86 



failure, previously stating the circumstances of the combustion 

 of pitcoal which differs materially from that of coke or charcoal. 

 When a furnace is charged with this fuel, it must be apparent 

 that the temperature is nuicli inferior to that which would be 

 produced, if the combustible were free from gaseous and vaporisa- 

 ble matter, these in assuming the elastic state absorb an immense 

 quantity of caloric which is in general completely lost ; besides 

 the coal fuses and agglutinates so that the afflux of air is not 

 as free as in the former case, and the ujjper strata of fuel cannot 

 burn fi'om the want of oxygen. We may therefore consider them 

 as exposed to distillation, so that the tar, gas, &c. Avhich proceed 

 from them make their escape throxigh the chimney unaltered : and 

 may conclude that in general coal does not produce more heat 

 than it would have done if previously coked or reduced to |ds of 

 its weight ,• it cannot even afford so much. But the volatile pro- 

 ducts are highly combustible, there is no reason why they should 

 not burn if oxygen be present, and if a stream of air be mixed 

 with the smoke and the mixture ignited, experience shews that 

 it is totally consumed. The first trial was made by Mr. Watt, it 

 was not nuich different from the stove of Dalesme which is men- 

 tioned in one of the first vohmies of the Philosophical Transac- 

 tions, and may be described in a few words as an inverted furnace, 

 the chimney proceeds from the ash-pit and the draught is down- 

 wards through the fuel. Let us suppose the fireplace to be nearly 

 filled with ignited coke and that a quantity of raw coal is laid on 

 the top, from the impression of the heat it will be in part vola- 

 tilised, and the smoke must be drawn through the fuel along with 

 tlie air which maintains the combustion, it will therefore be con- 



