1883.] on the Size of Atoms. 195 



a centimetre, being something less than a quarter wave-length of 

 violet light in the oxide. 



The exceedingly searching and detective efficacy of electricity 

 comes to our aid here, and by the force, as it were, spread through 

 such a film, proves to us the existence of the film when it is consider- 

 ably thinner than that l-100,000th of a centimetre, when in fact it is 

 so very thin as to produce absolutely no perceptible effect on the 

 reflected light, that is to say, so thin as to be absolutely invisible. 

 If in the apparatus for measuring contact electricity, of which the 

 drawing is before you ('Nature,' vol. xxiii. p. 567), two plates of 

 freshly polished copper be placed in the Volta condenser, a very 

 perfect zero of effect is obtained. If, then, one of the plates be 

 taken out, heated slightly by laying it on a piece of hot iron, and 

 then allowed to cool again and replaced in the Volta condenser, it is 

 found that negative electricity becomes condensed on the surface thus 

 treated, and positive electricity on the bright copper surface facing 

 it, when the two are in metallic connection. If the same process be 

 repeated with somewhat higher temperatures, or somewhat longer 

 times of exposure to it, the electrical difference is augmented. These 

 effects are very sensible before any perceptible tint appears on the 

 copper^ surface as modified by heat. The effect goes on increasing 

 with higher and higher temperatures of the heating influence, until 

 oxide tints begin to appear, commencing with buff', and going on 

 through a ruddier colour to a dark-blue slate colour, when no farther 

 heating seems to augment the effect. The greatest contact-electricity 

 effect which I thus obtained between a bright freshly polished copper 

 surface and an opposing face of copper, rendered almost black by 

 oxidation, was such as to require for the neutralising potential in my 

 mode of experimenting * about one-half of the potential of a Daniell's 

 cell. 



Some not hitherto published experiments with polished silver 

 plates, which I made fifteen years ago, showed me very startlingly 

 an electric influence from a quite infinitesimal whiff of iodine vapour. 

 The effect on the contact-electricity quality of the surface seems to 

 go on continuously from the first lodgment, to all other tests quite 

 imperceptible, of a few atoms or molecules of the attacking substance 

 (oxygen, or iodine, or sulphui', or chlorine, for example), and to go 

 on increasing until some such thickness as l-30,000th or l-40,000th 

 of a centimetre is reached by the film of oxide or iodide, or whatever 

 it may be that is formed. 



The subject is one that deserves much more of careful experi- 



^ mental work and measurement than has hitherto been devoted to it. 



I allude to it at present to point out to you how it is that by this 



* First described in a letter to Joule, published in the 'Proceedings of the 

 Literary and Philosophical Society of Manchester ' of Jan. 21, 1862, where also I 

 first pointed out the demonstration of a limit to the size of molecules from 

 measurements of contact-electricity. The mode of measurement is more fully 

 lescribed in the article of ' Nature ' (vol. xxiii. p. 567), referred to above. 



