A fine tungsten filament at 400°C in active nitrogen is quickly covered with a layer which produces a lowering of the resistance of the filament. This lowering of the resistance is apparently the result of a lowering of the temperature of the filament which occurs because heat is more readily conducted away from the coated filament by the gas than from the clean filament. At a given temperature, 15-20 percent more heat is conducted away by the gas with the layer present. The same effect is produced by activating the nitrogen by bombardment with electrons. In a tube with a large nickel anode the production of the active agent could be detected down to 10.8±0.5 volts. In a tube with a hot tungsten spiral anode of small area the effect could be detected down to the ionizing potential (16.3 volts). Traces of oxygen cause the spontaneous formation of a layer which renders the filament insensitive to active nitrogen. Traces of hydrogen prevent the formation of the nitrogen layer and remove it if already present. Experiments with a tube of small volume containing tungsten disks of large area show that the amount of nitrogen adsorbed is of the order of magnitude of the quantity necessary to produce a layer of single atoms. Activation of a stream of nitrogen reduces the thermionic emission from a tungsten filament in the stream. Similarly, the striking of an arc in nitrogen under suitable conditions gives a momentary slight increase of current followed by a great decrease, the emission apparently being cut down by the active form of nitrogen produced in the arc. Upon subsequent breaking of the same arc by reducing the anode voltage the current drops and then quickly rises to its much higher former value at that voltage. With an interrupted arc in pure nitrogen, and in argon containing a small percentage of nitrogen, an emission of the $D$ lines persists for several hundredths of a second after the extinction of the arc. The sodium is evaporated from an oxide-coated filament and excited by some active form of nitrogen produced in the arc, presumably atomic nitrogen.

• Received 15 June 1928

DOI:https://doi.org/10.1103/PhysRev.32.799