No abstract

No abstract

Some years ago, while still at the University, I carried out a number of experiments on the electrothermic production of ultramarine ; powdered mixtures of sodium sulphide, china clay and carbon were. interposed between carbon electrodes in a closed crucible furnace. I observed a t that time that in order to keep the electrical resistance and the temperature of the charge fairly low so as to avoid decomposition of the ultramarine as soon as it was formed, it was necessary to use very finely divided carbon, such as lampblack. With charges made up of powdered coke which was coarse compared to the lampblack, I could not get any appreciable current to pass between the carbon electrodes up to potentials of 250 volts. Subsequently, I have found repeatedly that the electrical conductivity of mixtures of finely divided substances is a function of the relative size of the components.Experimental A series of tests was made in order to get values of a more quantitative nature. Two substances were selected, the physical properties of the one as divergent as possible from those of the other: a black metal powder, tungsten, and a white insulator powder, thoria. The advantages in this selection are manifold. Both tungsten and thoria will stand very high temperatures and can, therefore, be made practically moisture-proof. It is a well-known fact that in all high resistance tests adsorbed moisture is a very disturbing factor. As regards metals such as copper and silver, these were not serviceable since they cannot be heated to high temperatures without partial vaporization, which though slight is sufficient to cover the surface of the insulator granules with a highly conductive film. Other factors that decided our selection in favor of tungsten and thoria were: (I) High state of purity;( 2 ) availability of both in extremely fine powdered form