54

and the yellow colour given by those ofbarium, and the green by those of boron,may depend upon a temporary productionof these bases by the inflammable matterof the flame.

Whenever a flame is remarkably bril-liant and dense, it may be always con-cluded that some solid matter is producedin it: on the contrary, when a flame isextremely feeble and transparent, it maybe inferred that no solid matter is formed.Thus none of the volatile combinations ofsulphur burn with a flame in the slightestdegree opaque ; and, consequently, there isno reason, from the phenomena of its flame,to suspect the existence of any fixed basisin sulphur.

5thly. These views will probably offerillustrations of electrical light. The voltaicarc of flame from the great battery, differs incolour and intensity according as the sub-stances employed in the circuit are different;and is infinitely more brilliant and dense withcharcoal than with any other substance.May not this depend upon particles of thesubstances separated by the electrical at-tractions ? and the particles of charcoalbeing the lightest amongst solid bodies, (as