HE general pattern of climate in the tropics is one of temperature |uniformity and rainfall variability. Rainfall is the climatic factor of greatest economic importance, and investigations into the water balance have the same significance as investigations into the heat balance in higher latitudes. Knowledge of rainfall-amount, seasonal distribution, variation, and effectiveness-in relation to the water needs of agriculture, industry, and urban communities is essential to the success of development schemes in low latitudes. Had this knowledge been available, such ill-fated schemes as the cultivation of cotton in the vicinity of Barberton, southeastern Transvaal, citrus culture in Natal, groundnuts in Tanganyika, and sorghums in Queensland might not have been undertaken. "What happens to the rain?" is a fundamental problem in agricultural climatology. No climate can be described as moist or dry from precipitation aggregates alone. Neither do measurements of precipitation alone tell the agriculturist of the availability of moisture in the soil. To find out what happens to the rain, it is necessary to study the water losses by evaporation and transpiration from the earth's surface. And thence one passes to water need. It was Thornthwaite' who introduced the concept of "potential evapotranspiration," or water need, into climatic literature, and this paper presents the results of the application of his concept to the Rhodesias and Nyasaland.