A mgnetically controlled null-point modification of the rotating crucible viscometer has been devised which enables measurements of the viscosity of liquids to be made at maximum temperatures of 1850° C. The instrument has been applied to measure the viscosity of the system CaO -SiO 2 over the composition range 30 to 58 mole % CaO. The flow is Newtonian . The viscosity isotherms are smooth functions of composition and donot exhibit the inflexion found in previous work carried out in (attackable) graphite crucibles. Up to 1700° C the energy of activation is constant for all systems. For compositions containing more than 38-8 mole % CaO, the energy of activation falls at temperatures 1700° C. The energy of activation is small compared with that for flow in liquid silica. Over the composition range studied it falls little (45 to 35 kcal mole -1 ) with increase of CaO content. The energy of activation observed is inconsistent in magnitude and variation with composition with the current theory of the structure of liquid silicates and glasses due to Endell & Hellbriigge. In liquid SiO 2 and systems containing less than about 12 mole % metal oxide the flow unit is SiO 2 . At about this composition a fundamental change occurs in the structure of the melt. For systems with a metal oxide content of more than about 12 mole % the flow unit is a discrete silicate anion, the size of which increases regularly with increasing concentration of SiO 2 . The simplest discrete silicate anions compatible with valency, stereochemical and electro-neutrality considerations are deduced for various compositions of the liquid. The structure thus suggested is consistent with observations on the viscosity of silica and of the system Na 2 O-SiO 2 ; and with results on the temperature of maximum density in silica, X-ray diffraction in glasses, the solubility of CaF 2 in silicates, and the partial molar volume results for simple alkali metal glasses.

Abstract-The optical layout of erbium-doped waveguide amplifiers (EDWAs) based on spiral and folded spiral planar lightwave circuits is considered. It is shown that layouts may be ranked according to their ability to provide maximum gain within a given chip area and an optimum layout based on a spiral containing both straight and curved waveguide sections is identified from an initial set. The analysis is initially based on simple geometric principles. A detailed optical simulation is then carried out using an algorithm that links a five-level rate equation model with beam propagation by the method of lines. The results confirm the choice of optimum layout.Index Terms-Erbium-doped waveguide amplifier (EDWA), optical amplifier.

No abstract