There is evidence in some cases of the emission of electrons from a body in which more than one work function is involved.
It is, however, not established that such an effect may occur at a single surface of a crystal of a pure element.
The emission of electrons from metals is in general very sensi tive to the action of gaseous contaminants. These effects, which are complicated and only partially understood, are mainly indi rect and caused by the modification of the surface arising from the interaction with the gaseous atmosphere. For any given atmosphere there is usually a fairly stable emission condition characterized by fairly definite values of A and b in equation (8.i). These values may, however, differ enormously from the standard values for the pure metal, but they are characterized by the remarkable fact that, no matter by what agencies the changes are caused, the new values always satisfy approximately a relation of the form b=const. x log A +constant • • • • (ILI). A similar relation has been observed in connection with the emis sion of positive ions by hot metals in gases. No observable effects on the electronic emission from hot bodies are caused by the inert gases and vapours such as helium, argon or mercury.
It is required by thermodynamics that the total photoelectric emission of electrons from any substance in equilibrium at tem perature T arising from the equilibrium (black body) radiation imprisoned within it should satisfy an equation of the same form as equation (8) and with an identical exponential factor. This makes it possible to regard the thermionic electron emission from a substance as an integrated photoelectric effect of its temperature radiation. It appears, however, that the absolute value of the emission calculated from optical and photoelectric data is less than one millionth part of the thermionic emission observed at convenient temperatures. Consequently this idea has not made much progress but it is not certain that it is wrong.
When wires of fresh metal are heated it is found that there is a temporary emission of positive ions which usually comes off at a lower temperature than the negative electron emission. Direct measurement has shown that these ions are positively charged atoms of the alkali metals. This emission generally decays roughly as an exponential function of the time at constant tern perature, but the decay may sometimes be preceded by an initial rise. Similar effects from involatile salts, such as aluminium phos
phate, are due mainly to alkaline contaminants. Recently very powerful and steady sources of positive ions have been got by coating metals with an admixture of an alkali salt, and an involatile body such as iron oxide.
If a wire has been heated in a vacuum long enough for the initial positive emission to disappear and gas is admitted at a low pressure there is, in general, an emission of positive ions which is a function of the pressure of the gas. It is believed that the ions in this discharge are charged atoms or molecules of the gas but this is an inference from indirect evidence and there are no direct measurements of the specific mass of the carriers. These discharges frequently exhibit slow changes with time in response to changes in the external conditions such as pressure or temper ature, and there is other evidence that the formation and decom position of surface films of adhering gas play an important part in their generation. For example, the emission in a mixture of nitrogen and oxygen such as air, is very much less than the emission in an equal amount of oxygen with the nitrogen absent. Thus the nitrogen exerts a definite inhibition on the oxygen emission. At low temperatures the emission in oxygen varies nearly as the square root of the pressure of the gas, but at high temperatures it is more nearly proportional to the pressure.
Most salts emit positive ions when heated. With salts of the alkali metals or the alkaline earth metals the ions are charged atoms of the metallic constituent. These are usually singly charged but may be multiply charged if the metallic element is polyvalent. With salts in which the metallic constituent is not very electropositive the positive emission usually observed is caused by alkaline contaminants. Some salts, such as calcium iodide, emit electrons and heavy negative ions in proportions which vary with the temperature but the two emissions may, at suitable temperatures, be comparable with each other.
In all cases when the remaining conditions can be held constant the emission of positive ions from hot bodies follows the same general temperature law as that for electrons, the saturation current being given by i= AT" cbIT, A, a and b being constants and a comparable with unity.
See 0. W. Richardson, "The Emission of Electricity from Hot Bodies" (2nd Edn., 1916). (0. W. R.)