In about 1767 Swiss physicist Horace-Bénédict de Saussure invented an
elaborate glass enclosure whose interior became very hot when it was
exposed to direct sunlight.

In about 1824 French mathematician and physicist Jean Baptiste Joseph
Fourier theorized that the energy in visible sunlight entered de
Saussure's glass enclosure unobstructed and warmed the interior; but
when this energy was re-radiated as longer-wave black radiation it
could not escape through the glass, and so it was trapped inside the
enclosure where it accumulated and caused the temperature to rise.
Fourier also theorized that the same mechanism warmed the Earth's
atmosphere and surface.

Subsequently, Swedish chemist Svante August Arrhenius added to
Fourier's ideas by theorizing that rising levels of atmospheric carbon
dioxide would raise the temperature of the Earth's surface.

But in 1906 American experimental physicist R. W. Wood published the
results of an experiment that demonstrated that a glass greenhouse was
not heated by trapped long-wave (infrared) radiation. In fact, the
glass windows excluded more infrared energy entering the greenhouse
than they trapped inside the greenhouse, which actually lowered the
temperature inside a glass greenhouse when compared with a comparable
quartz greenhouse. Following Fourier's original logic, Wood concluded
that the Earth's atmosphere and surface were not warmed by trapped
infrared radiation.

More than a century later, serious monographs on the greenhouse
effect, such as the one published in Wikipedia, have begun to
acknowledge that the greenhouse effect does not work in greenhouses.
They insist, nonetheless, that this effect does work in the Earth's
atmosphere.

It is very curious that the claimed atmospheric greenhouse effect
produced by carbon dioxide has not been tested directly and reported
on. But while no one appears to have deliberately measured temperature
as a function of carbon dioxide concentration, numerous agricultural
experiments have been conducted to study the effect of elevated levels
of carbon dioxide ("enriched CO2") on the growth of plants in
greenhouses and in unusual natural settings. Oddly enough, these
agricultural studies, which are often conducted using carbon dioxide
at very high concentrations, do not seem to report the presence of any
unusually high temperatures.