1886.] on Dissociation Temperatures, &c. ^11 



There is one other circumstance connected with dissociation, proved 

 by experiment, which however, requires explanation. It is considered 

 as a sure sign that dissociation is going on when a flame whose tem- 

 perature is raised becomes longer ; this it is said can only be 

 accounted for by dissociation taking place. I agree with this con- 

 clusion, but the experiments by which it has been proved have been 

 made, like others referred to, in narrow tubes or passages in which 

 the dissociating action of the heated surfaces must come into play. 

 It is not alone the heat to which the gases are raised that in these 

 cases causes dissociation and increases the length of the flame, but 

 also the influence of the heated surfaces in contact with the com- 

 bustible gases, more especially if these gases contain hydrocarbons. 

 The extension of the flame is also partly due to the obstruction which 

 the surfaces ofi'er to the recombustion of the dissociated gases through 

 want of space. If the same flame be allowed free development in a 

 space unencumbered by surfaces, as in my radiation furnace, no such 

 extension of its length would be observed; but, on the contrary, it 

 would get shorter with increase of temperature. This action can be 

 best observed in a regenerative gas-burner whose flame is shorter the 

 greater the intensity of the temperature, and therefore of the light 

 produced. On the other hand, flame may be extended almost to any 

 length if conducted through narrow passages ; this may be seen in 

 regenerative furnaces which will send the flame to the top of the 

 chimney if the reversing valves are so arranged that the flame, instead 

 of passing through the furnace chamber, is made to burn directly 

 down into the regenerators. No proper combustion can then take 

 place in the brick checkerwork of the regenerative chambers, and the 

 flame will consequently continue to extend until cooled down below a 

 red heat, being ultimately converted into dark smoke ; thus in this 

 case, the extensive surfaces offered by regenerators will act both ways, 

 by preventing combustion, and by assisting dissociation. 



It will now be understood that regenerative furnaces themselves 

 offer special opportunities for making experiments, most questions 

 being best settled by the results obtained in actual work. If dissocia- 

 tion sets in we see the consequences in want of heat, reduced output, 

 and in destruction of furnace and material. If the causes of dissocia- 

 tion are removed we immediately become aware of the circumstance 

 by a rise in temperature, increased output, longer furnace life, and 

 saving of material. 



Similar results may be obtained' with other furnaces, but the 

 beneficial action will not be so great as in the case of the regene- 

 rative furnace, because the intensity of heat obtainable in them is 

 much lower. 



In applying the principle of heating by radiation, or free develop- 

 ment of flame, to boilers, it is necessary to prevent the flame in its 

 active stage of combustion from touching either the sides of the boiler 

 or its brickwork setting. The flame is allowed free space to burn in, 

 and thus good combustion is obtained, after which the jjroducts of 



