Two main general approaches to water quality improvement exist: treatment of pollution at its source or occasionally after it has entered a stream, and augmentation of the stream's flow to help it assimilate loads of waste beyond its natural capacity. A third possibility in certain situations is the diversion of wastes out of a stream's drainage entirely. In practice, these methods can be varied and combined in any number of ways to fit a need. To take the last one first, diversion of whole wastes as received from their sources is a total and dramatic means of coping with a pollution problem stemming from collectable wastes, but it often has disadvantages. One of these, of course, is the possibility that the pollution problem may be simply transplanted elsewhere--that the water in which the wastes eventually end up will suffer. Another is loss of water from the stream system. If, as is usual, a town gets its water out of the local river or a tributary and does not give it back after use--preferably well cleaned up--other users downstream are not going to have as much water available to them, and the essential processes and ecology of the river itself may suffer. The only place such wholesale diversion of wastes has been seriously considered in the Potomac Basin is at metropolitan Washington, whose sewage could feasibly be piped across Chesapeake Bay and the Delmarva peninsula and well out into the Atlantic--possibly, as has been suggested, in combination with sewage from Baltimore. It would be a permanent means of disposal, but very expensive in terms of both investment and operating costs. Furthermore, though in the estuary no downstream users would suffer a loss of water supply, the water content in metropolitan sewage has at times risen as high as 80 percent of the flow of the river above the upstream intakes. The effects of such a subtraction of fresh water on the estuary itself--changes in flow, and in the penetration of salt water upriver, with an inevitable alteration in valuable fisheries and the whole balance of aquatic life established through millennia--could easily turn out to be disastrous. Standard treatment of pollution at its source consists of the primary and secondary processes we have glanced at, sometimes adjusted to specific industrial wastes. It has to be brought up to peak efficiency along the Potomac, for this is a "known factor" of great significance. Plants can and must be improved physically where necessary, and qualified operators provided for them. Collection systems have to be improved or enlarged in many places. Diminutive plants, doomed to inefficiency by their size and the financial impossibility of hiring expert workers for them, need to be eliminated in favor of regional waste collection and treatment facilities, which are quite feasible, particularly in the watershed units of the upper Basin.