The main problems of early wine-spirit thermometers, with special reference to the eighteenth century, are considered with a holistic approach based on historical sources, physical mechanisms, and mathematical relationships. Thermometers were hardly comparable and were vulnerable at extreme temperatures, including calibration. Wine-spirit and the vapor pressure exerted at different temperatures, especially in proximity of the boiling points of spirit and water, are analyzed, because they caused the failure of glass tubes. The method suggested by Réaumur of completely removing air from the tube when it was sealed and the opposite one by Micheli du Crest of leaving some air inside are discussed, as well as advantages and disadvantages of using wine-spirit at different mole fractions of ethyl alcohol and water. The original, so-called “true Réaumur” thermometer, its calibration, scale, and response are investigated. The equations that evaluate the deviation from linearity for various factors are derived. Equations are given to convert readings taken with the “true Réaumur” and other early thermometers to Celsius, and at the same time correcting them from the departures due to wine-spirit and the particular calibration. Finally, the direct Celsius scale has been found to be known earlier than believed, in 1740.