ELECTRON EMISSION FROM (FLOWING SOLIDS. 
291 
cm. 2 and gradually increased to 7'7 x 1 0~ 6 amperes cm. 2 during two hours’ heating ; but 
the increase did not stop here, and further heating on the two following days 
increased the emission to 1'1 x 10 -5 amperes cm. 2 at 1500° C., at which value it 
remained constant for some time. This gradual increase in the thermionic current 
was not due to an increasing gas pressure, for this was kept constant by continual 
pumping ; nor did it depend upon the passage of the thermionic current, for it took 
place when the potential difference was removed from the electrodes during part of 
the heating, and, as will be seen from the figures already given, the increase continued 
on one occasion when the cathode was left cold over night with mo electric field on. 
Experiments showed that the current at 1500° C. and 0'005 mm. pressure was 
approximately saturated with 20 volts, but that complete saturation was never 
obtained, the current increasing very slightly with each increase of the applied E.M.F. 
up to 400 volts. 
It seems possible that the smallness of the negative emission from the freshly coated 
platinum may be due to an electrical double layer at its surface which tends to 
prevent the escape of electrons, so that a large electric field has to be applied to 
counterbalance this action. Continued heating may have the effect of removing this 
double layer and hence causing the gradual increase of the thermionic current to a 
steady maximum value. The existence of such double layers is not easily proved, and 
further experiments in this direction are necessary. 
The Chemical Action Theory of the Electron Emission. 
The connection between ionisation and chemical action has been investigated by 
many observers and very inconsistent results have been obtained. The inconsistency 
is perhaps most marked in the case of the alkali metals. By recent experiments 
Haber and Just # and FREDENHAGENt claim to have shown that when the alkali 
metals enter into chemical action there is a liberation of electrons. On the other hand 
Broglie and Brtzard| have concluded that the ionisation which occurs when sodium 
oxidises is due to the formation on the surface of a host of microscopic bubbles, and 
that the effect is merely an example of “ ionisation by bubbling.” A similar result has 
been announced by ReboulJ who also states that the presence of water vapour is 
essential. A large number of chemical reactions which are not accompanied by 
ionisation have been investigated by L. Bloch|| ; these even include certain cases 
where the action is accompanied by flame as in the combustion of sulphur or arsenic. 
But the literature of this subject is so extensive that it is impossible to give an 
adequate summary of it in the present paper. 
* Haber and Just, ‘Ann. der Phys.,’ vol. xxx., p. 411, 1909, and vol. xxxvi., p. 30S, 1911. 
t Fredenhagen, ‘Verb. d. Deut. Phys. Ges.,’ 14, p. 384, 1912. 
\ Broglie and Brizard, ‘ Comptes Rendus,’ vol. 149, p. 923, 1909. 
§ Reboul, ‘ Comptes Rendus,’ vol. 151, p. 311, 1910. 
|| BLOCH, ‘ Ann. de Chim. et de Phys.,’ vols. 22 and 23, 1911. 
2 p 2 
