SYMPOSIUM ON SCIENCE AND RECONSTRUCTION 85 



a dense clond of yellow or brown smoke. After the fuel 

 bed and eheckerwork were well heated, little more trouble 

 from this source would be encountered. A remedy which 

 was deyeloped at Streator worked out quite successfully, 

 was to make a fresh charge of fuel just before shutting 

 down the set and making only one or two runs from it. 

 During the succeeding lay -oyer period this coal coked 

 gradually and by the time the set was again put into 

 operation it had become completely coked. Then upon 

 blasting the machine could be brought rapidly up to 

 temperature, and after one or two runs had been made, the 

 addition of fresh fuel produced yery little if any smoke, 

 since the tarry yapors could be ignited and completely con- 

 sumed. 



As has been preyiously explained, the blow and run 

 periods are so timed as to bring all three chambers to the 

 proper operating temperatures at the same time. Since 

 the caking property of the coal tends to make the fuel bed 

 more resistant to the passage of air than coke fuel, it is 

 eyident that with a giyen air-blast pressure ayailable, it 

 will take a longer blowing time to get the coal fuel to the 

 proper temperature. Howeyer, since there is considerable 

 rich yolatile matter giyen off from coal, this upon burning 

 in the carburetor and superheater tends to oyerheat these 

 chambers before the generator is hot enough. Since the 

 temperature most fayorable for i)roperly fixing the gas oil 

 is quite well defined, excess temperature tends to break up 

 the oil too much with the formation of lamp-black which 

 rapidly fouls the checker-work and chokes the connections. 

 If a great part of the combustible gas is burned outside the 

 machine at the stack, the latter becomes oyerheated, and 

 at any rate fuel is wasted. Seyeral methods for remedying 

 this condition haye been tried, but practically all of them 

 result in a great deal smaller production of gas in a giyen 

 time than is possible with coke, the reduction amounting 

 to as much as 30 per cent. A method was worked out at 

 Streator which obviates this difficulty and at the same time 

 increases capacity, and this method will be briefly de- 

 scribed. 



The gas produced during the blow period becomes 

 momentarily greater in heating yalue as the blow proceeds, 

 due to the increasing percentage of carbon monoxide. At 

 the end of a three-minute blow, for example, it may haye a 



