of the Spontaneous Combustion of Charcoal. 91 



of air and moisture absorbed ; and this should, according to 

 the explanation suggested, occur, the alkali formed being much 

 heavier than its metallic base. It appears that, to produce the 

 ignition, the charcoal should not only be reduced to powder 

 soon after its formation, but that the sooner it is so reduced 

 the more certain and considerable will be the effect. Now this 

 fact also is entirely consistent with the explanation ; because 

 when the pulverization has been delayed, air and moisture 

 will have gradually produced the alkali, by a process imper- 

 ceptible because the minute portions of potassium would be 

 at comparatively distant intervals from each other, and thus 

 would not be in sufficient quantity at any one place to pro- 

 duce a sensible effect. 



Colonel Aubert pulverized a mixture of charcoal and sul- 

 phur; and he found that, under these circumstances, no igni- 

 tion ever occurred. The reason is obvious ; for the potassium, 

 which has been conceived to be the cause of the combustion, 

 entered, during the trituration, into combination with the sul- 

 phur. 



He also triturated charcoal with nitre; and he again found 

 that the spontaneous combustion was prevented. Now nitre, 

 by mingling with the potassium, would check its too rapid ab- 

 sorption of oxygen ; and the effect of his experiment is in this 

 way sufficiently accounted for. 



The presence of potassium seems to account for the circum- 

 stance, that when charcoal is moistened and subjected to heat, 

 carbu retted hydrogen is set at liberty. In this instance it 

 would appear that the water is decomposed, the hydrogen 

 evolved, and the oxygen united with the potassium to form 

 the alkali. If the heat be continued, carbonic oxide would be 

 evolved ; the oxygen absorbed in the first part of the opera- 

 tion being again detached from the metallic base. Now this 

 explanation corresponds precisely, I believe, with the order 

 in which, in such an experiment, these gases are produced. 



All the circumstances observed by Mr. Hadfield and Col. 

 Aubert appear therefore perfectly reconcilable with the sup- 

 position, that the spontaneous incandescence is owing entirely 

 to the oxidation of the potassium liberated from the wood 

 during the manufacture of the charcoal. 



Dr. Thomson in the second volume of his History of Che- 

 mistry, published since my paper was read to our Society, has 

 thrown additional plausibility upon my explanation, by his at- 

 tempt to show that pyrophorus owes its property of catching 

 fire, when in contact with oxygen, to a little potassium which 

 is reduced to the metallic state during the formation of the 

 pyrophorus. 



N2 



