Thomson's theory or combustion.

371

combine with potash, or with lime, in a crucible or glass tube,and likewise when melted phosphorus is made to combine withlime heated to redness, (page 165). He supposes that in allprobability barytes and strontia exhibit the same phenomenonwhen combined with melted sulphur or phosphorus; and per-haps some of the metals when combined with phosphorus.

The phenomena Dr. Thomson explains thus: The sulphur andphosphorus are in the melted state, and therefore contain caloricas an ingredient; the alcalies, earths and metals which producethe phenomenon in question, contain light as an essential ingre-dient. The sulphur or phosphorus combines with the base of themetal, earth or alcali; while at the same time the caloric towhich the sulphur or phosphorus owed its fluidity, combineswith the light of the metal, earth or alcali; and the compoundflies off under the form of fire.

Thus the process is exactly the same with combustion, except-ing as far as regards the product. The melted sulphur or phos-phorus acts the part of the supporter, while the metal, earth oralcali occupies the place of the combustible. The first furnishescaloric, the second light, while the base of each combines toge-ther. Hence we see that the base of sulphurets and phosphuretsresembles the base of products in being destitute of light, theformation of these bodies exhibiting the separation of fire likecombustion, but the product differing from a product of combus-tion in being destitute of oxigen, Dr. Thomson distinguishes theprocess by the title of semi-combustion ; indicating by the term,that it possesses one half of the characteristic marks of combus-tion, but is destitute of the other half.

The only part of this theory which requires proof is, that lightis a component part of the earths and alcalies. But as potashand lime are the only bodies of that nature, which we are certainto be capable of exhibiting the phenomena of semi-combustion,the proofs must of necessity be confined to them. That limecontains light as a component part has been long known, Meyer

A a 2