1253 



2,2 Volts, if the current passes froni the sulplmric acid to tlie 

 iiiolybdemnnpoint. It is about 0,75 Volt in the opposite direction. 

 This mncii smaller E.M.F. of polarisation will according to Armac^nat 

 be found if the anode consists of a metal soluble in the liquid. 



A peculiar phenomenon, not observed by Armagnat occurs when 

 the moh bdenuni- or copperelectrode is reduced to half a m.m. or 

 less. While the characteristic F W in the tiist quadrant changed but 

 little — only the slope ot the straight pai-t will become smaller, 

 hence the real resistance is increased — , the negative branch of the 

 characteristic in the 3'^ quadrant is straight and runs through the 

 horizontal axis. Only when the applied A'. JY. /'. exceeds the amount 

 of about 20 Volts a strong current may suddeidy be observed. The 

 high apparent resistance may then be restored again as soon as the 

 applied E. M. F. is decreased to the former value. The particular 

 phenomena of electrolysis occurring in a very sujall electiode in 

 consequence of strong current density evidently cause the dissymmetry 

 so well to be marked here. The accompanying colloidal oxide 

 JYOj (^14 points to a possible relation between these phenomena and 

 those taking place in the aluminium rectifier. Indeed, the characte- 

 ristic J of a rectifier, obtained by placing in a solution of atnmo- 

 niasulphate an aluminium wire of 1 m.m. thickness an J cm. length 

 opposite a large platinum electrode, agrees with the charactei-islic 

 iTlI. The rectifying power is attributed by Schulze and Taylok ') 

 to a thin film of oxygen fixed by a layer of aluminium-hydroxide. 

 ScHULZK also found the same behaviour for many other metals. In 

 the 3"^ quadrant the characteristic A remained horizontal to about 

 25 Volts. (Jsually the resistance suddeidy diminished with a stronger 

 E. M. F., a layer of aluminium hydroxide detaching from the point. 

 The rectifying power was restored again as soon as the applied 

 E.M.F. was diminished. Clarenck Greene") has shown that the 

 horizontal part of the characteristic of the aluminium rectifier is due 

 to an E. M. F. of polarisation counteracting the applied E. }f. F. 



Summarizing one may say that the characteristics of the electro- 

 lytic detectors generally have the shape of the curve F\ with two 

 parallel straight parts. It may happen that one branch is not fidly 

 developed. It is not necessary that the centre of the characteristic 

 occuis at the origin. 



^ 6. Characterbitics of crystal recti/iers. 

 Molybdenite-platinum. In considering tlie characteristic of a 



~^A. H. Taylor. Wiedemanns Annalen 30, pp. 984. 998, 1016. 

 2) Clarence Greene. Phys. Review, 2"^ Ser. Vol. Ill, 1914. 



81 

 Proceedings Royal Acad. Amsterdam Vol XXI. 



