Galileo's first thermoscope of 1592 and the mercury thermometer devised by Fahrenheit in 1715 are separated by over a hundred years of research on thermal expansion. During that period scientists have learnt to compare precisely the temperature of bodies; on the way they discovered the abnormal expansion of water and refuted the ancient belief that our body is colder by night than by day (Santorio, before 1634). It was assumed that body temperature changes are due to absorption or emanation of subtle matter called warmth. At first it was believed that it consists of tiny and very hard atoms endowed with mass. Later (in the period 1760 -1780), when Lavoisier proved that the change of mass in glowing samples can be traced back to combustion reactions, the warmth became an indestructible and weightless liquid. In this form warmth theory (after the appearance of the notions of specific heat and latent heat of state transition) agreed perfectly well with Black's exact calorimetric measurements and made prediction of the final temperature of any mixture of solids and liquids possible, as long as the masses and initial temperatures of the components were known. Supported by Lavoisier's authority, the theory survived almost until the middle of the 19th century. As late as 1824 Sadi Carnot published his R\'eflexions sur la puissance motrice du feu, using the notion of warmth to prove that a machine transforming all the heat it absorbs into work is impossible. In his opinion the functioning of heat machines was only possible due to the work performed by the warmth passing from the heater to the cooler, just as the falling water makes the mill wheel move. However, the first observations putting into doubt the existence of the warmth date from the beginning of the 18th century. The appearance of heat at the boring of gun-barrels made Benjamin Rumford think (in 1798) of the possibility of transforming mechanical work into heat. In the same period transmission of heat by radiation with many features characteristic of light was observed and not much later heat produced by electricity flow was noticed. Now no improvement of the properties of warmth was possible to make it compatible with the latest discoveries. It was eventually replaced by the notion of energy, a magnitude that remains constant in any isolated system, independently of the processes taking place.