In a river on land, one can estimate the average annual flux of a constituent (kg/sec) as: Q x C, where Q is the mean annual discharge (cubic meters per second) and C is the flow-weighted annual mean concentration of the constituent (kg per cubic meter). Does the same formulation apply to the ACC? I say yes; my brother says no. I say that the ACC meets all the criteria for defining a river (net unidirectional flow of water) and so its discharge can be used in the same way as on land. For example, in addition to using Q at Drake Passage to calculate the flux of a constituent, one could also use it to estimate a global mean water replacement time of the ocean (V/Q). My brother says that the ACC is not a river because it has no beginning or end; it flows around in a circle and so has no net flow. Consequently, the net flux of a constituent is zero. I counter that if the ACC is continually exchanging water with three oceans, how can it be characterized as flowing in a circle. For every cubic meter of water flowing through Drake Passage, doesn't there have to be a cubic meter's worth of water flowing into the ACC from one or more of the oceans? If so, would it be correct to think of the ACC as analogous to a river on land and the three oceans as secondary circulation cells each with a net flow of zero (what goes in equals what goes out)? Any thoughts?