BACTERIOLOGY OF SEWAGE 165 



doubt that the albuminous material present is being split up by 

 ordinary putrefactive bacteria. In the mains and where open 

 systems of purification are at work, aerobic forms play the chief 

 part, while in the closed methods anaerobic organisms are those 

 chiefly concerned. In contact and percolating systems there is 

 evidence that at first the purifying action of bacteria is materially 

 furthered by physical processes. Thus Dunbar has shown that 

 when such a substance as coke is used in a sewage filter-bed a 

 considerable amount of the albuminous material is removed in 

 a very few minutes by adsorption, for, albumin, being of a 

 colloidal nature, is readily deposited under such circumstances 

 in the pores of the coke in the form of films. After a time such 

 a filter-bed becomes clogged, but on access of oxygen being 

 allowed, it regains its adsorptive properties probably from the 

 oxidation of the material adsorbed. 



During this stage, as in the whole purification process, four, 

 and it may be five, processes are at work : First, the action of 

 ordinary bacteria splitting up the higher albuminous molecules ; 

 secondly, the action of nitrifying bacteria building up nitrites 

 and nitrates from ammoniacal products ; thirdly, the action of 

 denitrifying bacteria which reduce nitrates to lower gaseous 

 oxides and to free nitrogen (the presence of which in filter beds 

 can be demonstrated) ; fourthly, the action of higher forms of 

 vegetable and animal life ; fifthly, it is possible that direct 

 chemical oxidation of the earlier products of bacterial action 

 may occur, and in any case the access of an abundant oxygen 

 supply to adsorbed material hastens its destruction. It is 

 possible, as is indicated by the work of Lorrain Smith and of 

 Mair, that perhaps too little weight has been attached to the 

 parts played by the two last processes specified, for in the later 

 .stages of the purification process there is a very marked 

 diminution in the number of bacteria present in the filter. 

 Much further work, however, is necessary before the part to be 

 assigned to each factor in operation can be properly estimated. 



Further, the details of the essentially bacterial part of the 

 process are obscure, and the relative parts played, even in an 

 open purification process, by aerobes on the one hand, and 

 anaerobes on the other, is little understood. When sewage is 

 drained off to rest a filter-bed, great quantities of oxygen are 

 sucked in, but as to how long the bed thus remains aerated, 

 authorities differ some maintaining that oxidation processes per- 

 sist even after the bed has been recharged, while others state that 

 the oxygen in the resting bed is consumed, and its place 

 by carbon dioxide and nitrogen. Certainly, at certain 



