The kinetics of halide-ion transfer to the vacuole of Nitella translucens have been further examined. The percentage of total tracer found in the vacuole can be expressed as the sum of two components, a fast component important at short times, and a slow component increasing as the level of tracer in the cytoplasm rises. In the individual cells in any given experiment the rate of transfer to the vacuole in the fast component is a quantized fraction of the total influx. This statement expresses the fact that in each experiment the cells can be divided into groups, each covering the whole range of influx values and having the same mean influx, whose mean vacuolar percentages are in the ratio 1:2:3 (with also a few cells having higher values of vacuolar percentage). There appears therefore to be a process of quantized discharge to the vacuole from an internal cytoplasmic compartment. Double-labelling with 36Cl and 82Br was used to measure both fast and slow components in the same cell. Although the bromide influx was only one-third of the chloride influx from equal external concentrations of O.6 mM, the vacuolar percentages for bromide and chloride were equal in cells loaded for the same time in bromide and chloride, whether the loading was for a short or a longer time. Therefore both slow and fast components are equal for the two ions, and in cells labelled for a short time in bromide and a longer time in chloride, the vacuolar percentage for bromide can be used as a measure of the fast component for chloride. The contribution of the slow component to the vacuolar transfer increases with time, but the slope of its time dependence is proportional to the influx in the cell. It is not quantized. Hence the rate of transfer to the vacuole can be specified only as a fraction of the total influx, and not as an absolute rate, and the exchange with the slow cytoplasmic compartment is very closely linked to the influx. The equality of the slow components of vacuole transfer for bromide and chloride (in spite of very different influxes) is consistent with the hypothesis that the slow exchange reflects exchange with pinocytotic vesicles, but is difficult to explain in terms which do not involve a common entry process for salt, in a unit involving both bromide and chloride. The significance of the quantized discharge, and the relation between the fast and slow components, are not yet understood. The increase in the slow component as the cell becomes labelled is very steep, and hence even the slow cytoplasmic compartment can represent only a part of the cytoplasm, in which the specific activity is higher than the rest.