On this page
The flask that keeps your tea hot was invented to keep something extremely cold. James Dewar needed a way to hold liquefied gases long enough to study them, and the vessel he built in 1892 is still used in every low-temperature laboratory. This post covers who Dewar was, the physics of the flask, how he liquefied hydrogen, and what students can learn from his work.
Who was James Dewar?
The Royal Institution's biography of James Dewar gives the outline. He was born in Kincardine, Scotland, in 1842, went to Dollar Academy and studied science at the University of Edinburgh. In 1875 he became Jacksonian Professor of Natural Philosophy at Cambridge, a post he held alongside his work at the Royal Institution in London. At Cambridge he worked on spectroscopy; at the Royal Institution he concentrated on cryogenics, the study of matter at very low temperatures. With Frederick Abel he invented cordite, a smokeless explosive. He set up the Davy-Faraday Research Laboratory in 1896 and was President of the Chemical Society from 1897 to 1899.
Why does a vacuum flask keep things cold?
Heat moves in three ways: conduction, convection and radiation. A Dewar flask tackles each one.
Conduction needs matter to pass vibrations along. A vacuum between two walls has almost none.
Convection needs a fluid to carry heat. No gas in the gap means no convection.
Radiation can cross a vacuum. Silvering the walls reflects most of it back.
The Royal Institution's page on Dewar's first vacuum flask explains how the design developed. In 1872 Dewar and Peter Tait at Edinburgh had made a vacuum-insulated goblet to keep things warm. Twenty years later he made a vacuum-jacketed flask with a narrow neck to keep things cold, and later improved it with an even narrower neck and silvering. It was first shown at the Royal Institution on Christmas Day 1892. Dewar never patented it; the Thermos company patented a version in 1904.
A quick calculation. The Stefan-Boltzmann law says a black surface at temperature T radiates σT⁴ watts per square metre, with σ = 5.67 × 10⁻⁸ W m⁻² K⁻⁴. A room-temperature wall at 293 K radiates about 5.67 × 10⁻⁸ × 293⁴ ≈ 418 W per square metre. Liquid hydrogen at about 21 K sends back only about 0.011 W per square metre. So almost all the radiation flows inward, towards the cold liquid, and the only defence is to reflect it. That is why silvering mattered so much.
Common mistakes: students often say the vacuum stops all heat flow. It stops conduction and convection only; radiation needs the silvering. Another is to forget the narrow neck, which is where conduction through the glass itself enters.
Our guide to heat conduction and thermal resistance works through conduction problems in detail.
How did Dewar make liquid hydrogen?
Dewar described it himself in his 1898 note on the liquefaction of hydrogen and helium in the Journal of the Chemical Society. Earlier researchers had seen only a momentary froth of hydrogen in a glass tube. Dewar wanted to collect it as a liquid at its boiling point "in an open vacuum vessel".
After a year building a larger liquid-air plant and many months of failures, he succeeded on 10 May 1898. Hydrogen cooled to −206 °C and at 180 atmospheres escaped continuously from a nozzle into a doubly silvered vacuum vessel kept below −200 °C. Liquid hydrogen began to drip into a second vacuum vessel. In about five minutes he collected 20 cubic centimetres before the jet froze, blocked by air frozen out of impure hydrogen. The yield was about 1 per cent of the gas.
He thanked his assistant Robert Lennox, without whose "engineering skill, manipulative ability, and loyal perseverance" the result "might have been indefinitely delayed". The work was dangerous: the Royal Institution notes that Lennox lost an eye in an explosion.
What did he get wrong, and fix?
In the same 1898 note, Dewar reported that a sample of helium appeared to condense when cooled in liquid hydrogen. He concluded there could not be much difference between their boiling points. A year later, in his 1899 report on solid hydrogen, he wrote that "pure helium is definitely proved to be more volatile than hydrogen". He also gave hydrogen's boiling point as 21 degrees absolute and its melting point as about 16 or 17 degrees absolute.
He was right to correct himself. Helium was eventually liquefied by Heike Kamerlingh Onnes at Leiden in 1908, at about 4 K. You can read that story in our post on Kamerlingh Onnes and superconductivity. Dewar was nominated several times for a Nobel Prize but never won, according to the Royal Institution.
What can a student take from Dewar's way of working?
Build the tool your question needs. He could not study liquid gases without a way to keep them, so he invented one.
Credit the team. His paper names the assistants who made it work.
Report what you see, then revise. His helium observation was wrong, and he said so in print.
Use simple physics on hard problems. Every feature of the flask follows from the three ways heat travels.
If the three modes of heat transfer still blur together, the Fermi Concept Library covers every concept in physics and chemistry with its own lesson.
Try this at home: fill two cups with hot water, wrap one in foil, and measure both temperatures after 20 minutes. Then explain the difference in terms of conduction, convection and radiation.
