The negative influence of dissolved oxygen in cold nuclear fusion tests is discussed with regard to the formation of a LiD membrane on the surface of palladium cathodes.