T HE time-honored and still useful methods of indirect animal calorimetry are based on measurement of the respiratory exchange and urinary nitrogen excretion, followed by conversion of oxygen consumption values to Calories. This procedure involves determination of the non-protein respiratory quotient (RQ), followed by reference to the classical Zuntz-Schumberg table of heat values of oxygen at various RQ's. In practice, the protein metabolism is often assumed to be negligible, especially under basal conditions, and the total RQ is used to estimate the heat production. Possible confounding sources of variation in RQ have been described, and the dangers inherent in an over-simplified interpretation of the RQ pointed out (Cathcart and Markowitz, 1927;Richardson, 1929;Mitchell, 1935). Nevertheless, as Mitchell (1935) has stated, accurate estimates of heat production can be obtained from measurements of the gaseous exchange, "at least within the RQ limits of 0.707 and 1.000." Except for hibernants, normal fasted mammals apparently rarely have RQ's of less than 0.70, the theoretical RQ for the oxidation of fat. For example, Benedict and Fox (1934) fasted rats until death, but in no case was an RQ less than 0.70 recorded. In contrast, many fasting RQ values of less than 0.70 have been reported in the literature on respiratory exchange in birds. These have occurred with both gravimetric methods (Mitchell et al.