August 8, 1907] 



A'A TURE 



from Ihe burning layer of gases, so that the next layer is 

 put in the same state, and steady combustion proceeds. 



Phosphorescence has been spoken of as degraded com- 

 bustion, and, though literally the appellation is correct, I 

 think it is liable to be misunderstood, .^gain, it is often 

 supposed that phosphorescence is necessarily associated 

 with the formation of incompletely oxidised products. This 

 may be the case in a chemical system which is capable 

 of affording different products at different temperatures, 

 but it is not an essential feature ; the phosphorescent com- 

 bustion of sulphur, for example, affords nothing but sulphur 

 dioxide. 



Temperature of ignition is, then, neither a temperature 

 at wliich combination suddenly begins nor one dependent 

 soIeI\' on the nature of the combining gases. It will vary 

 with the proportion in which the gases are mixed and with 

 thrir pressure and other circumstances. Notwithstanding 

 the simplicity of this conception, it must be admitted that 

 there are many obscure facts connected with the ignition 

 of gases. The inflammability of gaseous mixtures is not 

 necessarily greatest when they are mixed in t:tc proportions 

 thecretically required for complete combination : the in- 

 fluence of foreign gases does not appear to follow any 

 simple law ; the presence of a very small quantity of a 

 foreign gas may exercise a profound influence on the igni- 

 tion temperature as in the case of tlie addition of ethylene 

 to hydrogen. When a mixture of methane and air is raised 

 to its ignition temperature, a sensible interval (about ten 

 seconds) elapses before inflammation occurs. These facts 

 are cognate to others which have increased upon us so 

 abundantly in connection with the influence of moisture on 

 chemical change. The study of the oxidation of phosphorus 

 in particular brings us among rocks and shoals. Apart 

 from the influence of moisture on the combination we 

 have the limitation of the process by a certain tension of 

 oxygen and by minute quantities of a vast number of 

 chemical substances, among which, in spite of much 

 labour, no other common bond can be found. We do not 

 know what oxide is initially formed in the oxidation, and 

 the existence of the oxides P,0 and P.O is as confidently 

 disputed as it is atTirmed. There is some reason for 

 believing that the phosphorescence connected with phos- 

 phorus succeeds the formation of one oxide and accom- 

 panies the formation of another. The state of the oxygen, 

 whether atomic, ionic, or molecular, which acts on phos- 

 phorus, the induced oxidation of other substances, the 

 ionisation of air accompanying the oxidation — these are all 

 matters concerning which there exists a bewildering litera- 

 ture that hangs over us like a cloud. The whole of my 

 .Address would, in fact, not suffice for a summary of 

 the state of our ignorance about the oxidation of phos- 

 phorus. The subject, simple as it appears at first sight, 

 is really involved with a vast number of unsolved chemical 

 problems the elucidation of which would throw much light 

 on chemical action in general. I may. perhaps, bequeath 

 the topic to some successor in this Chair as one which 

 may serve to illustrate the advance of knowledge since 

 these present days of darkness. 



The structure of flames has always been regarded as 

 denendent unon the chemical changes taking place in the 

 differentiated regions, but until recent times little attention 

 has been given to any question beyond the cause of the 

 bright luminosity of hydrocarbon flames. In a flame such 

 as that of hydrogen or carbon monoxide, where we have 

 '-ome reason to suppose that the same kind of chemical 

 transaction is taking place throughout the region of com- 

 bustion we should not expect to find a differentiation of 

 structure, and, as a matter of fact, we do not find any. 

 Erroneous ideas have gained currency from the use of 

 impure gases, and hydrortpn is still described as burning 

 with a pale blue flame, although Stas long ago stated that 

 if the aas is hiahly purified, and the air freed from dust, 

 the flame even in a dark-room can only be discovered by 

 feeling for it; a fact consistent with the line spectrum of 

 water lying wholly in the ultra-violet. The presence of a 

 very small quantity of free oxygen in carbon monoxide 

 destroys the perfect simplicity of the single shell of blue 

 flame with which the purified gas burns, and in other 

 flames small quantities of gaseous impurities or of atmo- 

 spheric dust give rise to features of structure and halos 

 which have been frequently supposed to pertain to the 



NO. 1 97 1, VOL. 76] 



flame of the combining gases. The fringe of a flame in 

 air may be often tinged by the presence of oxides of 

 nitrogen. 



No flame better illustrates the relation of structure to 

 chemical processes than that of cyanogen, where the two 

 steps in the oxidation of the carbon are clearly marked 

 out in colour. .-Apart from hydrocarbon flames, very fevir 

 others have been carefully explored from this point of view. 

 There is, unfortunately, no gas composed of two com- 

 bustible gaseous elements ; and, though such gases as the 

 hydrides of phosphorus and sulphur do not fall far short of 

 this, the experimental difficulties of an exact exploration 

 of their flames are very great. We are thus prevented 

 from studying the flame of a composite combustible in its 

 simplest form. 



The flames of hydrocarbons have naturally been the 

 subject of most frequent investigation. The use of single 

 hydrocarbons instead of the mixtures present in coal-gas 

 and other common combustibles has simplified the study 

 considerably. Two problems stand out prominently : one 

 is to trace the steps in the oxidation of the hydrocarbon, 

 the other to account for the bright patch of yellow 

 luminosity. With regard to the question of the luminosity, 

 I do not think there is any longer doubt about its being 

 due essentially to the separation within the flame of minute 

 solid particles of what is practically carbon. The separ- 

 ation seems to be adequately explained by the high tempera- 

 ture of the blue burning walls of the flame, which decom- 

 poses the unburned hydrocarbon within. In a similar way 

 arsenic and sulphur and phosphorus are liberated within 

 flames of their hydrides ; but these elements, being volatile, 

 do not appear as solids unless a cold object be placed 

 within the flame. In the case of the hydride of silicon the 

 liberated element at once oxidises to form the solid non- 

 volatile oxide, which gives a bright glow. 



The mode in which a hydrocarbon yields carbon by the 

 application of a high temperature has been the subject of 

 experiment and of hypothesis ; but neither the view of 

 Berthelot, that the carbon results from a continuous 

 coalescence of hydrocarbon molecules with elimination of 

 hydrogen, nor that of Lewes, according to which the form- 

 ation and sudden decomposition of acetylene is the essence 

 of the phenomenon, appears to me to be in harmony with 

 the experimental facts; and I am not aware that either 

 view has secured any support from other workers in this 

 field. It is certainly not easy to ascertain experimentally 

 the changes undergone by a single hydrocarbon as its 

 temperature is raised, and at the last it may be objected 

 that the course of events in contact with the solid walls 

 of a containing vessel is not necessarily the same as that 

 within the gaseous envelope of a flame. I am glad to 

 think that there is promise of further light on this subject 

 from the work of Prof. Bone. 



The course of oxidation of hydrocarbons has been the 

 subject of very careful and fruitful study. The old view 

 that a selective or preferential oxidation of the hydrogen 

 always took place, that with a restricted supply of oxygen 

 the hydrogen was oxidised and the carbon set free, is, 

 I think, no longer maintained by anyone who has studied 

 the question. The explosion of ethylene with its own 

 volume of oxygen, which leaves us with practically all the 

 carbon oxidised and all the hydrogen free, is fatal to this 

 view. Again, when hydrocarbons are burned in a flame 

 with a restricted supply of air, as is the case in the inner 

 cone of the flame of a well-aerated Bunsen burner, there 

 is clearly no separation of solid carbon, and the products 

 of combustion when withdrawn and analysed disclose the 

 presence of much free hydrogen and no unoxidised carbon. 

 In describing this experimental fact I have spoken of it 

 as the preferential oxidation of carbon. I have always 

 thought it pedantic to quarrel with that expression; for, 

 in speaking of a chemical transaction, we usually include 

 only a description of the initial and final states of com- 

 bination. I should be sorry, however, to detach the ex- 

 pression from the facts it describes and to exalt it into 

 a general doctrine. That would be quite inadmissible, and, 

 if there is any danger of misunderstanding, it would be 

 better to avoid using the expression. 



The admirable researches carried out in the University 

 of Manchester by Prof. Bone and his collaborators have 

 afforded most valuable information as to the oxidation of 



