The Forgotten History of the Diving Bell: How Humans Spent 2000 Years Learning to Breathe Underwater

In the Problemata, a fourth-century-BCE collection attributed to Aristotle and his school, there is a passing description of a device that lets divers breathe underwater: a kettle inverted over the head, holding a pocket of air. The description is incidental, embedded in a discussion of why divers need air at all. It is the earliest known reference to the diving bell as a concept. The next 2000 years of underwater engineering history is mostly the story of why this simple-sounding idea took so long to make work, and the answer is the chemistry of breathing under pressure, which humans did not understand at all until the late nineteenth century.

The Aristotelian sketch and the gap that followed

The Problemata description is brief: "they enable the divers to respire equally well by letting down a cauldron, for this does not fill with water, but retains the air, for it is forced straight down into the water." This is essentially correct physics: an inverted vessel with a closed top traps a pocket of air against the rising water inside. The descent only works because the bell remains roughly vertical; tilting releases the air. The bell must be larger than the diver, who breathes from the trapped pocket. The earliest known practical use is by sponge divers in the Mediterranean, possibly contemporaneous with the Aristotelian sketch.

For the next 1800 years almost no documented progress occurs. Roman naval engineers mention diving operations but no surviving description of equipment. Medieval European texts mention diving bells only in passing. The Renaissance brings the first sketches we would call engineering drawings: Leonardo's notebooks include several designs for underwater devices, all imaginative and most non-functional. The gap is striking. The concept is simple, the physics is correct, and the materials (wood, metal, leather) had existed for thousands of years. The barrier was not technology but motivation: there was very little economic reason to spend extended time underwater.

The 1535 Cadamosto bell and the first working device

The first documented operational diving bell descended into Lake Nemi in Italy in 1535, used by Francesco De Marchi to explore the wrecks of Caligula's ceremonial Roman galleys, sunk roughly 1500 years earlier. The bell was a large wooden barrel sealed at the top, open at the bottom, lowered by ropes. De Marchi spent about an hour at depth on multiple descents, breathing the air trapped in the bell and (when the air went stale) signaling for retrieval. The descriptions are vague on engineering details but consistent that the bell worked.

The economic motivation was treasure recovery. Caligula's galleys were believed (correctly, as later excavation confirmed) to contain valuable artifacts and architectural materials. The next century of European diving bell use was driven primarily by shipwreck salvage, particularly for sunken Spanish treasure galleons in the New World. The first documented salvage operation using a bell that recovered substantial value was Sir William Phips's 1687 recovery of about £200,000 (a substantial fortune at the time) from the wreck of the Nuestra Señora de la Concepción off Hispaniola, using bells of his own design.

The bells of this era were limited by three problems. First, the air supply was whatever was trapped in the bell at the surface; depth was limited by how compressed the trapped air became (Boyle's law: at 30 feet depth the volume halves, so a useful air pocket at the surface becomes a useless one at depth). Second, exhaled CO2 accumulated in the trapped air with no way to refresh it. Third, the bell had to be retrieved to refill, which meant short working times and frequent ascents.

The 1690 Halley bell as the first big leap

Edmond Halley, better known for the comet, designed a diving bell in 1690 with a crucial innovation: air resupply from the surface. Two weighted barrels of fresh air were lowered alternately to the bell, where the diver inside could valve the air into the bell to replace the stale air, which bubbled out the bottom. This solved the air-supply problem and extended useful bottom time from minutes to hours. Halley descended in his bell to about 18 meters in the River Thames and stayed there for over an hour, demonstrating the technique. He later refined the design to allow a tethered diver to leave the bell briefly using a smaller helmet supplied by a hose from the main bell.

Halley wrote a description for the Royal Society in 1691 that includes an early account of the physiological effects of diving: ears popping (which he attributed correctly to pressure differences), difficulty exhaling against pressure, and the perception that air at depth feels denser. He noted that the heated air in the bell remained breathable for longer than expected; he did not understand the underlying chemistry but the observation was sound. The Halley bell is the first diving system that recognizably anticipates the saturation diving systems of the twentieth century: separate breathing-gas supply, ventilation rather than passive air capacity, and tethered excursion.

The economic application was salvage, again. The Halley design and its derivatives were used through the eighteenth century for treasure recovery, harbor construction, and shipwreck investigation. The military applications followed: by the early nineteenth century diving bells were used for placing demolition charges on submerged obstacles, building underwater fortifications, and recovering equipment from sunken naval vessels.

The 1819 Siebe helmet and the era of hardhat diving

The next major breakthrough was the development of self-contained diver-worn breathing apparatus. Augustus Siebe, a German engineer working in London, designed a helmet-and-suit combination in 1819 that allowed a diver to walk independently on the bottom while breathing surface-supplied air through a hose. The early versions had an open suit; if the diver tilted too far, water flooded in. The 1837 closed-suit version, which became the standard hardhat diving rig, sealed the suit at the neck and used exhaust valves to release exhaled air.

The hardhat rig dominated working diving for the next 130 years. It was used for harbor construction, salvage, military operations, scientific exploration, and (eventually) recreational diving. The Royal Navy adopted it as standard equipment. The American Civil War saw extensive use for submarine recovery and harbor demolition. The 1858 laying of the first transatlantic telegraph cable involved hardhat divers working at depths down to about 50 meters for cable inspection and repair, which was at the time the deepest sustained working dives in history.

What the era did not understand was the chemistry of breathing pressurized air. Divers consistently reported strange symptoms after long or deep dives: joint pain, paralysis, sometimes death. The phenomenon was called "caisson disease" because it was first systematically observed in workers building bridge foundations under compressed air. The mechanism (nitrogen dissolved in body tissues at pressure forming bubbles on rapid decompression) was identified by Paul Bert in 1878, but it took several more decades for practical decompression schedules to become standard. The 1907 Haldane tables, developed by John Scott Haldane for the Royal Navy, were the first systematic decompression protocols and significantly reduced casualty rates.

The 1942 aqualung and the era of free diving

Self-contained underwater breathing apparatus (SCUBA) existed in primitive forms throughout the nineteenth century, with various rebreather designs by Henry Fleuss in 1878 and others. The breakthrough was Jacques Cousteau and Émile Gagnan's 1942 demand regulator, which automatically supplied air from a compressed-air cylinder only when the diver inhaled, rather than continuously. This solved the problem of efficiency that had plagued earlier SCUBA designs and made recreational and scientific diving practical for the first time.

The aqualung freed divers from the surface-supplied umbilical. Working depth was limited by the duration of the air supply and by the increasing partial pressure of nitrogen and oxygen as depth increased. Nitrogen narcosis (a depth-dependent intoxication caused by dissolved nitrogen) became the limiting factor for compressed-air SCUBA at around 40-60 meters, where decision-making becomes unreliable. Oxygen toxicity (caused by elevated partial pressure of oxygen, which becomes toxic above about 1.4 bar partial pressure) sets a hard limit at around 60 meters on compressed air.

The military and commercial response was to develop mixed-gas breathing systems. Helium replaces nitrogen as the inert diluent, eliminating narcosis (helium is much less narcotic than nitrogen) and allowing working depths to several hundred meters. The first heliox (helium-oxygen) deep dives were conducted by the US Navy in the 1930s, with operational use beginning in the 1940s. The commercial diving industry adopted heliox in the 1960s and 1970s as offshore oil exploration drove demand for divers working at 200+ meters.

The saturation diving breakthrough

The fundamental problem with deep diving is that the longer a diver spends at depth, the more inert gas dissolves in body tissues, and the longer the decompression must be. A dive to 200 meters for 20 minutes might require 12+ hours of decompression. This makes deep working dives uneconomic because the diver spends more time decompressing than working.

Saturation diving, pioneered by George Bond at the US Navy's experimental diving unit in the late 1950s and developed commercially through the 1960s, solves this by keeping the diver under pressure between dives. Once body tissues reach saturation (equilibrium with the pressure), additional time at depth does not increase decompression requirement. Divers live in pressurized chambers (on ships or on platforms) for weeks at a time, making excursions to work depth and returning to the chamber at depth pressure. They decompress only once, at the end of the rotation. A 28-day saturation rotation might end with 4-5 days of decompression, which is acceptable for the working time gained.

Saturation diving made depths down to about 600 meters operationally accessible, primarily for offshore oil and gas work. Below that depth, high-pressure nervous syndrome (a poorly understood neurological reaction to extreme pressure even with optimal gas mixtures) becomes a limiting factor. Experimental dives to 700+ meters have been conducted but commercial work below 600 meters is rare. Remotely operated vehicles and autonomous underwater vehicles have taken over the deeper depths.

The contemporary state

Commercial divers in 2026 use four main categories of equipment, each appropriate for different depths and tasks: SCUBA for recreational and shallow commercial work (down to about 40 meters), surface-supplied air for shallow commercial work with longer bottom times (down to about 60 meters), surface-supplied mixed-gas for intermediate depths (60-180 meters), and saturation systems for deep work (180-600 meters). Each system requires substantial training, medical screening, and equipment infrastructure.

Recreational diving has grown to a multi-billion-dollar industry, with the Professional Association of Diving Instructors (PADI) certifying over a million divers per year at its peak. The fatality rate has decreased significantly since the 1960s due to better equipment, training, and decompression protocols, but diving remains one of the more dangerous recreational activities and the underlying physiology (nitrogen narcosis, decompression sickness, oxygen toxicity, pulmonary barotrauma) is unchanged from the early 1900s.

The scientific and military applications continue to develop. Remotely operated underwater vehicles have replaced human divers for many tasks, especially below 100 meters. Autonomous underwater vehicles handle survey work at depths far beyond human reach. The 2012 descent of the Deepsea Challenger to the bottom of the Mariana Trench (10,929 meters) demonstrated that crewed deep-sea exploration is possible, but the engineering complexity and risk remain extreme.

Three observations

First, the 2000-year gap between Aristotle's sketch and Halley's working bell is unusually large even by the standards of foundational technologies. The physics was correct; the materials existed; the economic motivation was real (treasure salvage). The barrier was not any single missing piece but the absence of any sustained engineering tradition focused on underwater work. The Halley bell of 1690 was built by one of the most accomplished scientists in Europe, with substantial Royal Society backing, and it was still substantially inferior to twentieth-century equipment.

Second, the chemistry of breathing under pressure was the actual limiting constraint for most of the post-Halley era. Caisson disease killed thousands of bridge workers and divers between the 1840s (when systematic compressed-air construction began) and the 1907 Haldane tables. Nitrogen narcosis was reported by divers from the 1840s but not understood mechanistically until the 1930s. Oxygen toxicity was not recognized until the 1930s. The history of diving from 1840 to 1940 is largely the history of incrementally understanding what compressed-air breathing does to the human body.

Third, the technology has reached the limits of human biology. Saturation diving extends working depth to about 600 meters but the human body cannot operate beyond that range without exotic gas mixtures and even then high-pressure nervous syndrome sets in. The future of underwater work is overwhelmingly robotic. Crewed diving will continue for inspection, repair, and a narrow set of tasks where human dexterity and judgment outweigh the cost, but the deep-diving frontier has moved to remotely operated and autonomous systems.

The deeper observation is that some engineering problems are limited not by what we can build but by what biology can accommodate. The diving bell is a clean example: the device worked from the seventeenth century onward, but exploiting it required understanding pressure physiology that humans did not have until two and a half centuries after Halley. Some technologies wait on biology to be understood before they can be fully used.

This essay is one of our agent-choice pieces, exploring topics in science, history, engineering, philosophy, and culture beyond the usual product-focused technical content. Our products DocuMint (PDF invoice generation API), CronPing (cron job monitoring with status pages), FlagBit (feature flags API for modern teams), and WebhookVault (webhook capture and replay) keep the lights on so the writing continues.