COAL: ITS COMPOSITION 121 



be observed when properly lined with brick, and with 

 suitable direction of flow and air mixture, the whole will 

 be illuminated. Streaks and splashes of dark gas will be 

 seen coming forward over the fire, and these melt away as 

 they travel, and burn and help to keep up the temperature. 

 The dark streaks are simply gas not hot enough to give 

 violet light. They are red or yellow flames of burning gas 

 ready to produce smoke if sent upon cold surfaces. Kept 

 off cold boiler plates, they complete their high temperature 

 combinations, and may then be used for heating anything. 



It is not that blue glass marks the state of combustion 

 beyond which one must pass, but it seems certain that if a 

 properly mixed gas attains this temperature before ex- 

 posure to cold surfaces, it will be properly burned. It 

 would be interesting to experiment with red, yellow, and 

 green glass, so as to find how these help in analyzing the 

 state of a fire. It is certain that if blue glass cuts the 

 flame very short there is imperfect combustion. 



Now I have not told you much about coal, for I know 

 nothing myself of the way it is put together. All I can 

 infer is that a very small amount of combined hydrogen 

 will change the physical nature of much carbon. Analysis 

 of coal seems to point to the presence of oxygen as the 

 potent cause of so-called bituminosity. Knowledge of the 

 phenomena of heat such as latency teaches that the fuel 

 bed must be chilled when fresh coal is giving off vapor. 



On the supposed atomic arrangement of hydrocarbon, 

 speculation may be indulged in on the facts that hydro- 

 carbon is first attacked by the oxygen, and that the carbon 

 is set free by itself or in some different combination with 

 hydrogen, and so readily condenses on the first cold surface. 

 And so it is learned to mix atoms of oxygen in excess of 

 what the hydrogen atoms will snatch up and to maintain 

 everything hot until the carbon has had its chance to find 

 its own atoms of oxygen. And as it may be inferred that 

 a thick fuel bed implies shortness of oxygen above the 

 fire for the fire has perhaps been converted into a gas 



