27 



removing substances which are not otherwise removed, and which, if 

 allowed to be burnt with the gas, lessen its illuminating power. These 

 light- damaging compounds are produced during the later portion of 

 the distillatory process, as I have proved by experiment. The same 

 retort was charged twice with the same weight of the same coal. 

 The gas produced by one charge was purified by lime only, that pro- 

 duced by the other charge was purified by lime and clay. The 

 illuminating power of the gas passing at each half hour's end was 

 determined, and it was found that the purification made no difference 

 for the first three or four half-hours. About the middle of the 

 charge, that purified by my process had slightly the advantage, and 

 at the close the difference in favour of that purified by the addition 

 of clay has been found as much as ten or twelve per cent. Thus it 

 is shown that the compounds removed by clay from gas produced 

 during the early stages of distillation however objectionable on 

 other accounts do not lessen the light-giving power of gas, whereas 

 those removed during the later periods of distillation reduce the light- 

 giving powers considerably. 



If conjecture be allowable, I would venture an opinion that cya- 

 nogen compounds, and other nitrogenized substances with which foul 

 clay abounds, are those which lessen light. My own investigations 

 lead directly to this inference, and, I think, explain an old Table by 

 Dr. Henry in this sense. In the ' Philosophical Transactions ' for 

 1808, he shows that the gas produced from 112 Ibs. of Cannel-coal 

 contained, after purification, the following quantities of nitrogen : 



Without assuming the absolute accuracy of these figures, we may 

 regard them as valuable indicators, pointing, I think, in the direction 

 I have ventured to conjecture. 



A beautiful reaction furnishes experimental proof of the damage 



