Eight Historical Temperature Scales and Their Measurement Principles

How temperature units developed

Temperature scales are measurement systems that assign numerical values to thermal conditions. A thermometer responds to temperature through a measurable physical change, such as liquid expansion, gas pressure, electrical resistance, or radiation intensity. A scale then converts that response into numbers by defining reference points, a zero point, and intervals between graduations.

Early thermometry was not simply a matter of choosing different labels for the same quantity. Seventeenth- and eighteenth-century scientists had to solve several practical problems: which thermometric liquid to use, how to seal the instrument, what fixed points were reproducible, and how to divide the distance between those points. Water freezing and boiling were attractive references because they were accessible, but they depended on pressure, purity, and experimental procedure. Human body temperature was convenient but not truly invariant. Mixtures of ice, salt, alcohol, mercury, and water all introduced their own limitations.

This history produced many temperature scales, most of which disappeared as instrument making and metrology became more standardized. The eight scales below—Newton, Rømer, Fahrenheit, Réaumur, Delisle, Celsius, Kelvin, and Rankine—show how different measurement principles developed. Some were empirical scales based on liquids and fixed points. Others, especially Kelvin and Rankine, were absolute thermodynamic scales tied to the concept of absolute zero.

Today, Celsius, Fahrenheit, Kelvin, and limited Rankine use remain relevant. Newton, Rømer, Réaumur, and Delisle are mainly of historical interest, but they remain useful examples of how measurement systems evolve when scientists need repeatable numbers for the same physical condition.

Newton’s temperature scale from 1701

Isaac Newton’s temperature scale was one of the earliest systematic attempts to quantify heat. In his 1701 work, Newton described “degrees of heat” rather than temperature in the modern sense. His starting point was practical: assign reproducible values to familiar thermal states and then extend the scale to hotter conditions.

Newton used melting snow as 0 on his scale and human body heat as 12 degrees. The thermometer he used contained linseed oil, whose volume changed with temperature. Liquid expansion was a common thermometric principle: as the liquid warmed, it expanded and moved along a graduated tube. However, linseed oil was not ideal for high-temperature measurement. It could be difficult to use reliably near its upper thermal limits, and its behavior was not as stable or convenient as mercury later proved to be.

To extend the scale beyond body temperature, Newton used cooling behavior and the melting points of materials. He compared how heated bodies cooled in air and used melting or freezing points of metal alloys as reference conditions. This was technically important because it moved thermometry beyond everyday temperatures and toward reproducible material transitions. Historical descriptions of the Newton scale include alloy melting points involving bismuth, lead, and tin, as well as much hotter states such as red-hot or glowing iron. These fixed points should not be read as modern high-precision values, but they show the principle: a scale could be anchored to observable phase changes or thermal appearances.

In modern linear comparison, the Newton scale is often related to Celsius by taking water freezing as 0 °N and water boiling as about 33 °N, so:

The Newton scale is no longer used for practical temperature measurement. Its significance is historical: it represents an early bridge between qualitative heat descriptions and calibrated thermometry.

Rømer’s scale and early fixed-point calibration

Ole Rømer, better known for astronomical measurement, also contributed to thermometer calibration. His work illustrates a key early problem in thermometry: a thermometer needed not only a responsive liquid but also a reproducible scale that different instrument makers could copy.

Rømer used sealed glass thermometers containing wine or an alcohol-water mixture. Alcohol-based liquids expanded substantially with temperature, making them useful for visible readings in narrow tubes. Their disadvantages included nonlinearity and limited high-temperature range, but they were practical for many early instruments.

The Rømer scale used water’s freezing and boiling points as calibration references. Commonly cited values place the freezing point of water at 7.5 °Rø and the boiling point at 60 °Rø. This means the interval between freezing and boiling water was divided into 52.5 Rømer degrees. Rømer’s zero was deliberately placed below the freezing point of water. That choice reduced the need for negative numbers in ordinary weather and domestic measurements, a practical concern when scales were intended for direct reading.

Using the usual water fixed points, the scale may be compared with Celsius as:

Rømer’s scale influenced later thermometer makers, including Fahrenheit, but it did not survive as a standard practical scale. Its main importance lies in fixed-point calibration: assigning numbers to repeatable thermal states rather than relying only on arbitrary tube markings.

Fahrenheit’s scale and mercury thermometers

Daniel Gabriel Fahrenheit made a major contribution to practical thermometry by improving both scale design and thermometer construction. His work was influenced by earlier fixed-point approaches, including Rømer-style calibration, but his thermometers became more reproducible and widely used.

Fahrenheit’s early scale evolved over time. In one historical arrangement, water freezing became 32 °F and normal human body temperature was placed near 96 °F. This body-temperature point was convenient for instrument checks, although it was not a stable metrological reference in the modern sense. The later familiar Fahrenheit scale used water freezing at 32 °F and water boiling at 212 °F under standard atmospheric conditions, creating a 180-degree interval between those two points.

A major technical advance was Fahrenheit’s use of mercury as a thermometric liquid. Compared with alcohol, mercury offered several advantages for many thermometers: it was visible in glass, had a broad useful liquid range for common laboratory and weather measurements, and generally behaved more predictably over useful intervals. Alcohol remained useful for low-temperature thermometers, but mercury became strongly associated with precision liquid-in-glass instruments.

The origin of Fahrenheit’s zero point is often discussed in relation to cold brine mixtures, but such historical explanations should be treated cautiously. Whatever the exact origin, the scale’s practical success came from reproducible instruments and convenient divisions, not from a single perfect natural zero.

The basic modern relationship is:

Fahrenheit remains in everyday use in the United States and in some sectors or countries that retain customary measurement traditions. It is also still encountered in weather reporting, cooking, HVAC work, and older engineering documentation.

Réaumur’s alcohol-based temperature scale

René Antoine Ferchault de Réaumur developed his temperature scale in the early eighteenth century to improve comparability in scientific measurements. The Réaumur scale set the freezing point of water at 0 °Ré. Its upper reference was associated with the boiling point of water, commonly represented as 80 °Ré.

The scale was closely tied to the thermometric liquid. Réaumur used an alcohol-water mixture and selected the graduation so that the liquid’s expansion between the freezing and boiling points of water corresponded to an 80-degree interval. This made the instrument’s principle easy to understand: temperature was inferred from liquid expansion, and the scale was built around a defined expansion range.

However, the same design also created limitations. Alcohol-water mixtures do not expand perfectly linearly over all temperatures. Alcohol also has a relatively low boiling point compared with mercury, making it less suitable for thermometers intended to approach boiling water. In addition, the boiling point of water depends on atmospheric pressure, so any scale using boiling water as a fixed point required attention to pressure conditions.

The usual Celsius comparison is simple:

Réaumur thermometers were used in parts of Europe, especially before Celsius-based metric practice became dominant. The scale later declined as scientific standardization favored Celsius and, for thermodynamics, Kelvin. Limited niche references to Réaumur may still appear in some traditional food or processing contexts, but it is not a general modern measurement scale.

Delisle’s inverted scale

Joseph-Nicolas Delisle developed one of the more unusual historical temperature scales because it ran in the opposite direction from most modern scales. On the Delisle scale, larger numbers represented colder temperatures.

Delisle’s early thermometer work used spirit of wine and an initial reference related to cellar temperature. Spirit of wine, an alcohol-based liquid, was suitable for moderate temperatures but not for reliable measurements near the boiling point of water. When Delisle later shifted the fixed point to boiling water, mercury became the more appropriate thermometric liquid because alcohol would be unsuitable close to that range.

The mature Delisle scale used water’s boiling point as 0 °D. A later recalibration assigned the freezing point of water to 150 °D. Because the scale was inverted, cooling from boiling water to freezing water increased the reading from 0 to 150. This may feel counterintuitive today, but it was not technically impossible. A scale only needs consistent fixed points and intervals; its direction is a convention.

The common conversion relationships are:

The Delisle scale saw historical use in Russia during the eighteenth and nineteenth centuries before disappearing from practical use. Its inverted numbering is a useful reminder that temperature scales are constructed systems. The physical state is real, but the numerical direction and zero point are design choices.

Celsius from inverted scale to modern centigrade use

Anders Celsius proposed a temperature scale in 1742, but the original arrangement was the reverse of the modern Celsius scale. In Celsius’s original form, water’s boiling point was 0 °C and water’s freezing point was 100 °C. The scale was therefore inverted: colder conditions produced larger values.

The scale was later reversed to the familiar modern arrangement: 0 °C for water freezing and 100 °C for water boiling under defined pressure conditions. The inversion made the scale more intuitive for many applications. In weather, biology, refrigeration, and ordinary laboratory work, increasing numbers naturally correspond to increasing warmth. The freezing point of water at 0 °C is also a convenient reference for everyday interpretation.

The exact attribution for the inversion is historically uncertain. It is commonly associated with figures such as Carolus Linnaeus or Martin Strömer, but the broader technical change was the adoption of a more intuitive orientation for practical thermometry.

Celsius was long called “centigrade” because the interval between freezing and boiling water was divided into 100 parts. The name Celsius later became standard to avoid ambiguity, since “centigrade” can also mean a hundredth-part division in other measurement contexts.

The conversion with Fahrenheit is:

Celsius is now the dominant everyday temperature scale in most countries and is widely used in science for relative temperature reporting. In strict thermodynamic calculations, however, Celsius temperature differences may be useful while absolute temperatures are normally expressed in kelvin.

Kelvin as an absolute thermodynamic scale

The Kelvin scale, associated with Lord Kelvin, was developed to express temperature on an absolute thermodynamic basis. Unlike empirical scales that choose convenient fixed points such as freezing water, Kelvin begins at absolute zero. Absolute zero is the zero point of thermodynamic temperature in the classical sense: the limiting condition at which thermal energy is minimized.

Kelvin uses increments equal in size to Celsius increments. A temperature difference of 1 K is the same size as a difference of 1 °C. The zero points are different: 0 K corresponds to −273.15 °C, while 273.15 K corresponds to 0 °C.

This makes Kelvin fundamentally different from Celsius or Fahrenheit in calculations involving thermodynamic behavior. Gas laws, entropy calculations, heat engine efficiency, radiation laws, and many equations in physics require absolute temperature. Using Celsius or Fahrenheit values directly in those equations would give incorrect results because their zero points are arbitrary relative references, not thermodynamic zero.

Historically, the kelvin was defined using the triple point of water, with the triple point assigned 273.16 K. Modern SI practice no longer defines the kelvin by a physical water sample. Since the 2019 SI revision, the kelvin is defined by fixing the numerical value of the Boltzmann constant. In practical metrology, however, calibrated fixed points and standardized procedures still matter for realizing and transferring temperature measurements.

Useful relationships are:

Kelvin is the primary absolute temperature scale in scientific work, engineering analysis, thermodynamics, and metrology. It is written as K, not degrees Kelvin.

Rankine and absolute temperature in Fahrenheit-sized units

The Rankine scale, developed by William Rankine, is an absolute thermodynamic scale like Kelvin. Its zero point is absolute zero. The difference is the interval size: Rankine uses Fahrenheit-sized increments rather than Celsius-sized increments.

This means that a temperature difference of 1 Rankine is the same size as a difference of 1 °F. Since Fahrenheit increments are smaller than Celsius increments, the absolute temperature of water freezing is 491.67 R, corresponding to 32 °F or 0 °C. Absolute zero is 0 R, equivalent to −459.67 °F.

The basic relationships are:

Rankine is most often encountered in engineering calculations that use Imperial or U.S. customary units. Examples include ideal gas law calculations in Imperial units, heat transfer work, Carnot efficiency, combustion analysis, and entropy calculations in older or unit-specific references. It gives engineers an absolute scale while preserving Fahrenheit-sized temperature intervals.

Kelvin has largely replaced Rankine in most scientific work because SI units dominate modern science and metrology. Rankine persists where calculations, equipment specifications, or legacy engineering practice remain tied to Fahrenheit and Imperial-unit systems.

Why multiple temperature scales exist

Multiple temperature scales exist because early thermometer makers made different choices about four basic design questions:

• What physical property should the thermometer measure?
• What liquid or sensing medium should be used?
• Which fixed points are reproducible enough for calibration?
• Where should zero and the scale intervals be placed?

Newton’s and Réaumur’s scales show the importance and difficulty of thermometric liquids. Rømer and Fahrenheit show the move toward repeatable fixed-point calibration. Delisle and early Celsius show that even the direction of a scale was a convention. Kelvin and Rankine show the later shift from empirical reference points to absolute thermodynamic temperature.

Many historical temperature scales disappeared as scientific communities standardized instruments and as metric practice spread. Celsius became dominant for everyday and scientific relative temperature measurement in most of the world. Fahrenheit continued where customary measurement traditions remained strong. Kelvin became essential for thermodynamics and fundamental science. Rankine survived mainly in limited engineering contexts using Fahrenheit-sized units.

Each scale reflects a historical solution to the same measurement problem: assigning stable, repeatable numbers to thermal states. The differences between them are not just mathematical conversions; they record the development of thermometers, calibration methods, scientific theory, and measurement standards.