs for positive and negative impulses. Secondary emission processes are included where necessary.

The theory applies to rise times such that equilibrium is established between the applied impulse voltage and avalanche growth at the corona cloud surface (1 μs for negative impulses and 100 μs for positive).

It compares well with experiment for small concentric (hemi) spheres where data are available. It is extended to rod-plane and sphere-plane gaps, where good agreement with experiment is also found. The enhanced strength of sphere gaps is successfully explained and the theory extends naturally to the case of uniform field breakdown.

Simultaneous measurements of electron D/μ (diffusion coefficient/mobility) values, attachment and ionization coefficients are reported over the range 100<E/p<210 in perfluorocyclobutane (c-C₄F₈) and the range 140<E/p<270 in perfluorobutene-2 (iso-C₄F₈). Electron drift velocities have also been measured in these two isomers over the same range. The D/μ values in iso-C₄F₈ are substantially lower than those in c-C₄F₈ and indicate that the difference in the molecular structures of these two isomers is of profound importance in this context. In general, the D/μ values are significantly lower than those in the perfluoroalkanes C_nF_2n+2 (n=1-4) or in SF₆ at comparable E/p values. Comparison of the attachment coefficients at any given D/μ in the range D/μ=1·6-3·2 V suggests that the (total) attachment cross section in c-C₄F₈ is smaller than that in iso-C₄F₈. The drift velocities in iso-C₄F₈ are consistently higher (by about 7·5%) than those in c-C₄F₈.

The electrohydrodynamic stability theory of Atten and Moreau yields precise criteria for the onset of cellular motion in a dielectric liquid subjected to uniform unipolar charge injection; but these are not in terms of convenient experimental variables.

An on-line computer program to give instantaneous evaluation of experimental conditions is described. It has been used to produce curves which enable the consequences of the theory to be discussed with respect to a typical nonpolar liquid. The results show that measurements made in the linear régime, with the techniques of uniform injection available at present, are free from errors of an electrohydrodynamic origin.

Polarity and magnitudes of charges produced by evaporating saline solutions of various concentrations have been measured. Characteristics of drops evaporating from a heated metal and a molten metal have been qualitatively determined in comparison with the rate of charge separation and the point of initiation of separation. Earlier results are confirmed and the relationship between the concentration of solution and the charge magnitude is suggested.

Aluminium films have been produced successfully on gallium arsenide wafers by the new technique of ion bombardment. This involves bombarding the substrate, which is in low-resistance electrical contact with the target, with charged particles of thermally evaporated aluminium produced in the ion source, and depositing the film without any thermal effect. The problems of the technique are discussed. The mechanism of ion bombardment (by ions produced from gas molecules) and mechanisms determining the target current have been investigated.

It is shown that, by this method combined with prebombardment by helium ions, pure aluminium films with any desired characteristics can be prepared on atomically idealized (clean) surfaces of gallium arsenide wafers under controlled conditions.

The formation of mixed crystals in equimolar KCl-KBr powder mixtures is studied at temperatures from 406 to 602°C. The progress of the homogenization is followed by measuring the change in specimen capacitance as a function of time. The results indicate that the following three processes are involved in the mixed crystal formation: (i) grain boundary diffusion, (ii) volume diffusion of Cl⁻ ions, and (iii) volume diffusion of Br⁻ ions. The experimental results for the activation energies associated with these processes are respectively 1·5±0·2 eV, 1·9±0·2 eV and 2·1±0·1 eV.

Dissolution studies were carried out on NaCl crystals, some pure and some doped with various impurities. It was observed that etch pits associated with beaks are obtained only in NaCl crystals doped with CaCl₂, if the etchant reported by Hari Babu and Bansigir is used. Of the 26 poisons studied with methyl alcohol as base, only a few of the cadmium and lead salts are capable of giving beaks in NaCl:CaCl₂, whereas with the other poisons etch pits without beaks are obtained. These investigations reveal that the poison, the solvent and the impurity decorating the dislocation line influence the formation of beaks in NaCl crystals. The substructure of the grain boundaries in NaCl:CaCl₂ crystals has been revealed for the first time, by employing the etching technique. Different types of grain boundaries, junctions between the boundaries and dislocation networks have been recorded.

The internal friction and the modulus anomaly of high-purity, slightly cold-worked aluminium have been measured against temperature. In particular, a complex peak with its maximum at 130 K was observed and studied. It has been found that this phenomenon could be due essentially to a mechanism of interaction between linear and point defects. The Bordoni peak could contribute to the low-temperature part of the anomaly.

An ultrasonic resonance technique has been used in the nondestructive testing of a set of carbon-fibre-reinforced plastic specimens. Both compressional and shear wave resonance as well as internal damping were measured. The results were compared with static modulus measurements, and repeated after the specimens had been cracked in torsion.

The measurements lead to some important inferences about the effects of surface treatment on the fibres and their adhesion to the matrix. The method may have some applications where rapid nondestructive tests of fibre type, alignment and density, or of structural integrity are required.

The distribution of path lenghts through a porous material can be examined experimentally by the transmission of particle beams. The assumption is made that there are no correlations in the sizes of pores and in the distances between pores. This leads to a distribution which depends on four basic parameters, namely the means and variances of the path lengths through and between pores. The first four moments of the distribution are calculated explicitly and are generally related uniquely to the four parameters.

The temperature dependence of the modification of the bismuth III polymorph, which occurs at 42·7 kbar, 20°C, is determined. The results are plotted on a recent pressure-temperature phase diagram for bismuth and give a phase line which is approximately parallel to the temperature axis. The intersection of this phase line with the III-IV line gives a triple point at 41 kbar, 175°C.

The quantum efficiencies of red-emitting gallium phosphide diodes produced by the double liquid epitaxy process using a wide range of p and n type carrier concentrations have been measured. These values have been compared with the calculated injection efficiencies; and near agreement with a theoretical model has been obtained, by incorporating an empirically derived relationship into the calculation to relate the minority carrier diffusion lengths to the carrier concentration. With liquid epitaxy growth techniques it does not appear to be possible to change the injection efficiency for electrons very significantly, the absolute value being normally in the range 0·6 to 0·8. Consequently large improvements in device performance will only be obtained by altering the luminescent characteristics of the materials.

The effects of illumination on the susceptibility of ferric borate are described, and the domain structures that occur in ferric borate are discussed. It is shown that the photo-induced changes agree with a simple domain wall model, the walls probably being of the Néel type. The time and temperature dependence of the photo-induced increases are consistent with a localized centre model in which the centres can exist in two energy states, each with a different anisotropy. The energy required to excite the centres is received indirectly from the photons, via Fe³⁺ ions.

When an electron beam is used to initiate Auger electron spectra, the secondary electrons backscattered from the solid can also cause ionization and Auger emission. This leads to an enhancement of the Auger yield, allowed for by defining a backscattering factor. This paper presents a simple model which enables the backscattering factor and energy dependence of the ionization cross section to be measured. Results are reported of measurements made on silver and silicon specimens.

Chemically prepared thallous sulphide powder, when mixed homogeneously with polymethyl methacrylate, applied on to an insulating support and given suitable electrical contacts, exhibits photoconductivity. Current-voltage characteristics in darkness and under illumination in varying conditions of heat treatment, current-illumination characteristics and the spectral distribution of photocurrent are reported for these layers.

The experimental values of the scattering cross sections of cellulose membrane filters and kaolin layers are determined as a function of wavelength, pore size and void volume, by using a photon diffusion model. The pore sizes and void volumes are determined experimentally by mercury penetration porosimetry. It is found that the rate of change of the scattering cross section with wavelength as a function of pore dimension enables pore sizes to be determined nondestructively from reflectance measurements.

In order to derive refractive indices and coefficients of absorption from reflectance and transmittance measurements it may be necessary to take into account the effects of surface layers. This requires the use of formulae for a double film supported on a substrate. It is shown that the very complicated formulae for reflectance R₂ and transmittance T₂ for normal incidence on such a double film may be replaced by expressions for (1 ± R₂)/T₂ which can be expressed explicitly in terms of the n and k values for each layer. These formulae are relatively simple and very convenient for computing purposes.

It is also shown that under certain circumstances a double layer may be replaced by an equivalent single layer.

The derivation of the components n and k of the complex refractive index of a thin film from measurements, at normal incidence, of reflectance R and transmittance T is complicated by the existence of multiple solutions of the relevant equations. It is shown that, although use of Tomlin's expressions for (1 ± R)/T rather than Heavens's explicit formulae for R and T separately simplifies the problem, it is still necessary to consider very carefully the multiple solutions. A procedure is given for determining the correct solutions for n and k, and also for accurately fixing the film thickness, provided that the film is a homogeneous single layer on a substrate.

Some experimental results have not yielded to this treatment and in these cases it has been found necessary to assume the presence of a surface layer (often, but not necessarily, an oxide layer) and to make use of the equations in the preceding paper which relate to a double layer on a substrate. It is shown how this may be done in such a way as to obtain the correct dispersion curve for the material, together with its thickness, and that of the surface layer.