730 PROCEEDINGS OF THE AMERICAN ACADEMY. 



If we plot the logarithms of these last numbers on the one hand and the 

 corresponding temperatures on the other, the resulting points lie almost 

 precisely on a straight line whose slant is given by the equation 



^|ii: = 0.0185, 

 or by the integrated form 



logio-- 0.0185 (71- T:). 



It is customary to express the influence of temperature on reaction ve- 

 locity by means of the quotient of the velocity at one temperature divided 

 by the velocity at a temperature ten degrees lower. This quotient can 

 be easily found in the present case from the above equation. Since 

 T,-T, = 10, 



^ = log-' 0.185 = 1.53. 



In other words, when the temperature rises 10° the velocity increases by 

 about fifty per cent. This increase is less than that of most reactions, 

 but corroborates the idea of van't HofF,* who writes concerning this 

 quotient: — 



" Bedeutend weniger als Verdoppelung zeigt nur die Zerlegung von 

 Phosphor- und ArsenwasserstofF. Das hangt aber vielleicht mit der 

 hohen Beobachtungstemperatur zusammen, denn in der grossen Mehrheit 

 der bis jetzt beobachteten Falle nimmt das Geschwindigkeitsverhaltnis 

 fiir 10° mit stiegender Temperatur ab." 



The Nature of the Catalytic Action. 

 At first sight the explanation of the catalytic action of the silver would 

 seem to be like that which is sometimes offered for the development of a 

 photographic plate ; namely, that the silver as it is produced becomes 

 supersaturated, and finally reaches such a concentration as to stop the 

 reaction except where silver is already present upon which it may de- 

 posit. If this explanation is correct the phenomenon we are studying 

 resembles closely the simple process of crystallization from a supersatu- 

 rated solution. In fact, I may point out that every case of such crystal- 

 lization is a case of autocatalysis, and that here also the reaction velocity 



* Vorlesungen ueber Tlieor. u. I'liys. Clietn. 



