Table of contents

An apparatus for testing stereoscopic acuity has been constructed in which the distance of a reference aperture P is compared with that of a test aperture Q selected from one of a tapering series of aperture plates. The apertures are seen as illuminated discs in a black field and all monocular aids to depth perception are eliminated. The observer has to state whether Q is in front or behind P, and the acuity is derived by plotting the percentage estimates against the distance of Q. Observations have been made for condition A, in which the observer is allowed to look first at P and then at Q, and for condition B, in which the observer fixates on P and views Q extra-foveally. From a comparison of the acuity for the two conditions for various angular separations θ between P and Q, deductions can be made about the relative contributions to stereoscopic vision of convergence and retinal disparity The results suggest that convergence makes a significant contribution and has become the predominant factor at θ = 20° For θ = 14°, the acuity for condition A is due entirely to convergence, since the blind spot interferes with the retinal disparity. Tests with flash exposures suggest that small eye movements do not play any important part in depth perception The presence of depth perception under conditions of gross diplopia was confirmed, indicating that fusion is not an essential item for three-dimensional vision.

By means of certain simplifying assumptions a formula is developed for the penetration into a ductile target by a high-velocity (`Munroe') jet. The action of the jet is divided into two stages, each making its contribution to the total penetration In the first stage a hole is formed by the lateral compression of the target as the jet penetrates it; the second stage begins when the last particle of the jet has ceased to act, the hole continuing to deepen until the residual energy in the target has been spent. At the high pressures set up by a Munroe jet the strength of the target plays only a subsidiary part in the phenomenon.

Work is described in which the two stages of penetration by a high-velocity (`Munroe') jet were separated experimentally. The large penetrations measured in lead targets are shown to result from the flow which takes place in the metal after the jet itself has been consumed For a given jet at a given stand-off it is possible to predict the penetration into a combination of targets from the results of a very small number of standard experiments. The method depends upon the calculation of a certain quantity which is constant for a given jet at a given stand-off, and examples are given of the determination of this quantity from experimental data.

In order to extend the study and use of the elastic properties of geological bodies to include effects involving the amplitude and frequency of elastic radiation, a new source of energy is needed

A large moving-coil transducer is described which, when attached rigidly to the rock, radiates sinusoidal vibrations of controllable amplitude, frequency and duration. The operation of the vibrator is analysed in detail by the method of electromechanical analogy.

For experiments in a chalk-mine, energy at each of the frequencies 300, 600 and 1,000 c/s. was radiated in the form of a square-enveloped pulse, of 20 msec duration. Vibrations were detected in the rock by means of a piezoelectric microphone, amplifiers and an oscillograph, with an overall magnification of about 10⁹. Oscillograms show the complex disturbances produced by the arrival of direct and reflected energy, at distances out to the limit of the mine at 450 ft. Measurements of amplitude at 130 ft. are used to examine the mechanism of the transfer of energy from the vibrator to the rock.

The power radiated was about 0.05 watt at 600 c/s., of which about one-sixth was in the compressional mode and five-sixths in the shear mode, but this power may readily be increased.

The suitability of square plates, vibrating in contour modes, for the determination of piezoelectric coefficients is considered. Piezoelectric measurements on such plates provide checks for the theoretical solution of the various modes and in particular for the distribution of displacement. The recently published solutions for square plates form a basis for this work and the reliability of these solutions is discussed. The materials used were sodium chlorate, sodium bromate, quartz, ammonium dihydrogen phosphate and ethylene diamine tartrate (EDT). New determinations of the elastic and piezoelectric coefficients of these materials have been made, with the exception of EDT, the data for which were published in a previous paper

The magnetic susceptibility of tantalum has been determined absolutely, the value obtained for the mass susceptibility being XTax10⁶=0.8490±0.0006 at 20° C. Small specimens of tantalum have been used to determine, by a comparative method, the susceptibilities of specimens of platinum, palladium and rhodium at 20°C. The values obtained are · XPtx10⁶=0.9712±0.0007, X Pdx10⁶=5.231±0.004, XRh×10⁶=0.9903±0.0008. All specimens were prepared from spectrographically standardized materials.

If due allowance is made for the lowering by the image force of the height of the potential barrier in contact rectifiers, current-voltage characteristics can be calculated which are in good agreement with observed characteristics. On application of a voltage in the blocking direction, electrons (or holes) pass over the barrier owing to their thermal energy, the number of successful passes, i.e. the leakage current, being exponentially dependent on temperature. As the applied voltage is increased and the barrier lowered, the power loss corresponding to the increased leakage current will eventually raise the temperature in the thin barrier layer sufficiently to increase the current in turn until the cooling arrangement provided cannot cope with the increased power loss and the rectifier becomes thermally unstable. The conditions for this instability to arise are investigated quantitatively. Amongst other parameters such as mode of operation, cooling, etc., the conductivity of the semiconductor, and in particular the height of the potential barrier, have a paramount effect on the maximum voltage that can be withstood. The simultaneous requirements for large resistivity and barrier height severely restrict the choice of materials which can be used for efficient rectifiers.

It is shown that the self magnetic field of a high current discharge plays an important part in determining the radial electron density distribution The effect of this magnetic field can be treated by adding a `magnetic potential' term V_m=f^r₀wH dr to the electrostatic potential V_s in the Boltzmann equation for the radial electron density distribution, w being the longitudinal electron drift velocity. The drift of electrons and ions to the tube walls is equalized at low arc currents by a radial electric field. Since the magnetic feld reduces the drift of electrons to the wall whilst having little influence upon the ions, increasing the arc current leads to a progressive reduction of the radial electric field Probe measurements in an arc at low pressures show that at currents over 50 amperes, the radial electric field is zero except for a narrow sheath at the wall The axial density is then mainly determined by the self magnetic field A theoretical expression is derived for the equilibrium density distribution assuming a balance between the diffusion rates due to the concentration gradient and the magnetic forces. Experiments in a mercury vapour arc at high currents support this theory

The excitation and Saha equations applicable to gaseous electrical discharges in thermal equilibrium are used to obtain convenient expressions for calculating the radial concentration variations of electrons and excited atoms when the radial intensity variation of one spectral line is known, and when the temperature on the axis of the discharge can be estimated.