April 1896

Svante Arrhenius

Arrhenius publishes the first quantitative CO₂–temperature calculation

Arrhenius used heat coming from the Moon as a stand-in for the radiation leaving Earth.

Philosophical Magazine, April 1896

The April 1896 issue of the Philosophical Magazine published Svante Arrhenius's “On the Influence of Carbonic Acid in the Air upon the Temperature of the Ground,” at volume 41, pages 237–276.

A footnote identifies the English article as an extract from a paper presented to the Royal Swedish Academy of Sciences on 11 December 1895, so the result predates the printing. Debates at the Stockholm Physical Society about the causes of ice ages, plus Arvid Högbom's carbon-cycle estimates, had prompted the work.

Moonlight as the measuring instrument

Arrhenius built his estimates from Samuel Langley's infrared observations of moonlight passing through the atmosphere, using Frank Washington Very's lunar-temperature estimate to reduce the data to Earth-temperature conditions.

The calculation was done by hand across seasons and latitude bands, producing tables rather than a single number: doubling CO₂ gave about 5–6 °C averaged over the latitude zones. The paper stated that geometrically increasing “carbonic acid” produced a nearly arithmetical increase in temperature — with an explicit warning that the rule held only over the range investigated.

An ice-age paper with an industrial footnote

The paper did discuss industrial CO₂, but its purpose was explaining glacial and interglacial temperatures. At the world's roughly 500 million tons of annual coal production, Arrhenius reckoned combustion was adding scarcely one-thousandth of the atmospheric stock per year, and put a doubling about 3,000 years away.

“Greenhouse law” is a later label, not his name for the rule. Fourier had argued that the atmosphere retained terrestrial heat, Eunice Newton Foote showed that CO₂-rich air warmed strongly in sunlight, and John Tyndall demonstrated infrared absorption by gases including CO₂ and water vapour. What was new here was the quantitative global calculation.

Sources

Researched 23 Aug 2026 4 sources 2 audit passes, 1 issue open

How this was checked

Researched from the web into a fact sheet, rewritten from that sheet, then audited against it by a different model. How the pipeline works →

What the sources leave uncertain

April 1896 is independently supported as the journal issue date, but no exact day of publication was found, so month precision is appropriate.
A footnote identifies the English article as an extract from a paper presented to the Royal Swedish Academy of Sciences on December 11, 1895. The intellectual result therefore predates its April 1896 publication.