The time course of the appearance in the vacuole of Nitella translucens and of Tolypella intricata of tracer from the outside solution has been studied over short periods of uptake. There are two components of chloride transfer to the vacuole, a fast component linear with time and a second component at longer times whose behaviour is reasonably well described in terms of a single rate constant for exchange; a constant fraction of the total entry is in the fast component and the apparent rate constant for the second component is proportional to the influx. In Nitella the path of rapid transfer involves chloride and sodium, and may also involve a small but variable amount of potassium, but in Tolypella potassium has a significant component of rapid transfer; these correspond to the cations for which chloride-linked components of cation influx have been shown by another worker. Over both parts of the time course the level of activity in the cytoplasm specifies, not the rate of transfer to the vacuole as would be expected, but the rate as a fraction of the influx; the processes of influx to the cell and transfer to the vacuole are intimately linked. It is difficult to explain the results in terms of static membranes and fixed compartments. An explanation in terms of the sequence, entry of salt by pino-cytotic vesicles at the plasmalemma, fusion of these vesicles with the endoplasmic reticulum after some loss of tracer to the surrounding cytoplasm, and transfer to the vacuole in minivacuoles formed from the endoplasmic reticulum, is consistent with the time course found. A model of this kind, involving transport by a dynamic membrane system, seems necessary to explain the results.