= Joule's free-expansion experiment
{c}
In his paper published in 1845, Joule connected two copper receivers with stopcocks, filled one with compressed dry air, and evacuated the other with an air pump. He immersed the connected receivers in water in an insulated arrangement, stirred the water, and read a sensitive <thermometer>. He then opened the stopcocks, let the air redistribute, stirred again, and compared the water's <temperature> with its initial value. Alternating control trials accounted for effects such as stirring and evaporation. The observation was no detectable net cooling of the common bath, rather than a record of the <gas> <temperature> at every instant. https://upload.wikimedia.org/wikipedia/commons/4/47/The_free_expansion_of_gases-_memoirs_by_Gay-Lussac%2C_Joule%2C_and_Joule_and_Thomson%3B_%28IA_cu31924012329300%29.pdf#page=37[J. P. Joule, On the Changes of Temperature Produced by the Rarefaction and Condensation of Air (1845), reprinted in Ames's 1898 collection, pp. 24–26]{ref} https://pmc.ncbi.nlm.nih.gov/articles/PMC4360093/[John Young (2015), section 5: Joule's paired-cylinder experiments and common-bath observation]{ref}
Joule also immersed the receivers and their connecting piece in separate water vessels. The source receiver cooled while the receiving side warmed, with nearly cancelling thermal effects. https://upload.wikimedia.org/wikipedia/commons/4/47/The_free_expansion_of_gases-_memoirs_by_Gay-Lussac%2C_Joule%2C_and_Joule_and_Thomson%3B_%28IA_cu31924012329300%29.pdf#page=39[Joule (1845), separate-water-vessel comparison, reprint pp. 26–27]{ref}
This distinguishes local transient changes from the final observation for the combined apparatus. Insulating its outer boundary suppresses external <heat> exchange; it does not prevent <heat> exchange between the <gas>, the receivers and the water inside that boundary.
An idealized modern version can use two rigid chambers joined by a valve, with <thermometers> in the chambers and on the walls. Prepare one chamber with <gas> and the other evacuated, disconnect the pump, close the apparatus to matter exchange, and let the initial <temperature> settle. Record initial and final equilibrium <pressure> and <temperature>, along with the wall response. Repeated closed-valve runs can reveal drift and valve-actuation effects; independent heating calibrations can estimate the <heat capacity> of the participating apparatus. This is a proposed measurement protocol, not a claim of a particular measured result. Fast local readings can characterize the release, but they cannot replace the final equilibrium readings in the endpoint <internal energy> balance.
Back to article page
