Fowler (p. 434) discusses also a somewhat different model, namely, a layer of " bad" semi-conductor between a metal and a good semi-conductor. This is essentially the model we use in the second half of this paper. Fowler's treatment does not seem to us correct however; he assumes that the electrical double layer between the metal and bad semi-conductor would bring the levels of the latter into, equilibrium with the metal, as in fig. 1 (a) . Actually, however, since there are practically no

more »

... rons in the conduction levels of the " bad" semi-conductor, a double layer of this type cannot be formed in-the small space available. (Cf. Mott 1938a.) [ 27 ] on July 20, 2018 http://rspa.royalsocietypublishing.org/ Downloaded from * With F= 104 V/cm., k -10, 293p K it is 0-48.