The opinion has recently been expressed1 that the poisonous action of ordinary coal-gas and carburetted water-gas is probably in part due to the “Illuminant” hydro-carbons, of which ethylene is the chief, and not simply to carbonic oxide. In consequence of the doubt existing on this point one of us was asked by the recent Departmental Committee of the Home Office on Water-Gas to investigate the matter, and the results of the experiments which we were then able to make appeared in the Committee's Report2. In the main series of observations the animal was placed in a respiration chamber through which a current of air was passing at a known rate. With the current, before it entered the chamber, a known percentage of coal-gas or carburetted water-gas was mixed. It was found that whether ordinary coal-gas or carburetted water-gas was used the symptoms observed were those of carbonic oxide poisoning, and corresponded exactly to the percentages of carbonic oxide present. We also found that Benzene, which is one of the “illuminants,” is present in proportions far too small to contribute to the toxic effects of coal-gas or carburetted water-gas. Finally, we endeavoured to investigate separately the action of ethylene. The ethylene we then used was prepared in the ordinary way from sulphuric acid and alcohol, and when about 10% of the gas was mixed with air and supplied to an animal very distinct toxic symptoms were produced.