SUMMARYThe measurement of evapotranspiration from large land areas is, like the measurement of evaporation from large areas of water, a formidable problem and not often accurately accomplished. Nevertheless, it is essential for an engineer to try to check the results of theoretical or empirical analyses, or small-scale experiments, against nature where natural phenomena are concerned and the results of an analysis of the hydrology of the Thames are presented herein. Conclusions favouring the Thornthwaite method are reached.