Table of contents

At high temperatures, the electric strengths of the alkali halides appear to decrease rapidly with rise in temperature. This behaviour is qualitatively similar to that predicted by Frohlich's high temperature theory of breakdown. It can also be explained in terms of space charge and thermal effects due to the transport of ions and therefore speculation exists about the operative mechanism. The ionic conductivity of alkali halide crystals depends upon the amount of bivalent impurity they contain. This fact has been used to determine the influence of ionic conductivity upon the electric breakdown of KCl and NaCl. No evidence is provided to support Frohlich's high temperature theory, the observed negative temperature coefficient of KCl being attributed to ionic phenomena. Changes of up to sixtyfold in the ionic conductivity of NaCl did not influence the impulse electric strength of this material except at temperatures below about - 50 c, when the effect of increased conductivity was to increase the electric strength.

A pulse method, developed for the study of effective mobility in insulating solids, is applied to evaporated specimens of vitreous Se. Charge carriers are generated near the top electrode by a 13 mμsec electron pulse and drift across the specimen in a pulsed applied field. The transit time of holes is measured with improved electronic apparatus having a bandwidth of 27 Mc/s. At 20°C the effective hole mobility μ lies between 0.13 and 0.14 cm²sec^-1v^-1. The temperature dependence of μ shows that it is controlled by a level of acceptor states lying = 0.14 eV above the valence band. The values of μ and are in close agreement with those found for single crystals of hexagonal Se which suggests that the centres controlling μ are common to both forms of Se. It is concluded that the hole conduction is determined by the short range order within the Se chain itself or in close proximity to it. The carrier density, however, depends on the arrangement of the chains in the solid. From x-ray data the density of acceptor states is estimated as 10²⁰ cm^-3, giving a value of about 60 cm²sec^-1v^-1 for the microscopic mobility.

The far ultra-violet absorption spectra of the heavier inert gases are characterized by strong, diffuse lines in the region between the ²P_3/2 and ²P_{1 2} series limits. These lines have large natural widths, owing to the high probability of auto-ionization of their upper levels. Their oscillator strengths have been measured in the case of krypton by photographic photometry, using a photomultiplier to calibrate the photographic plates. The f-values found, 0.04 for the strongest line and 0 10 for the sum of all the lines, are much smaller than had been expected on the basis of dispersion measurements. Rough measurements of the continuous absorption in krypton from the ²P_1/2 series limit to 500 Å have been made directly with a photomultiplier, and it is shown that absorption in this continuum, the diffuse lines and the resonance lines probably accounts for only about two-thirds of the refractive index measured in the quartz ultra-violet region.

Some studies of the electronic spectrum given by magnesium in a vacuum are in oxygen 18, and in a mixture of ¹⁶O and ¹⁸O have been made. The measurements of isotope shift are given for the green and for the ultra-violet system. The isotope effect in the green system confirms the vibrational analysis of Mahanti and Lagerqvist. The bands in the ultra-violet region are again attributed to a polyatomic molecule containing magnesium and oxygen atoms. The assignment is partly supported by the isotope shift, but the molecule only contains a single oxygen atom.

The construction, instrumentation and calibration of a pressure-driven shock tube is described. Experiments are reported on interactions occurring between a magnetic field and ionized argon moving at shock Mach numbers 8 to 23 and the conditions for these interactions are shown to be well defined in terms of the Lundquist and magnetic Reynolds numbers.

The modifications to the fluid flow which result from these interactions are discussed in terms of magnetic body forces on the fluid and steady state flow patterns associated with different magnetic field configurations have been photographed. These show that a radially symmetric magnetic field may act as a nozzle to plasma flow, that an asymmetric field may shear the plasma flow and that when the interaction is strong enough a reflected shock wave is produced in the plasma.

The characteristic electron energy loss spectra of the elements of the first group of transition metals, Ti to Cu, have been obtained using the reflection technique. In the spectrum of each metal one loss peak is identified as a plasma loss, and another, except in Ti and V, as a lowered plasma loss. Two other series of loss values are associated with ionization of the M₁ and M_2,3 shells, whilst a peak observed with most metals at about 5 ev loss could be ascribed to ionization of the M_4,5 shell. Other losses have been interpreted as due to interband transitions, or as combinations or multiples of other loss values. The plasma loss values were found to agree with the theoretical values of hw_p, assuming all 3d and 4s electrons to be `free', for the first elements of the series and then to fall below the predicted values.

Optical designers have always had reason to be interested in the question: how well can the practical performance of optical systems be predicted from calculations based on geometrical optics? A large number of experimental investigations have indicated that the scalar wave theory (Huyghens-Kirchhoff diffraction theory) can predict the intensity distribution in the image of a point object well enough for the purposes of instrumental optics. Thus a natural approach to the above question is to consider how well ray-theoretic evaluations of image quality agree with those based on the scalar wave theory. In those favourable cases where the agreement is good, that is to say where the effects of diffraction can be disregarded without introducing unacceptably large evaluation errors, geometrical optics can be expected to provide an adequate basis for practical optical design. The problem of delimiting the favourable cases then arises.

Some results bearing on this problem are obtained in the present paper. In §2 it is shown how, by the use of distribution functions, the contrast transmission properties of ray-theoretic images can be discussed in a way which overcomes the well-known difficulty connected with infinite ray-densities. A sharpened form of a result of Hopkins and Miyamoto is obtained (equation (2.18)) and used to discuss the sense in which the wave-theoretic image of an extended object by a small-field monochromat with given geometrical aberrations approaches the ray-theoretic image as the wavelength λ->0.

In §3 the relations between wave-theoretic and ray-theoretic quality evaluations of photographic images are considered. The importance of the relation between photographic spread and optical spread to the above question is clarified, and it is shown that if emulsion spread is held fixed while the aberrations are multiplied by a constant factor A, wave-theoretic and corresponding ray-theoretic photographic quality evaluations need not approach equality as A->infinity; they will do so, however, if the emulsion spread is also multiplied by A.

In §4, computed quality evaluations are examined for model photographic systems with 1, 2, 4 and 8 wavelengths of fourth-power aberration, working at focal ratios 2, 4, 8, 16 over a range of focal settings in each case. Three wave-theoretic evaluations (3.9) are considered, together with the corresponding ray approximations (3.12). It appears that, over this range of special cases, the ray theoretic evaluation (3.13) can safely be used at focal ratios F less-than-or-eq, slant 4 and the error of the ray approximation (3.12) depends much more strongly on the focal ratio F than on the amount of fourth-power aberration.

The amount of gravitational mass contributed by electromagnetic energy is not clear from previous investigations. In the case of the exterior field of a charged sphere, the charge appears not to contribute at all to the gravitational mass, but other electrostatic and magnetostatic fields give different results.

The field of a circular wire carrying a steady current is investigated by the theory of relativity. An exact solution is obtained for a wire without mass, but this contains a singular surface without physical significance. An approximate solution without this singularity is then obtained, and this shows that the energy of the magnetic field generated by the current contributes to the gravitational mass of the system. The order of magnitude of the effect is discussed with a view to its experimental detection.

Infra-red absorption measurements from 1 to 34 μ have been made on germanium containing varying amounts of antimony and copper at temperatures down to 20°K. The results indicate that the three acceptor levels introduced by copper into germanium can be identified by their effect on the low temperature absorption spectra of suitably doped samples. The appearance of the levels is in good agreement with the existing theory, and in addition to this, an excited state of the shallow acceptor level has been identified which has not been previously reported.

A corrected proof is given of the result of García-Moliner and Simons obtained in 1957 showing the equivalence of the variational formulation and the transport (Boltzmann) equation in the case of an applied magnetic field and effective crystal boundaries. The proof in the paper takes into account the existence of a hyper-surface (in phase-space) of discontinuities of the distribution function.

The (t, p) reactions induced in the target nuclei ¹⁰B, ¹²C, ¹⁶O and ²⁷Al have been investigated at an incident triton energy of 5.5MeV. The angular distributions of the observed proton groups have been compared with a theory of double stripping and good agreement has been obtained with the theoretical predictions for transfers of 0, 1, 2 and (more tentatively) 3 units of orbital angular momentum.

A new level in ¹²B at an excitation of 4.297 MeV is reported. Spin and parity assignments have been made for the levels of ¹⁴C at excitations of 6.582 MeV (1^-), 6.725 MeV (probably 3^-), a hitherto unreported level at 7.009 MeV (0⁺) and at 7.335 MeV (probably 2^-). The existence of levels in ¹⁸O at 3.634 and 4.448 MeV has been confirmed and the spins and parities have been determined to be 0⁺ and 3^- respectively. The spin and parity of the state in ¹⁸O at 3.915 MeV excitation have also been determined as 2⁺.

The behaviour of the ¹⁶O(t, p)¹⁸O reaction leading to the ground and first four excited states of ¹⁸O has been found to be consistent with the ¹⁸O configurations predicted by the intermediate coupling theory.

The atomic mass excess of ²⁹Al has been found to be -9.647±0.015 MeV and the excitations of forty-two new levels of ²⁹Al have been measured.

Eleven resonances have been observed in the yield of gamma radiation produced by the bombardment of chlorine with protons of energies between 500 and 1000 kev. These have been assigned to their parent isotopes either by the use of isotopically enriched targets or from the gamma rays emitted.

The decay schemes of six capture states of ³⁶A have been studied in detail and interpreted in terms of bound states at 1.97, 4.17, 4.45, 4.94, 5.85 and 6.85 MeV. Angular distribution measurements have confirmed that the spin of the 4.17 MeV level is 3; from the modes of decay, tentative spin and parity values have been assigned to some of the other levels.

A series of detailed measurements of the absorption spectrum of high quality single crystals of n-type indium antimonide has been made in the wavelength range 15 to 130 μ, and over the temperature range 4.2 to 90°K.

Some new absorption bands have been found, and an explanation is offered for the positions and temperature dependence of all the principle absorption peaks in terms of multiple phonon interactions involving phonons with energies corresponding to temperatures of 259, 224, 169 and 62°K.

The experimental results for the elastic scattering of high energy electrons from ⁶Li have been fitted using a charge distribution derived from a smoothed finite potential V(r)= -V₀+½Kr², r < r₀ and V(r)= -Ae^-μr, r > r₀, which has identical parameters for s- and p-nucleons. An r.m.s. radius of 2 72±0.08 fermi and a central nuclear density of 0.14±0.01 particles per cubic fermi are obtained. The form factors for inelastic scattering are calculated for three models which fit the elastic scattering and it appears that the inelastic scattering is much more sensitive to choice of parameters. The total energy of the ⁶Li nucleus is calculated on the independent particle model, including a correction for the spurious energy of the centre of mass.

An attempt was made by the author a few years ago to combine the geometrical approach of Putter and the statistical approach of Briggs and Spencer to the problem of ionosphere drift measurement. The combination gave rise to a method of wind measurement which was nearly as simple as that of Putter and yet vigorous enough to eliminate the effect of random changes in the ground diffraction pattern. However, the method could not be proved valid when the ground pattern was statistically anisotropic. The present paper develops a method involving the use of five statistical parameters and a few simple algebraic formulae, which takes into account the effects of random motion as well as anisotropy. The intuitive clarity of the geometrical picture is, however, unfortunately lost.

The results of the measurement of the specific heats in the range 76°K to 250°K of two samples of manganous sulphide are reported. The samples used crystallized in the zincblende and wurtzite structures and are the same as were used in the magnetic susceptibility measurements reported previously. The large anomaly in the specific heat of sample 1 confirms the onset of antiferromagnetic ordering in the zincblende structure at 100°K. The absence of a similar large anomaly in specimen 2 above 76°K shows that the wurtzite structure remains unordered at this temperature, but there is a small anomaly at 110°K A study of the x-ray powder photographs, and the calculation of Madelung's constant for several possible types of stacking faults, show that this anomaly in the specific heat is consistent with magnetic ordering within stacking faults

The possibility that the presence of ordered stacking faults may depress and smooth out the Néel temperature of the rest of the sample is also considered, and is suggested as a possible explanation for the absence of a Néel temperature above 76°K in specimen 2 whilst a specimen composed entirely of the wurtzite structure was shown to order above this temperature by Corliss et al.

The microwave Zeeman spectra of atomic chlorine and bromine have been observed. From measurements on several lines in the spectrum of the ²P_{3 2} ground level, the electronic g-factors are found to be g_J(Br) = 1.333921 ± (5 × 10^-6) g_J(Cl) = 1 333927₅ ± (3 × 10^-6). An estimate of relativistic and diamagnetic corrections to the g-factors, based on self-consistent field wave functions, is in satisfactory agreement with the experimental results. A comparison is made with the g-factors of fluorine and iodine.

Correction to Proc. Phys. Soc.76, 539