Abstract
A 180° magnetic velocity analyzer tube has been used to observe the energy distributions of the thermionic emission from various crystallographic directions of a single-crystal tungsten filament. The distributions were the same in all of the directions and were not Maxwellian. An energy-dependent reflection coefficient for the tungsten surface, previously proposed by Nottingham, is capable of explaining the shape of the distributions quite well. The tube permitted measurements of the differences between the true work functions of the various directions. The changes of true work functions with temperature between 1700°K and 2000°K were also measured for all directions except the (110). The non-Maxwellian character of the energy distributions and the temperature variations of the work functions can largely explain the discrepancy between the emission constant, A=120, of Richardson's equation and the Richardson-plot emission constants obtained for the various directions of a tungsten crystal by Nichols and by Smith.