--a pair of meters geared together, one for the gas, the other for the air--and this would introduce extra complexity and extra expense. On the whole the idea cannot be recommended, and the action of the British Home Office in prohibiting the use of all such mixtures except those unavoidably produced in otherwise good generators, or in burners of the ordinary injector type, is perfectly justifiable. The derivation and effect of the other gaseous and liquid generator impurities in acetylene were described in Chapter II. Besides these, very hot gas has been found to contain notable amounts of hydrogen and carbon monoxide, both of which burn with non-luminous flames. The most plausible explanation of their origin has been given by Lewes, who suggests that they may be formed by the action of water-vapour upon very hot carbide or upon carbon separated therefrom as the result of previous dissociation among the gases present; the steam and the carbon reacting together at a temperature of 500° C. or thereabouts in a manner resembling that of the production of water-gas. The last generator impurity is lime dust, which is calcium oxide or hydroxide carried forward by the stream of gas in a state of extremely fine subdivision, and is liable to be produced whenever water acts rapidly upon an excess of calcium carbide. This lime occasionally appears in the alternative form of a froth in the pipes leading directly from the generating chamber; for some types of carbide-to-water apparatus, decomposing certain kinds of carbide, foam persistently when the liquid in them becomes saturated with lime, and this foam or froth is remarkably difficult to break up.