LUEDEKING — CHEMISTRY OF COISIBUSTIOIS . 373 



Considering the ease with which the Cyanogen Titanium reac- 

 tion can be obtained in the same flames in a few minutes' time, 

 it may seem strange that the prussian-blue reactions are not 

 stronger. However, the conditions of experiment are of neces- 

 sity very unfavorable. As it is, only a part of tlie Cyanogen or 

 Hydrocyanie acid is absorbed, the greater quantity being carried 

 away by the current of Carbon dioxide. 



The best way of experimenting is to cool the test tube with ice 

 and thus condense most of the aqueous vapor of the gases; much 

 of the Cyanogen can thus be retained. 



By this mode of experiment a prussian-blue reaction could be 

 obtained in the combustion of coal gas, various candles, and 

 alcohol. 



It is necessary to guard carefully against ammonia in the atmo- 

 sphere, as otherwise we have an inducing agent for the formation 

 of cyanogen. 



We proved then, firstl}', that in the combustion of ordinary 

 fuels Cyanogen occurs in the reducing flame, being generated by 

 the process of combustion. 



The final answer to our proposition can, however, not yet be 

 given ; for, is it not possible that the combustibles referred to 

 above contain Nitrogen ? If so, we have an inducing agent ; for 

 by our mode of experiment both nitro and amido compounds 

 mixed with alcohol were found to produce large amounts of Cya- 

 nogen in their combustion, and consequently it need not be atmo- 

 spheric nitrogen that causes the formation of Cyanogen or 

 Hydrocyanic aci 1 in our experiments. 



Coal gas we know contains nitrogenous compounds, and of 

 tallow we should also expect the same, as well as of alcohol, for 

 in alcoholic fermentation volatile nitrogenous bases are formed. 

 We should least expect paraffin to contain Nitrogen, but cannot 

 positively say. 



It was necessary, therefore, to prepare a combustible entirely 

 and certainly free from all nitrogenous matter. 



Marsh gas is such a combustible. I prepared 50 litres of this 

 substance out of pure acetate of soda, and lime and soda, previ- 

 ously heated. This Marsh gas was passed from a copper gas- 

 ometer through a Buiisen burner placed under a bell-jar, and the 

 experiment conducted as in previous cases. The air for combus- 



