The index of refraction of air at STP is about 1.0003. When was this established, and to what precision?

As an analogy, you can find historical summaries of values for the speed of light, showing its convergence on the current value. Is there a similar timeline available for the index of refraction of air? I'm particularly interested in early values and specific dates (year level, not days or months).

Since the ability to measure optical angles was very advanced by the time vacuum technology became common, I'd assume that the early values (1800's) would be very close to the current value. Of course, the current value is actually changing (since the amount of CO2 is changing), but that's a different question.


According to this lecture, Newton did measure $n_{air}$ compared with vacuum, tho' they give no details. Perhaps the original source, "Newton, Opticks, The Fourth edition, Ed. W. Innys, London, 1730," provides more details. It's available at Gutenberg.org .

Peeking at Newton's book, I find

The Refraction of the Air in this Table is determin'd by that of the Atmosphere observed by Astronomers. For, if Light pass through many refracting Substances or Mediums gradually denser and denser, and terminated with parallel Surfaces, the Sum of all the Refractions will be equal to the single Refraction which it would have suffer'd in passing immediately out of the first Medium into the last. And this holds true, though the Number of the refracting Substances be increased to Infinity, and the Distances from one another as much decreased, so that the Light may be refracted in every Point of its Passage, and by continual Refractions bent into a Curve-Line. And therefore the whole Refraction of Light in passing through the Atmosphere from the highest and rarest Part thereof down to the lowest and densest Part, must be equal to the Refraction which it would suffer in passing at like Obliquity out of a Vacuum immediately into Air of equal Density with that in the lowest Part of the Atmosphere.
There may be more details elsewhere but I haven't read the entire chapter.

There are plenty of papers dating back at least to the 1950s which provide empirical formulas for calculating the index based on gas ratios and relative humidity of the sample. But to your first question: any published value for $n$ is published with the calculated uncertainty; if that's missing then the default is to assume $\pm 1$ in the last digit, or half that. Since you quoted a number without reference to a specific sample, I suspect that's as precise as one can get for "average atmosphere at sea level and $25^{\circ} C $ .

| improve this answer | |

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.