1. Today the Meter Is Defined by the Speed of Light and Time, Not a Metal Bar
The definition of the meter changed in a surprising way. In 1983, an international meeting called the CGPM defined the meter as "the length of the path travelled by light in vacuum during a time interval of 1/299,792,458 of a second" [2]. Usually, we measure the speed of light after defining length. However, in this definition, the number for the speed of light is fixed first. The National Institute of Advanced Industrial Science and Technology (AIST) explains that this actually means "defining the speed of light as 299,792,458 m/s" [1]. The wording of this definition was updated in 2018 and has been used since May 20, 2019, but the length itself did not change [2]. If we stopped using a bar, the length of the meter might have changed. To understand why we switched from a bar to light, we must look back at what the first meter was based on.
2. The Original Meter Was Based on the Distance from the Equator to the North Pole
The original meter was defined as one ten-millionth of the distance from the Equator to the North Pole along a meridian (a line connecting north and south). The French Academy of Sciences established this in the late 18th century. Delambre and Méchain measured the distance from Dunkirk to Barcelona between 1792 and 1798 [3]. Their measurements resulted in a meridian length of 10,000 km. Today, the value is 10,019.66 km. The difference is about 0.02%, which is about 0.2 mm per meter (calculated from AIST values). AIST explains that this difference was due to the limits of measurement technology at the time [1]. However, you cannot put the Earth on a desk. For everyone to use the same length, we needed something we could hold in our hands.
3. The Metal Meter Bar Reached Japan in 1890, but It Had Weaknesses
The Meter Convention was signed by 17 countries in Paris in 1875, creating the International Bureau of Weights and Measures (BIPM) [2]. Japan joined in 1885 (Meiji 18) and received the "National Prototype Meter of Japan" (No. 22) from the BIPM in 1890 [1]. The International Prototype Meter, approved at the first CGPM in 1889, was an alloy of 90% platinum and 10% iridium. The distance between the marks on this bar at 0°C (the temperature of melting ice) was one meter [2]. At that time, the difference between national prototypes and the international prototype was kept within 0.01 mm [1]. AIST notes that prototypes are physical "objects" that can break or change over time. This conflicted with the ideal of the metric system: a standard based on an "immortal natural object." This weakness was recognized even then [1]. So, what could be used that would not disappear and could be copied anywhere with the same length?
4. In 1960 the Meter Was Redefined Using the Wavelength of Light
The idea of using light as a wave existed for a long time. In 1873, Maxwell and in 1892, Michelson measured light wavelengths compared to the meter prototype, showing it was possible [1]. At the 11th CGPM in 1960, the meter was defined as "1,650,763.73 wavelengths of the radiation corresponding to the transition between the 2p10 and 5d5 levels of the krypton-86 atom in vacuum." This ended the definition based on the prototype bar used since 1889 [2]. Why was krypton-86 chosen? AIST explains that this light had the narrowest line width among many lamp lines. A narrower line width means the wavelength is more precisely defined. However, the width was still about 400 MHz, which limited the standard's accuracy to about 10^-7 to 10^-8 [1]. Interestingly, in the same year (1960), ruby lasers and helium-neon (HeNe) lasers were reported. Because lasers have even narrower line widths, they promised accuracy of 10^-10 to 10^-11 if used for definitions [1].
5. Fixing the Speed of Light Let Scientists Define the Meter by Time
In the 1960s and 1970s, research flourished to find the speed of light by measuring laser wavelengths and frequencies. From this time, length and time became closely related [1]. In 1967, the definition of the second also changed from Earth's orbit to 9,192,631,770 periods of radiation from cesium-133 atoms [1]. In 1973, an advisory committee set a provisional value for the speed of light at 299,792,458 m/s, and values from lasers in various countries matched. The CGPM recommended this value in 1975. In 1983, without changing this value, the meter was defined as the path of light [1][2]. AIST lists three reasons why defining the speed of light as a constant is better than defining by wavelength: the accuracy of the speed of light was limited by krypton's accuracy; krypton lamp standards could still be used even if the definition was abolished; and it was difficult to decide which laser to use for the definition [1]. The BIPM resolution also stated that stabilized laser light is easier to reproduce than krypton lamps [2]. Because the value of the speed of light did not change, no noticeable step in length occurred within the relative uncertainty of about 4×10^-9 for the best realizations [2]. AIST notes that the premise for this change was the principle of the constancy of the speed of light in special relativity (the idea that the speed of light is constant regardless of time or direction of observation) [1].
6. How Much More Accurate Did the Meter Become with Each New Definition?
According to AIST, the accuracy of Japan's national standard improved over time. In the era of the prototype bar (1889–1960), accuracy was about 10^-7. With the krypton lamp (1960–1983), it was 10^-8 to 10^-9. With the iodine-stabilized HeNe laser (1983–2009), it was 2.1×10^-11. With the optical frequency comb (2009–present), it is 6.8×10^-14 [1]. Comparing the prototype's 10^-7 with the optical comb's 6.8×10^-14, the number is about 1.5 million times better. The 1983 definition actually fixes the value of the speed of light. There are multiple ways to create a meter: measuring by light flight time, by wavelength, or using a list of recommended lights [1][2]. This means we can keep improving how we make a meter without changing the definition itself.
7. Try Comparing Two Rulers at Home to See Why Measurement Scales Differ
Try placing two rulers from home side by side. Align the 0 positions and compare the marks up to 30 cm. If you find a discrepancy, the rulers might not be bad; the thickness of the lines or the printing position can change how they look. In the era of the prototype bar, the difference between national bars and the international prototype was kept within 0.01 mm [1]. Comparing rulers at home is just a way to experience the feeling of "aligning" standards. For more reading, AIST's "SI Brochure: Meter" organizes the changes in accuracy of Japan's national standards [1].