growths of carbon soon appear, but do not obstruct the orifices during 100 hours' use; if, however, the gas-supply is checked till the flame becomes thick, the growths appear more quickly, and become obstructive after some 60 hours' burning. On the assumption that acetylene begins to polymerise at a temperature of 100° C., Gaud calculates that polymerisation cannot cause blocking of the burners unless the speed of the passing gas is so far reduced that the burner is only delivering one- sixth of its proper volume. But during 1902 Javal demonstrated that on heating in a gas-flame one arm of a twin, non-injector burner which had been and still was behaving quite satisfactorily with highly purified acetylene, growths were formed at the jet of that arm almost instantaneously. There is thus little doubt that the principal cause of this phenomenon is the partial dissociation of the acetylene (i.e., decomposition into its elements) as it passes through the burner itself; and the extent of such dissociation will depend, not at all upon the purity of the gas, but upon the temperature of the burner, upon the readiness with which the heat of the burner is communicated to the gas, and upon the speed at which the acetylene travels through the burner. Some experiments reported by R. Granjon and P. Mauricheau-Beaupré in 1906 indicate, however, that phosphine in the gas is the primary cause of the growths upon non-injector burners. According to these investigators the combustion of the phosphine causes a deposit at the burner orifices of phosphoric acid, which is raised by the flame to a temperature higher than that of the burner. This hot deposit then decomposes some acetylene, and the carbon deposited therefrom is rendered incombustible by the phosphoric acid which continues to be produced from the combustion of the phosphine in the gas. The incombustible deposit of carbon and phosphoric acid thus produced ultimately chokes the burner.