ON GASEOUS COMBUSTION. 499 



gas (e.g., hydrogen or carbonic oxide), which thus becomes the ruling 

 factor. The behaviour of copper oxide towards mixtures of hydrogen 

 with excess of oxygen, and of nickel oxide towards mixtures of carbon 

 monoxide with excess of oxygen, have so far proved exceptional in this 

 respect, the observed rates in both cases being more nearly proportional 

 to the partial pressure of the oxygen than to that of the combustible gas. 

 In the case of copper oxide there is definite proof of the formation of a 

 condensed film of ' active ' oxygen at the surface, which actually bums 

 up the hydrogen before it can reach the still more active oxygen 

 chemically combined with the copper. 



The catalysing powers of all the metallic and non-reducible oxide 

 surfaces examined are highly stimulated by previous exposure to the 

 combustible gas, which is undoubtedly rendered ' active ' by association 

 with the surface. This stimulus is usually very durable, but in most 

 cases it is at once destroyed by a short exposure to oxygen. Although 

 as a general rule oxygen has per se no stimulating effect on a catalysing 

 surface, cases to the contrary have been encountered ; but even in these 

 exceptional instances the effect is neither so marked nor so durable as the 

 corresponding effects always observed in respect of the combustible gas. 



One notable feature with regard to the catalytic combustion of carbon 

 monoxide over a fireclay surface is the fact that the rate of combustion 

 of a mixture 2CO + 0, at 500° is about doubled by a thorough drying of 

 the gases the ' reaction constant ' increasing from about 0'09 to about 

 0'20. This remarkable result can hardly be explained on the supposition 

 that, in the case of the undried mixture (saturated at 18°), steam acts 

 merely as a diluent ; it apparently exercises a specific retarding influence 

 out of all proportion to its relative mass. 



Of. scarcely less interest are some recent results bearing upon the 

 effects of a hot fireclay surface (at 500°) upon the relative rates of com- 

 bustion of methane, hydrogen, and carbon monoxide. In the previous 

 section attention was directed to the fact that in ordinary explosive 

 combustion the affinities of methane and other hydrocarbons far exceed 

 those either of hydrogen or of carbon monoxide for oxygen. In contact 

 with a hot surface, however, the order is completely reversed, owing to 

 an apparently selective action of the surface in rendering the combustible 

 gases active. This circumstance is sufficient to invalidate the conclu- 

 sions of certain earlier investigators, notably those of Landolt, 1 concern- 

 ing the relative combustibilities of various gases, owing to their having 

 sucked off the products of partial combustion from the inner regions of 

 coal-gas flames through platinum tubes and the like. It may be taken 

 for granted that the introduction of a hot solid into a mixture of burning 

 gases is in itself sufficient to upset the regular conditions of explosive 

 combustion. 



The conclusions drawn by Bone and Wheeler 2 as to the catalytic 

 combustion of hydrogen derive collateral support from the recent re- 

 searches of Sabatier and Senderens on the remarkable powers of many 

 metallic surfaces (and especially nickel) of rendering this gas ' active ' 



Vber die cliemischen Vorg/inge in dcr Flamme des Leuchlgas. Habilitationsachrif f . 

 Breslau, 1856. 2 Loc. cit. 



