Direct and indirect calorimetric experiments were performed on flowers of the sacred lotus, Nelumbo nucifera, and compared with temperature measurements. To this end, a simple, light and cheap heat-flow calorimeter of the twin type was developed to monitor the heat output of lotus flowers in an outdoor pond. Each side of the calorimeter consisted of a water jacket as a heat sink surrounding a 730 ml concentric can as a calorimetric vessel. The vessel and heat sink were thermally connected via a Peltier element but otherwise thermally isolated. Both water jackets were housed in a styrofoam box and connected in parallel to a thermostated water bath. The calorimeter exhibited a mean sensitivity of 25.8 mV W-1, a time constant of 8 min and a 24 h baseline stability better than 18 of the chosen range. This differential calorimeter was placed around lotus flowers approximate to 1 m above the water level. Direct calorimetry was accompanied with indirect calorimetry by measuring oxygen consumption rates of the flowers with open-how respirometry, and the patterns of temperature change were recorded with thermocouples. Flowers maintained mean temperatures of ca. 30.7 degrees and 34.2 degrees C at mean calorimeter temperatures of 18.4 degrees and 30.4 degrees C, respectively, demonstrating good thermoregulatory ability. Metabolic heat production averaged ca. 0.51 W at the low temperature and 0.25 W at the high temperature. Dry heat loss to the calorimeter averaged -0.62 W and -0.17 W, respectively, which indicated that there was a small condensation of atmospheric water vapor inside the calorimeter at the low temperature, but net evaporation from the flower at a level of ca. 33% of heat production occurred at the high temperature.In a set of laboratory experiments on cut lotus flowers, a heat-flux budget was constructed from measurements of heat production (open-Row respirometry), heat loss (gradient-layer calorimeter of the Benzinger/Kitzinger type), and evaporative heat loss (gravimetric). Heat production rate was ca. 0.3 W and was balanced almost completely by evaporative heat loss into the calorimeter air (25 degrees C; 37% relative humidity). Therefore, total heat Bur by convection, conduction and radiation was essentially zero, despite the flower's heat-producing receptacle prevailing ca. 5 degrees C higher than the calorimeter air. Heat from the receptacle was apparently transferred to the petals which, in turn, lost it mainly through evaporation. Equivalence of direct and indirect calorimetry substantiated the assumed caloric equivalent of oxygen consumption of 21.1 J ml(-1) and indicated that there was no conservation of energy in metabolic processes during thermogenesis.