V. ARTIFICIAL SELF-PURIFICATION OF SEWAGE. 1. GENERAL OBSERVATIONS. [Sidenote: Enumeration of more important experiments.] A great many experiments have been made during the last ten years with artificial processes for the self-purification of sewage, and amongst the more important the following may be mentioned: London experiments. Sutton „ Exeter „ Manchester „ Leeds „ Sheffield „ Leicester „ York „ Hamburg „ [Sidenote: Experiments have not been conducted on uniform lines.] A casual observer might, therefore, consider himself justified in thinking that all these experiments had added a great deal to our knowledge of the intricate changes taking place in these processes, but such a conclusion would not be justified in reality. For beyond settling questions of local importance by chemical analysis, the experiments, owing to a variety of causes, have not materially enhanced the stores of our information, indeed not unfrequently the results obtained are apparently contradictory and bewildering. An experiment must be looked upon as a question addressed to nature, and the answer will depend on the way the question has been put. If this way differs in every case it must be clear that the answer, too, will differ in every case, and it is this absence of uniformity which greatly reduces the general value of these experiments. These remarks must not be misunderstood to convey the impression as if the experiments had not been conducted with care and skill! Far from it! Some of them have been made with the greatest skill and care and with the very evident desire to arrive at correct conclusions, and it is only when they are placed side by side with other experiments, with a view to deducing from them general conclusions concerning the processes at work, that great difficulties are experienced. The result of each experiment is governed by a large number of factors, which by slightly different manipulations may attain in this ever-fluctuating process different weights, so that the results may be contradictory, and it is only by arranging these factors on a common basis, as it were, and by addressing the questions to nature in the same systematic and uniform way, that good general results may be expected. It is well known, for instance, that in some cases septic tanks have not given good results, whilst in others they have worked very well; again, continuous filtration has failed in some experiments, whilst in others, notably in the York experiments, it has given good results. If, therefore, in future the mistake of the past is to be avoided, it will be necessary to settle on a common line of action in all experiments. [Sidenote: Attempt to evolve general theory.]