Like HowStuffWorks on Facebook!

How Pneumatic Tubes Work

How Air Moves Stuff

The metal-and-glass intricacy of pneumatic tube systems seems like a perfect example of Victorian-era technology, the sort of elaborate period gadgetry that you might see Robert Downey Jr. tinkering with in one of the recent Sherlock Holmes movies. But actually, the idea of pneumatics — that is, using pressurized gas to produce mechanical motion — goes back to Hero of Alexandria, a Ptolemaic Greek mathematician, inventor and author who lived in the first century A.D. [sources:Merriam-Webster, Britannica].

Hero apparently was a pretty observant guy. He noticed that the wind, even though it didn't have a visible substance, could push pretty hard on things. That led him to deduce that air was actually composed of tiny, invisible, moving "particles," what today we call molecules. He went on to figure out that if you compressed those moving molecules by jamming them into a tight space or passageway, they'd try to escape, and in the process, push a solid object that was in front of them. He also deduced that if you could create a vacuum — basically, an empty space — that air molecules would try to rush into it. By means of these principles, he wrote, "many curious and astonishing kinds of motion may be discovered" [source: Woodcroft].

Hero used his understanding of how pressurized gases behave to create gadgets like a primitive steam engine and a singing toy bird, but it wasn't until the 1810 that a British engineer, George Medhurst, published a plan for a pneumatic tube transport system[source: Woodcroft].

Medhurst noted that if air was subjected to 40 pounds per square inch of pressure — only about two-and-a-half times the amount that the atmosphere exerts on us at sea level — air molecules would be propelled at 1,500 feet (457 meters) per second, or about 1,000 miles (1,609 kilometers) an hour. When pushing a canister, the speed would only be about 100 miles (160 kilometers) an hour, but that was blazingly fast for 1810 [source: Medhurst].

Unfortunately, Medhurst wasn't as good at actually building machinery as he was at designing it, and he died in 1827, before he actually could construct his vision [source: Stephen and Lee].