486. When the spark from a good induction coil traverses a glass tube containing a rarefied gas, certain phenomena are observed which vary greatly with the kind of gas and the degree of exhaustion. There is one appearance, however, which is constant in all the gases which I have examined, and within very wide limits of pressure, viz.: the well-known dark space round the negative pole. I have long been impressed with the idea that this dark space coating the pole was in some way related to the layer of molecular pressure causing movement in the radiometer, and the following experiments were instituted with the object of testing this hypothesis. A glass bulb (fig. 1) was furnished w ith platinum wire terminals sealed into the glass, ending outside in loops and. inside in aluminium poles; the positive pole being a wire and the negative pole a disk about 10 millims. diameter, bare in front and covered with mica at the back. The bulb being full of dry air and connected with the Sprengel pump, was exhausted. An induction coil capable of giving sparks 68 millims. long in air when actuated by 3 Grove’s cells, was connected with the terminals, the disk being always negative except when otherwise stated.