The use of earthenware as a filtering material for the purification of water is probably of very ancient origin, but the modern modification, the porcelain filter, according to Gautier (1890), is due to Nadaud de Buffon (1861). Earthenware filters in general serve the useful purpose of aërating the water and retaining suspended matter and are therefore valuable in rendering boiled and rain water clear and palatable. The first application of bacteriological methods of examination by pupils of Robert Koch showed, however, that the great majority of these and other water filters gave no protection against disease organisms. Hesse (1885–6) found that clays and asbestos retained micro-organisms to a marked extent, and, although Plagge (1886) confirmed this observation he demonstrated that but few filters prevented the direct transmission of bacteria and that even the best after a few days contaminated the filtrates indirectly. This indirect contamination Plagge attributed to the growth of micro-organisms within the filter mass and so pronounced is its influence that porcelain filters after being a few days in action give filtrates richer in germs than the original water. Schäfer (1893) states that porcelain filters do not allow organisms to grow through them when the water filtered is of such a character that the organisms cannot multiply in it. He points out that the waters usually filtered for domestic purposes do not allow B. typhosus and V. cholerae to grow, and that in such cases porcelain filters are efficient. Unfortunately, however, drinking waters do occur in which these organisms can multiply, and then indirect contamination is inevitable if the filter be allowed to act more than a few days without sterilisation.