The Forgotten History of the Pneumatic Tube: How Compressed Air Moved Mail Through 19th Century Cities

Between roughly 1860 and 1950, every major Western city ran a pneumatic tube network. Berlin's was the largest, with 400 kilometers of tube and 90 stations at peak. London had 130 kilometers; Paris had 467 kilometers across two networks; New York had 43 kilometers connecting 23 post offices and moving 6 million letters per day at peak. The pneumatic tube was the dominant short-distance urban communication infrastructure for nearly a century, faster than horse-drawn dispatch and competitive with the telegraph for local point-to-point messaging. Most of the infrastructure was scrapped between 1950 and 1990, and the engineering capacity that built it has largely been forgotten.

The basic engineering problem

Sending a physical object through a tube using compressed air requires solving three coupled problems: the air handling (compressors, vacuum pumps, or gravity to move air through the tube system), the carrier (the container holding the message, sized to fit snugly enough to be pushed by the air but loosely enough to slide), and the routing (how the system decides which branch each carrier goes down).

The air handling has two basic configurations. Pressure systems put compressed air behind the carrier and let it push the carrier through. Vacuum systems remove air ahead of the carrier and let atmospheric pressure behind push it. Pressure systems are simpler but require sealing the compressor end; vacuum systems are easier to seal because atmospheric pressure does the work, but require larger vacuum pumps. Most large urban networks used vacuum or a combination of pressure and vacuum at different segments depending on terrain and traffic.

The carrier evolved through several generations. Early systems used leather-jacketed cylindrical carriers with felt seals; later systems moved to all-leather or vulcanized rubber. The sealing problem is fundamental: too tight and friction stops the carrier, too loose and air leaks past it. The mid-19th century solution was leather flaps that flexed outward against the tube wall, creating a seal that increased with pressure differential.

The routing problem was solved by mechanical switches at branch points, sometimes controlled by labels on the carrier read by operators at the branch, sometimes by inertial mechanisms triggered by features of the carrier itself, and in the most sophisticated networks by electromechanical switches signaled by the sending station. The Berlin system in its mature form used an addressing scheme similar in spirit to a modern packet network: each carrier had a destination tag, and the switches at each branch read the tag and routed accordingly.

The 1854 London beginning

The first commercial pneumatic tube system was Joshua Latimer Clark's 1853 London Pneumatic Despatch installation, connecting the London Stock Exchange to the Electric Telegraph Company's central office on Lothbury, a distance of about 200 meters. The system handled telegraph messages, which arrived at the central office and needed to be physically delivered to traders at the Stock Exchange; the pneumatic tube cut delivery time from minutes (with messengers running through crowded streets) to under thirty seconds.

The commercial success of the London system inspired rapid replication. Berlin's system started in 1865 with a single tube; Paris started in 1866; New York opened its first commercial pneumatic tube in 1893. The decades from 1860 to 1900 saw essentially every major European and North American city install some form of pneumatic infrastructure, with the largest systems growing to hundreds of stations and hundreds of kilometers of tube over the next several decades.

The Berlin network at peak

The Berlin Rohrpost (pipe post) is the canonical example of what a mature pneumatic network looked like. By the 1930s, the network had grown to about 400 kilometers of tube connecting 90 stations across the city. The stations were typically located in post offices and major commercial buildings, with the larger stations capable of receiving and dispatching dozens of carriers per minute.

The operational characteristics: carrier speed averaged 35-40 kilometers per hour, meaning a message could traverse the city center in under fifteen minutes including switching time. Cost was modest (a Rohrpost stamp cost roughly the same as a regular letter, with delivery in minutes instead of hours), which made the service accessible to ordinary commercial use rather than only to government and financial customers. Peak traffic in the 1920s exceeded 30,000 carriers per day.

The institutional structure supporting the network was substantial. The Reichspost employed dedicated Rohrpost technicians, the equipment was custom-built by a small number of German firms, and the network was integrated with the broader postal system so that mail entering at any post office could be sent by Rohrpost to a destination post office for last-mile delivery by foot or horse.

The Paris double network

Paris ran two separate pneumatic networks. The Petite Pneumatique started in 1866 and served the central districts with short-message delivery; the Grande Pneumatique started in the 1870s and connected major post offices with longer-distance traffic. The combined network reached 467 kilometers at peak, the largest in the world, and was the canonical reference for European pneumatic engineering through the early 20th century.

The Paris system had a few innovations not seen elsewhere. The pneus (the carriers, which gave their name to "pneumatique" and ultimately to French slang for any quickly-sent message) were small lead canisters with screw-on caps, more robust than the leather-jacketed carriers used elsewhere. The system handled telegrams as well as direct messages, and continued in service much longer than other European networks: the last segments of the Paris pneumatic network were not retired until 1984, more than 120 years after the system started.

The New York system and the Lamson chutes

The New York pneumatic post operated from 1897 to 1953, connecting 23 post offices across Manhattan with 43 kilometers of tube. The system moved up to 95,000 letters per day at peak, with carriers reaching speeds of 50 kilometers per hour through the larger 8-inch tubes that distinguished American systems from the smaller European tubes.

The American story diverges from the European in one important way: the pneumatic system in the United States was supplemented by a much larger network of department-store pneumatic systems supplied by the Lamson Engineering Company, which dominated the industrial pneumatic market from the 1890s through the 1930s. Lamson tubes connected sales floors to cashier rooms, allowing a clerk to send the payment and receipt up to a centralized cashier and receive change back, a system that was the dominant retail point-of-sale architecture in American department stores for forty years.

The Lamson installations were not connected to the postal network, but the engineering vocabulary was similar and the manufacturer was the same. The total length of Lamson tube in American commercial buildings in the 1920s was almost certainly larger than the postal pneumatic network, though the figures are not well-documented because the installations were private.

The competitive landscape

The pneumatic tube competed with several alternatives. Horse-drawn dispatch was slower and more expensive but more flexible (no fixed infrastructure). The telegraph was faster for true point-to-point traffic but required a telegraph operator at each end to convert between language and Morse code. The telephone, after about 1880, was faster for short messages but required someone at the receiving end to take the call. The pneumatic tube fit the niche of "fast delivery of physical documents within a city," which mattered for original signatures, financial instruments, legal documents, and confidential traffic that could not be encoded into telegraph signals.

The competitive position was strong enough that pneumatic systems continued expanding through the 1920s, even as the telegraph and telephone matured. The eventual decline was driven not by direct displacement but by changes in what people were sending: the rise of mass-produced typed documents, the disappearance of the office boy as a profession (who would carry documents both via pneumatic and directly), and the shift to telephone communication for most short-message traffic that had previously gone by tube.

The decline

The pneumatic systems declined gradually from about 1930 onward, with most North American postal pneumatic networks shut down by 1950 and most European systems by 1990. The decline was not driven by any single competing technology but by a combination: cheaper and more reliable telephone service replaced the call-routing use, motor vehicles replaced the urgent-document use, and fax machines (starting in the 1960s and reaching ubiquity by the 1980s) replaced most of the surviving signed-document use.

The infrastructure was expensive to maintain. The tubes corroded, the leather carriers wore out, the air compressors needed constant attention, and the labor cost of operating switching stations did not scale down with declining traffic. Most networks reached a point where the per-message cost exceeded what customers were willing to pay, and the operators (usually national post offices) closed them.

What remained was hospital pneumatic systems, used for moving lab samples and medications within large hospital campuses, and a handful of industrial installations for specific high-value transport problems. The Czech city of Prague is the only major city that has reactivated parts of its pneumatic network in the 21st century, and only as a heritage and tourist demonstration rather than functional infrastructure.

What was lost

The engineering knowledge to build pneumatic networks at city scale is not gone, but it is no longer ordinary professional knowledge. The handful of engineers who maintained the surviving systems through the late 20th century are mostly retired or dead, and the equipment is no longer manufactured at scale. Rebuilding a Berlin-scale pneumatic network in 2026 would require reconstructing institutional capacity that took decades to develop and was lost in less than a generation after the systems closed.

The cultural memory is even thinner. Most current city residents in former pneumatic-network cities have no idea the systems existed, even when they walk through buildings that still have abandoned tube stations or pass over manhole covers labeled for tube access. The pattern is the same as for other lost urban infrastructure (gaslight, horse stables, ice harvesting): a generation-long forgetting follows the closure of any infrastructure that was once load-bearing.

Three observations

First, the pneumatic tube is a case study in technology that competed successfully against newer alternatives for decades, not because it was technically superior but because it filled a niche the alternatives did not fully cover. The lesson is that displacement is often slow and depends on complete coverage of the application surface, not just superiority on one dimension.

Second, the decline was about institutional sustainability rather than technical obsolescence. The systems could still work; they just could not pay for themselves once the use cases shifted to alternatives. The lesson is that infrastructure has a maintenance economics that is at least as important as its technical design.

Third, the loss of operational knowledge has been faster than the loss of the physical infrastructure. Parts of the Berlin and Paris networks still physically exist, but the human capacity to operate them is mostly gone. The lesson is that civilizational technical capacity is more fragile than the artifacts the capacity produces.

The deeper observation is that the modern city's communication and transport infrastructure looks much less like 1900 than 1900 looked like 1800. The pneumatic tube as a piece of urban memory is mostly a curiosity now, but it was a working system that moved hundreds of thousands of messages per day across great cities for the better part of a century. Almost no one alive has used one in its original commercial form, and the engineering tradition that built it is, for practical purposes, gone.