The dielectric properties of a number of synthetic aluminosilicate minerals have been determined by means of a.c. measurements. A Hartshorn—Ward dielectric test set was used for the measurements in the range 74kc/s to 2mc/s and at lower frequencies a bridge with inductively coupled ratio arms was employed, The temperatures at which the measurements were made were within the range —196 to 35000. The lithium minerals studied have reasonably low loss values which increased with decreasing frequency, The loss/log,(frequency) plots for 4,—eucryptite give maxima within the measuring range, which are ascribed to relaxation of trapped electrons. The measurements on nepheline hydrate and its ignition product show that these two materials have very similar values for the dielectric loss and for the temperature coefficient of the loss, This suggests that the water in nepheline hydrate is strongly bound in the crystal and does not make a large contribution to the loss of this material at the temperatures used in the measurements, The barium feldspar, celsian, and the synthetic strontium polymorph of feldspar both exhibit conduction of a d.c. nature at the higher temperatures used. The conduction in the potassium feldspar, sanidine, has a large contribution from d.c. conduction circa 25000 and indications are that it will be mainly of that type at somewhat higher temperatures. The anhydrous sodium chabazite has loss peaks within the measuring range. These are attributed to cation jumps. A slightly hydrated sample of sodium chabazite has much lower loss values, with no maxima, in the region where the peaks occur for the anhydrous sample.

The first part of the thesis deals with the tight-binding method of band structure calculations particularly as applied to 8 bands' end the second part with the role of the d band in magnetism. A fundamental error in the tight-binding method arises from the use of a restricted set of trial fumigates A method of calculating an upper bound to this error is given and an exact calculation at the error is made for a chain of d-functions. The general conclusion is that the tight-binding method is only accurate when the nearest neighbor's approximation is a good one. A method of calculating correction terms is given which is valid in cases of strong bindings the use of a 'muffin-tin' crystal potential is advocated. A criterion for choosing the 'atomic' functions is applied to the problem of calculating d bands. A correct choice ensures reasonable accuracy for states which do not interact with the sp band. A tight-binding calculation of the d band in copper is described; this employs a convenient method of reducing the energy integrals to the smallest number of independent ones and a new method of calculating three-centre integrals. A band treatment of spin waves in metals is averts The thermo-dynamic, properties of a ferromagnetic metal are discussed in terms of spin wave excitations and independent particle excitations. Herring's and Kittel's conjecture on the additivity of the two contribution to the specific heat, and to the deviation of the magnetization from saturation, is confirmed. The status of Stoner's collective electron theory is discussed. A formal proof is given of the existence of spin craves of long wavelength and a general method of calculating their energies, which depend strongly on the band structure, is outlined. Herring's and Izuyama's results on spin-wave energies are obtained as special cases.

This thesis is concerned with a study and evaluation of the use of ferrites to produce non-reciprocal attenuation in slow-wave structures for high-power travelling-wave amplifiers. Two structures' are investigated in detail, one suitable for a high-power "0"-type amplifier, the other for a high-power "M"-type amplifier. The "0"-type amplifier structure investigated is the folded waveguide. Not only is this structure eminently suitable for high-power applications, but it is a relatively easy matter to achieve non-reciprocal attenuation in it by means of rectangular waveguide-typeisolators. As a result, the folded-waveguide has been'the subject of two previous and independent investigations. In each case a theory of the folded-waveguide was developed and a non-reciprocal structure was constructed. Although the two theories are identical, the non-reciprocal structures built differ considerably. One structure was for c.w. operation at X-band with an 11 watt beam, the other for pulsed operation at S-band with a 0.8 megawatt beam. In neither case was the structure, with or without ferrites, tested in a working amplifier. The work on the folded-waveguide structure described here is a continuation of that which led to the development of the non-reciprocal X-band structure. The design, construction and performance of an electron gun for this structure are described, and the results of "0"-type amplifier measurements on it are given. Although the design gain of the tube was 15db., no amplification was obtained. The failure of the tube to amplify is discussed. It is shown that this was due to a design fault. By neglecting to take into account space charge debunching of the electron beam, the structure had been designed with too low a coupling impedance. A detailed analysis of the folded-waveguide structure and ways of producing non-reciprocal attenuation in it follows. It is concluded that, in general, when the structure is designed so as to be suitable for amplifier operation, it is no longer possible to incorporate rectangular waveguide-type isolators in it. The "M'-type amplifier structure investigated is the ridge loaded ladder line. The basic theory of the structure is given, and the results of dispersion, coupling impedance and "0"-type amplifier measurements on an X-band circuit are presented. The design of an "M"-type amplifier making use of this circuit is then described and predicted performance figures for it are given. The development of a non-reciprocal attenuator system for this tube is discussed, and the results of measurements on the non-reciprocal structure are presented and evaluated. It remains for the structure to be tested in a working "M"-type amplifier.

Measurements of infrared and Raman spectra have been used to obtain structural information on a range of transition-metal-oxygen complexes. In order to apply the Raman technique with mercury arc excitation to as many of the systems as possible, the studies have generally been restricted to the colourless or pale yellow complexes of the metals of Groups IVa, Va and VIa in their highest oxidation states. By recording the Raman spectra and complementary infrared spectra of alkaline solutions of oxyvanadium(V) anions at various pH values, it has been possible to identify the species present in such solutions and to suggest structures for these. This technique has also been used to investigate the possible existence of monomeric tetrahedral oxyanions of niobium and tantalum in solutions of high pH, and the existence of low molecular weight polymers in isopolytungstate solutions. A study has also been made of the species present in acid solutions of the early transition metals; the complexes present are mainly anionic, and formulae and structures are proposed for them. During this study, the presence of complexes with two cis oxy groups was postulated. The infrared and Raman spectra of a number of dioxy complexes were therefore measured, and the cis dioxy configuration found to be more widespread than generally realised. The spectral characteristics of such 6. compounds, and also of some trioxy complexes, have been tabulated, and an explanation proposed for the relative existence of cis and trans dioxy systems. The spectral characteristics of a number of peroxy complexes have also been recorded; it is thought that the results may be used to identify the mode of bonding of peroxy groups in unknown complexes. Finally infrared and Raman spectroscopic measurements have been used to suggest structures for a number of early transition metal peroxyoxalato complexes.

No description available.

The intention throughout this work has been to correlate the many factors involved in the metabolism of the nitrogen fixing species A. vinelandii in order to present a comprehensive unified whole. The various techniques employed have been strictly adhered to, after a probationary trial period used to evaluate their different potentialities; with the result that comparison of values obtained over a period of some months has been attempted. Continuous culture has been shown to be a useful asset in the production of cell material and to have possibilities of an interesting development in further studies of the growth metabolism of the bacteria. The ultrasonic method of disruption of Azotobacter has been extended and developed into a satisfactory routine procedure associated with centrifugal separation. An investigation of the third step in the reduction of nitrogen (i.e. N - N bond reduction) has been carried out and forms a stage in the complete examination of reduction of nitrogen-nitrogen compounds undertaken in this laboratory. Electron microscopy has been used to examine the progress of disruption and also to establish some morphological details of the bacteria.

A linear AH....A model, representing hydrogen bonding, which incorporates a symmetrical double minimum in the potential energy, and interaction between the symmetric and antisymmetric motion, is examined in detail. The asymmetric system A-H....B is also briefly examined. Using harmonic oscillator functions as a basis for expansion, solutions of the two dimensional wave equation are obtained. Three tests for the convergence of the matrix expansion are satisfied. A set of potential parameters describing a typical hydrogen bonded system is chosen. A method for obtaining the interaction constant is described.

The distributions of polyanions in binary phosphate glasses of sodium, cadmium, magnesium and lead oxides were determined by a paper chromatographic technique. The effect of making these glasses with rigid exclusion of moisture and the effect of quenching rates on the constitution of these glasses were investigated. The results indicate that the distributions are a function of cation and the mean chain length E. The breadth of these distributions is directly related to thw shape of the heat of formation curve for the glasses (from the constituent oxide), and to the cation radius to charge ratio of the cation to the extent this determines the heat of formation. The glasses made under dry conditions and also' •by fast quenching had anion distributions the same as those made in the conventional manner. The thermodynamic properties of PbO, Pb0+Si02, PbO+ P205, Pb0+B203 and PbO+PbF2 melts were obtained by measuring the emf of the cell, Pb / Pb0 containing melts / 02 : Pt The measurements were usually carried out in the temperature range 850°C to 1050°C. From the emf and its dependence on temperature the activities of PbO, and the partial free energies, heats and entropies of solution of Pb0 in Pb0 SiO2' Pb0 + P205, Pb0 + B203 and Pb0 + PbF2 melts were derived. The corresponding thermodynamic properties of the other components in these binary were derived using the Gibbs-Duhem relationship and these were combined with those for Pb0 to obtain the integral thermo-dynamic properties. The interpretation of observed Pb0 activities for the lead borate and Phosphate melts, more basic than the ortho composition, in terms of structural models seems to indicate the presence of three and four coordinated boron atoms in the borates and the possible existence of five coordinated Phosphorus atoms in the phosphate. The structural information gained on the phosphate glasses has been extended to the silicate and borate systems. Various available models, which relate the thermodynamic properties and structure of these polyanionic melts are assessed;It seems that Meadowcroft and Richardson's model is the only one capable of relating quantitatively the structure and thermodynamic properties of such melts. The thermodynamic data obtained for Pb0 PbF2 melts shows these to be regular solutions, with an ideal entropy of mixing as defined by Temkin. Unsuccessful attempts to study the anionic distribution in silicate glasses by a chromatographic separation technique are described in the appendix.

The sorption and encapsulation properties of a variety of zeolitic minerals have been investigated using the rare gases helium, neon, argon and krypton as sorbates. High pressure (up to 15,000 p.s.i.) and high temperature (up to 580°C.) techniques were used to effect sorption of the gases. The high pressure sorption apparatus was described in detail. Minerals investigated included the crystalline silica polymorphs, sodalites, cancrinites, stilbite, heulandite and the synthetic zeolite Rb-M. Preliminary results were reported for mesolite, scolecite, thomsonite, attapulgite, sepiolite, beryl and cordierite. Sorption isotherms and desorption kinetic isotherms were determined in most cases, enabling calculations to be made of heats of solution and activation energies for diffusion. Where macro-crystalline samples were used (silica polymorphs, stilbite and heulandite) sample surface areas were estimated using a projected area method, and diffusion coefficients thus evaluated. The desorption kinetic isotherms were characteristically sigmoid in shape, having a non-isothermal initiation period. Diffusion in such a system had not been considered previously, and an applicable solution to the diffusion equations was developed and described in detail. Heats of solution were determined for the following systems: helium in a-, β1- and β2- tridymites, and a- and β- cristobalites; neon in β2-tridymite, β-cristobalitei, basic-cancrinite and heulandite; argon in stilbite and basic sodalite; krypton in Rb-M, basic sodalite and extracted basic sodalite. Activation energies for diffusion were also calculated for the above pairs, and in addition for argon in Rb-M, stilbites and heulandites; and krypton in stilbites. Diffusion coefficients were evaluated for helium in a-tridymite and a-cristobalite; neon in a-, β1- and β2-tridymites, heulandite and stilbite; argon in various stilbites and heulandites; and krypton in stilbites. Using the Lennard-Jones 12-6 potential detailed theoretical calculations were made of the interaction energies between the rare gases and the oxygen frameworks and cations in the structures of sodalite, β2- tridymite, a-cristobalite and β-cristobalite. In all cases agreement was good between calculated and experimental parameters using the London and Slater- Kirkwood dispersion constants.

No description available.