New evidence is presented from Carnot's memoir to show that he defined “calorique” (entropy) differently from “chaleur” (heat in the general sense). Calorique is the thermal quantity conserved in reversible cycles; the unit of calorique melts one kilogram of ice; the unit of chaleur raises the temperature of a kilogram of water one degree centigrade, an irreversible process.
Although Carnot used these terms interchangeably for constant temperature (reversible) processes, which is justified when the temperature scale is not specified, he distinguished between chaleur and calorique for work producing reversible processes operating over a finite temperature range, where the distinction becomes necessary.
Carnot's major conclusions on heat engines are deduced on general principles and are not dependent upon any caloric or mechanical theory of heat. Carnot believed in the mechanical theory of heat, but in the last parts of the memoir he introduced, as axioms, conservation theorems for calorique (correctly) and for chaleur (incorrectly) which have tenets in common with the now-discredited caloric theory. These theorems do not require that these quantities be substantive in nature.
