1886.] on Dissociation Temyeratures, dx. 475 



than are reached in the arts, as materials used in furnace-building 

 would not withstand such temperatures for any length of time ; but 

 still I must call attention to the circumstance that if the influence of 

 the inner surfaces of the tubes on the combustion of the gases therein 

 could be removed, the dissociation temperatures arrived at would be 

 found still higher. I cannot admit that Bunsen's explanation of the 

 cause of the second and third explosions is quite satisfactory, as it 

 is not the cooling of the gases alone which renders the subsequent 

 explosions possible, but also the thorough re-mixture of the gases by 

 diffusion after each explosion. This I will illustrate by means of 

 the diagrams exhibited, Figs. 1 to 6, which represent: — 



1. A tube filled with an explosive gas mixture which is shown white. 



2. The same tube immediately after an explosion has taken place, 

 the white margin indicating the unexploded mixture close to the 

 sides, and the deep-red, towards the middle of the tube, the exploded 

 gases. The white is shown as merging into deep-red by degrees, 

 because close up to the sides the surfaces prevent explosion or 

 combustion altogether; nearer the middle partial combustion takes 

 place, whilst only in the middle of the tube the gases find sufficient 

 space for complete combination. 



3. The same tube after the burnt and unburnt gases have mixed 

 by means of diffusion, which is coloured light-red. 



4. The same tube immediately after the second explosion, coloured 

 light-red at the sides, turning into deep-red by degrees towards the 

 middle. 



5. The same tube after diffusion has done its work a second time, 

 coloured a deeper shade of red. 



6. The same tube after the third explosion, coloured nearly 

 deep-red throughout, but still a lighter shade on the sides. 



In Bunsen's mode of determining dissociation at high tem- 

 peratures we have only to deal with the obstruction which surfaces 

 offer to combustion, leaving out their dissociating influence at high 

 temperatures which affect most of Deville's results. For that reason 

 Bunsen arrives at much higher dissociation temperatures than Deville, 

 and his mode of experimenting possesses the advantage that it may 

 lead to a proper settlement of the question of temperatures at which 

 dissociation would set in when taking place in a space unencumbered 

 by surfaces. 



I should wish some one more experienced than I am with purely 

 physical investigation to make the following experiment : — 



Take a narrow tube of about the same size as Bunsen used for his 

 experiments, and a hollow sphere of the same capacity, in both of 

 which Bunsen's experiment should be repeated. The sphere offering 

 less surface than the tube in proportion to the quantity of gas it 

 contains, the dissociation temperature should be found higher in the 

 former than in the latter, if my views are correct. The results that 

 would, in my opinion, be obtained are shown approximately by the 

 red and white coloured surfaces in the diagram (Figs. 7 to 10). 



