Measurements with Piche evaporimeters have been conducted within a thermometer screen over a grass surface at several heights. The results show a good correlation with model calculations but differ somewhat in absolute value. Comparisons with the Penman potential evaporation results show a considerable overestimation. With a simple factor, or with a lower measuring position, the results can be brought in a better agreement. The idea of G. Stanhill was verified and proved to be useful only for long periods.