AbstractExperiments carried out at B.C.U.R.A. in 1939‐40 in connexion with a programme of work on producers for road transport threw doubt on the conventional “two‐zone” theory of producer beds taught in the text‐books. In particular, an overwhelming proportion of the fuel was consumed within the first few particle thicknesses. Temperature measurements and gas analyses of samples taken at a series of points throughout the bed gave evidence of an apparentincreasein the total energy of the gases (per unit quantity of ingoing air) on passage through the bed. The only possible explanations are: (a) systematic errors in temperature measurements, (b) systematic errors in gas analyses such as would arise if energy in some latent form were yielded up to the sampling probes. Check measurements with suction pyrometers rendered (a) highly improbable. The most probable explanation was considered to be that carbon monoxide produced as a primary combustion product of the surface reaction between carbon and oxygen bums in the voids where it reacts with the surviving oxygen. Experiments in which air was streamed through a carbon tube at 850° showed that the oxygen removal and nearly the whole of the carbon consumption took place within a path‐length roughly equal to a single tube diameter. This length being quite insufficient for all the oxygen to reach the carbon surface (even with a certain degree of turbulence), this experiment is taken as providing substantial evidence as to the truth of the above working hypothesis. It appears, moreover, that mere thermal quenching is insufficient to “freeze” the reaction between CO and surviving oxygen. This would be at once intelligible if the reaction followed a “chain” mechanism as it doesin vitro.Attempts were therefore made to inhibit the gas phase reaction by introducing small quantities of volatile chlorine‐containing materials with the ingoing air, which was carefully freed from moisture. Only minimal quantities of CO2, now appeared in the outgoing gases, and the oxygen survival and carbon consumption were now appreciable after path lengths of about 10 tube‐diameters. Finally, it was found possible to muse partial inhibition of CO2, formation by the use of solid inhibitors (clean fused silica) remote from the carbon surface. These results were obtained at temperatures such that no appreciable reduction of CO2, to CO would take place.The above findings carry important implications in regard to technical fuel beds, which begin to function 85 producers after a path length equivalent to a single particle diameter has been traversed by the ingoing air. In particular it is ordinarily necessary to run with excess air since the physical mixing of the broad stream of producer gas with secondary air is otherwise incomplete. A burner which climates this disadvantage and gives combustion gas “by first intention” has been designed and is in the proc