Guglielmo’s Secret: The Enigma of the First Diving Bell Used in Underwater Archaeology

Abstract

The earliest employment of a breathing apparatus in underwater archae­ology took place in July 1535, when Guglielmo de Lorena and Francesco de Marchi explored a Roman vessel sunk in Lake Nemi near Rome using a one-person diving bell that Guglielmo had invented. Francesco’s book on military architecture contains first-hand documentation of the exploration and of the diving bell; however, the author refrained from describing the bell’s most intriguing feature, the air-supply mechanism. The diving bell was very small, so that the air it contained could support a diver for only a few minutes; however it had a mechanism that expelled breathed air while maintaining sufficient air pressure inside to prevent the water level from rising and this mechanism allowed the diver to work submerged for hours. Francesco did not describe this mechanism because he had taken an oath to keep Guglielmo’s invention a secret. This article analyses Francesco’s text to unveil this secret.

Keywords:

• Lake Nemi
• diving
• Roman barges

Introduction

Two Roman barges lay for many centuries at the bottom of Lake Nemi, a small lake in a volcanic crater some 30 km south of Rome. Each barge was some 70 m long and 20 m wide at the beam; they served as platforms for floating palaces, built for Emperor Caligula in the first century ad. The palaces were in ruins but the wooden hulls were virtually intact, presumably indicating intentional destruction and scuttling. These barges were the objective of the first two instances of underwater archaeology on record.

In 1446, Cardinal Prospero Colona, driven by Humanist interest in anything Roman, commissioned renaissance scholar Leon Battista Alberti to explore the ships. Genovese seamen who ‘swam like fish’ attached ropes to one of the ships and Alberti attempted to raise it by means of cranes built on floating platforms.¹ He obviously did not realize how large and heavy the ship was, so all he managed to do was to break off and bring to the surface a piece of the stem and some planks from the bow. Incidentally, this was the first damage inflicted to the ships by would-be explorers, later explorations, particularly in 1827 and 1895, caused substantial harm.² Alberti mentions this exploit very briefly in his Ten Books on Architecture;³ he described it in more detail in another book, named Navis, which is now lost.

In 1535, Guglielmo de Lorena and Francesco de Marchi used a diving bell invented by the former to explore the same ship from end to end; they collected and brought to the surface pieces and items they found on the ship, they observed many details of the construction and they even measured the length, beam and depth of the hull. Francesco de Marchi has left a detailed account of the exploration and of the findings in his book on military architecture.⁴ He has also left a detailed description of the diving bell, which starts with a statement that it had a mechanism that expelled breathed air and prevented the water level from rising, but he had taken an oath to keep it a secret.⁵

Alberti deserves credit for performing the first act of aquatic archaeology ever. His was not the first attempt to salvage a sunken ship but it was the first undertaken for the sole purpose of learning about it. Guglielmo de Lorena and Francesco de Marchi deserve credit for being the first to practice archaeology underwater using a breathing apparatus. Diving bells had been used before for military purposes and for salvage; even Alexander the Great allegedly used one in the siege of Tyre. Early Modern authors, including Leonardo da Vinci, describe various devices for air supply to submerged divers, mainly masks or helmets with long snorkels, but this was the first actual use of an underwater breathing apparatus in aquatic archaeology.

In 1927–32, Italian archaeologists, with the support of Mussolini’s government, drained the lake and exposed the two vessels, removed piles of construction debris from the bilges, extracted the hulls from the mud and transferred them to a museum built nearby. Unfortunately the hulls were destroyed by World War II artillery fire in 1944, but not before detailed documentation and measurements of the ships were made, published in Le Navi di Nemi by Guido Ucelli, the archaeologist who was in charge of the entire operation.⁶ Now that we have those data the archaeological findings of the early explorations are not very interesting, but the design of Gugliermo de Lorena’s diving bell is intriguing.

The diving bell

Francesco de Marchi made several dives to the ship and his companion made several more. He describes explicitly the first two dives he made on 15 July 1535; one lasted for half an hour and the second for about an hour. It is evident from his description of the ship, from the detailed observations of its construction, from the way they measured it and from the two mule-loads of material they brought to the surface that they made multiple dives with a cumulative duration of many hours. De Marchi writes that typical diving time was between one and two hours, the limiting factors being cold and fatigue, not shortage of air. Apparently, there was an ample supply of air in the diving bell but de Marchi does not say how that was accomplished. On the other hand, the compact size of the bell, according to the dimensions he does provide, proves that the air it contained was enough for only a few minutes. This is the enigm­a of de Lorena’s diving bell.

De Lorena’s invention was a small personal diving bell or perhaps a large diving helmet, suspended by rope from a boat or floating platform, supported on the diver’s shoulders and reaching only down to the middle of his upper arms above the elbows (Figure 1). The diver could walk around on the bottom with the bell on his shoulders and his hands and arms were free to do work, but he had to be very careful to keep the bell in an upright position at all times. Walking and working while balancing the heavy and clumsy contraption on the shoulders must have been quite difficult and time-consuming but it was obviously feasible. The divers walked all over the ship, which was about 70 m long, but refrained from going below deck for fear of losing their balance. They used strings to take measurements of the ship and collected two mule-loads of timber, nails, lead pipe, bricks and other items they found lying around. The total amount of diving time and work required for doing all that must have been extensive; this was certainly not a quick-look dive. According to de Marchi, Guglielmo de Lorena had used his diving bell to find a sunken galley near Civita­vecchia and to salvage its artillery, which must have involved a significant amount of walking and working underwater.

FIGURE 1. Gugliermo de Lorena’s diving bell.

This is how Francesco de Marchi describes the diving bell:

The oath I have taken prevents me from explaining the mechanism by which the exhaled air exited the instrument and water could not get in. This instrument was a round barrel made of oak two fingers [about 4.5 cm] thick. It was five palms [about 1.25 m] long and three palms [about 75 cm] wide, with a securely fastened bottom, which, when in the water, remained at the top. Six iron hoops held the barrel together with another hoop made of lead more than two fingers thick installed around the open end to make the vessel sink better. The outside was caulked and greased like a ship to make the vessel watertight. The person inside could look out through a thick piece of crystal, one palm long and half a palm wide set in the sidewall. Two strips of iron embraced the shoulders of the person inside so that the head would not reach the bottom [i.e. top]. A girth attached to these strips went down behind the back and between the thighs; it was attached to the sidewall in front by means of a clasp, which could be opened very easily […] The instrument did not extend below the middle of the upper arms, so that it was possible to work [with the hands] but almost by groping due to the lack of sufficient light […] Out of the water, this instrument was too heavy for a single man to carry; underwater it seemed to weigh not more than 40 lb. due to the air trapped inside […] This instrument was useful for staying one or two hours under water […] It was possible to work under water, to saw, cut, tie ropes, operate hammers and other tools, but without much force due to the [resistance of the] water […] From the waist up, it felt as if one was in a hot oven but from the elbows down one felt a great cold […] The breath [i.e. exhaled air] exited the instrument and water did not enter but there was no tube or pipe for connection with the air out of the water.

According to the dimensions in Francesco’s description, the internal volume of the diving bell was 0.41 cubic metres. The upper part of the diver’s body took up a small fraction of this volume so a volume of 0.4 cubic metres (400 litres) is a good-enough approximation for our estimates.

The oxygen intake increases linearly with the intensity of physical exercise; actual consumption by each individual depends on his or her fitness and body size. The oxygen consumption of a ‘booted’ diver, i.e. a diver wearing a helmet and heavy boots, walking on a hard bottom with no mud, has been measured at 1.5 litres/minute STPD (Standard Temperature and Pressure, Dry) that is, normalized to atmospheric pressure, 0° C, with no humidity.⁷ Working underwater while balancing the heavy diving bell on one’s shoulders was surely at least as strenuous as the task of the booted diver; a similar consumption is a reasonable, perhaps even a conservative assumption.

Endurance, or the length of time a diver could stay submerged in the bell, depends on whether some ventilation device existed, through which exhaled air could exit the bell. With such a device, the concentrations of oxygen and carbon dioxide in the inhaled air would be constant but the volume of air in the bell would decrease and the water level would rise with every breath. When the water level in the bell rose to the level of the diver’s mouth, he would no longer be able to breath. The volume of air in the bell from the initial water level to the level of the diver’s mouth, divided by the volume he exhaled per minute would set the absolute upper limit on endurance. If there were no ventilation device, then when the diver exhaled into the bell and air did not get out then the volume would remain constant but the concentration of oxygen would decrease and that of carbon dioxide would increase; these concentrations would then set the limits on endurance. When either one of these concentrations reached a certain threshold, the diver, especially one engaged in a physically strenuous activity, would experience increased breathing rate, accelerated heart beat, impaired coordination, chocking sensation and other adverse symptoms.⁷

At an exertion level that requires oxygen consumption of 1.5 litres/minute in STPD conditions, the Respiratory Minute Volume (RMV), namel­y the volume of air breathed-in per minute, is around 35 litres.⁸ At this rate, the initial air volume of 400 litres in the bell could apparently suffice for 12 minutes. However, the volume of air inside the bell decreases linearly with increasing pressure. At a depth of 10 m, the pressure is two atmospheres, absolute, and the volume of the air in the bell decreases to half its volume at atmospheric pressure. This means that the water level inside would rise to half the height of the bell, 61 cm above the rim. We know from de Marchi that the lower rim of the bell was above the elbows, to allow use of the hands during the dive. According to anthropometric data the distance from a spot half the length of the upper arm above the elbow, where the lower rim of the bell was, to eye level is about 0.215 the height of the individual.⁹ Variation of this ratio with the height are negligible, it is 0.22 for a person 1.5 m tall and 0.213 for one 1.95 m tall. For a man 1.7 m tall, which was the average height at that time, this translates into 36 cm. Simply put, if the water level in the bell rose by 61 cm to half the height of the bell it was above the height of diver’s nose and he drowned. When the Italians drained the lake to expose the ships, the first one broke the surface as the level dropped by 5 m and the bottom was exposed when the water level dropped by 12 m; the depth of the second ship was a between 15 and 22 m.¹⁰ Assuming that the divers explored the first ship, which was also closer to the shore, they had to descend to at least 10 m, which the rising water level inside the bell would not allow. Even at smaller depths, say 5 m, the diver’s mouth and nose would be below water level within two minutes.

This simple analysis shows that the diver could not just breathe the air trapped inside the bell, there had to be a way to remove exhaled air and replace it with fresh air, which is in line with De Marchi’s reference to a mechanism ‘by which exhaled air could exit the instrument’. The analysis shows more than that; it shows that air had to be somehow injected into the bell as the depth increased, just to keep the water level from rising and drowning the diver; 400 litres of air (STPD) had to be injected before the diver could reach a depth of 10 m.

3. The air supply mechanism

Francesco states that the secret mechanism kept the water level from rising and that they were able to dive for two hours. To achieve that, the mechanism had to be capable of continuous or frequent supply of fresh air to the submerged diving bell while removing the exhaled air. The newly supplied air would keep the volume of air, and the water level, inside the bell more or less constant; it would also provide the diver with sufficient breathing air for as long as he could stay under water before cold and fatigue forced him to the surface.

Thus, we know what Guglielmo’s secret mechanism did but we do not know how it worked, but we can use the few hints that his companion’s text does provide to put together a plausible assumptions about the nature of that mechanism. There are actually two mechanisms to consider, one for removing exhaled air and the other for replenishing the bell with fresh air.

Francesco writes that the exhaled air (the ‘breath’) exited the diving bell. This means that the diver inhaled compressed air from the diving bell and exhaled it into a device that allowed it to get out to the surrounding water. That could be simple enough; the diver could inhale through his nose and exhale through his mouth into a tube going through a hole in the roof of the vessel, sealed all around to keep the vessel airtight. The pressure in the tube would be the same as the pressure inside the bell and in the diver’s lungs; the outlet of the tube would be higher and therefore the pressure there would be lower. With the diver’s lips serving as a valve, this tube could let the exhaled air out into the water, working like a snorkel in reverse. A valve, probably also in the roof of the bell, that the diver could open to let out air, is another possible solution. Figure 2 shows a notional example of such a valve.

FIGURE 2. Notional design of a ventilation valve, closed (left) and open (right).

The air pressure inside the bell was equal to the water pressure at the depth of the water level inside and therefore higher than the pressure of the water on its top. This difference in pressure kept the valve shut (left figure). The diver could pull the rod down to open the valve, let some air out and then close it again. The air ejected from the bell would be mostly exhaled air, which was warmer than the rest of the air in the bell and therefore at its top.

Which of the two solutions did Guglielmo use? His companion gives us an answer, albeit in a roundabout way. In one of his dives, Francesco de Marchi took some bread and cheese into the bell. Perhaps it was lunchtime or perhaps, as a contemporary wit said, as a native of Bologna, de Marchi could not conceive being, even for a short while, where food was inaccessible. In his own words:

I brought with me, to eat, four ounces of bread and one of cheese. The bread, being black and not fresh, crumbled and many fish rushed at me to eat the crumbs. The worst was that, since I did not wear pants, they bit me in that part that anyone can imagine.

A person cannot eat bread and cheese with a tube attached to his mouth; therefore, this little anecdote precludes the possibility of an exhalation tube. However, one wonders why de Marchi decided to share this piquant but embarrassing episode with posterity. Furthermore, he writes that looking through the glass window implanted in the sidewall of the bell, the fish seemed to be as long as his arm, whereas in reality they were the size of his little finger. The glass could have acted as a lens, but it is hard to believe that he could explore the ship from end to end, collect so many items and observe the details of its construction while looking through a lens that magnified observed objects more than tenfold. The episode of the fish and bread should probably be taken with a grain of salt.

Francesco de Marchi was quite obsessive about keeping the secret of his companion’s invention. After somebody stole from his house the findings he had brought up from the bottom of the lake, Francesco wrote that the theft was really an act of industrial espionage; the thief was only trying to discover the secret mechanism. It is conceivable that he wrote the episode only to disguise his secret, in which case the mechanism was an exhalation tube. On the other hand, if we accept the episode as true, the mechanism had to be some sort of valve.

Either way, the mechanism could get rid of the exhaled air but it would cause the water level inside the bell to rise, since the quantity of air inside the bell would decrease while the pressure remained the same. Francesco was aware of the linkage between ejected air and rising water level; he says explicitly that the secret mechanism allowed air to exit without letting more water in. The only way to do that is to bring in more air from outside.

Air can be pumped into a submerged vessel via a snorkel or it can be transported down in some containers. More than a century and a half after Guglielmo’s time, Denis Papin invented the first method. Others, including Leonardo da Vinci and Battista della Vella, had ‘invented’ the snorkel earlier; even Aristotle mentions it, but using lung suction alone their snorkels would be useless even at a depth of 1 m. In 1689, Papin suggested pumping air into a submerged diving bell with bellows and a flexible tube. His method failed because the bellows of his time were not powerful enough. Even if bellows could be powerful enough to drive air into Gugliermo’s bell, Francesco dismisses this possibility by stating explicitly that there was no tube going from the bell to the surface.

Edmund Halley implemented the second method a quarter of a century later, in 1714. Like Papin, he did not invent the method; Aristotle describes large cauldrons lowered into the water vertically, with their bottoms up, so divers could periodically swim to one of the cauldrons, insert their heads inside, take a breath and swim back to look for pearls or sponges.¹¹ Halley resupplied a submerged diving bell with air carried in casks, open at the bottom like small diving bells and lowered from the surface. When a cask reached a depth below the diving bell, the pressure inside was higher than the pressure in the bell and the air rushed into the bell through a hose. With two alternating casks, one going up while the other went down, Halley had a continuous supply of compressed fresh air that allowed diving for long periods of time. Additionally, with supply from many casks, the pressure in the bell was high enough to keep the water level inside the bell close to its bottom rim. All the while, a valve in the roof of the bell let out hot, breathed air.

Halley achieved exactly what Francesco de Marchi had insinuated: a mechanism that let exhaled air out of the bell and kept the water level inside from rising, without a tube to the surface. Is it possible that Guglielmo de Lorena invented Halley’s method nearly two hundred years before Halley?

Unless we suggest a viable alternative to Halley’s method for replenishing the air in a submerged diving bell and keeping the water level down without a snorkel of some sort, we must admit that this is the only possible solution to the enigma of Gugliermo de Lorena’s diving bell.

Only circumstantial evidence supports the argument that Gugliermo used the metho­d Halley would invent some two hundred years later. However, there was nothing in Halley’s invention that de Lorena could not accomplish with the technology available to him, given the same design idea. The idea is the key. The hypothesis assumes that two engineers who faced the same problem and possessed the same technology came up with the same design idea, which is not a far-fetched assumption at all. Furthermore, they both could have taken the idea from Aristotle.

The facts are:

a.	There had to be a way to supply fresh air to the diver, otherwise he would be able to stay submerged for only a few minutes at a shallow depth whereas Francesco de Marchi writes, from personal experience, about dives of one to two hours at depths from 5 to 12 m.
b.	Francesco unequivocally excludes a snorkel solution.
c.	An undisclosed mechanism existed; it was capable of expelling breathed air and maintaining the pressure and volume of the air in the bell constant in spite of the expelled air.
d.	Halley’s method was the only mechanism for replenishing the air in the bell and for maintaining the pressure without a snorkel.
e.	Gugliermo de Lorena had the technology required to implement Halley’s method.

Considering these facts, the hypothesis that de Lorena used Halley’s method is more than just plausible.

By this hypothesis, de Lorena used two diving bells, not one. De Marchi describes the first; his oath prevented him from writing about the second. As shown in Figure 3, the diver donned the main bell, actually more a helmet than a diving bell, as de Marchi describes. The second bell, an inverted cask with an open bottom, a hoop of lead around the open end and a plugged tube sticking out of its top lid, travelled periodically up and down, into the water and back to the free air above. As the cask descended, the air pressure inside increased; when it reached below the diving bell the pressure in the cask was higher than the pressure in the bell. All the diver had to do was to grab the cask, pull it so that the tube was inside the bell and remove the plug to let the compressed air from the cask into the bell. Then, the assistants in the boat above would pull the cask out of the water. By repeating this procedure, the diver had a continuous supply of fresh air and could stay submerged as long as the cold water and his fatigue allowed. The diver could expel the breathed air by periodically operating a valve or by exhaling into a tube, like a modern snorkel in reverse.

FIGURE 3. Mechnism of air supply.

This mechanism contains some hypotheses within a hypothesis. Gugliermo could use containers made of animal skins instead of a cask, but those would have to be inflated, probably by means of a bellows. A cask was simpler and, having devised the diving bell, he surely knew how to make and use a cask. Halley used a flexible hose longer than the cask with lead weights at its end, not a plugged tube, to transfer air from the cask to the bell. The plugged tube is simpler, it works just as well and for a diver who can move around and use his hands and forearms in the water rather than being inside the bell like Halley’s divers, it is easier to manipulate. Halley had to use an exhaust valve because his bell was large and usually held more than one diver. Gugliermo’s device served a single diver and therefore both a valve and a reverse snorkel were viable solutions. The snorkel was simpler to implement but the valve was more convenient to operate; and, if we grant any credibility to Francesco’s bread and cheese episode, we must conclude that Gugliermo used a valve.

Each ventilation solution required a different mode of operation. With the snorkel, the diver would release air continuously and the water level in the bell would rise until the next cask arrived with a new load of air. With a valve, the diver could breathe normally and open the valve to ventilate the bell only when a new cask arrived. For an RMV of 35 litres per minute STPD, as used above, and assuming an air cask with a volume of 100 litres, the assistant in the boat would have to raise and lower the cask once every three minutes, which is quite easy to do.

Conclusion

If we accept Francesco de Marchi’s text as factual, and we have no reason to believe otherwise, some version of Halley’s invention is a plausible solution to the enigma of Guglielmo de Lorena’s diving bell. In my opinion, this is the only plausible solution but it remains a hypothesis nonetheless.

Guglielmo de Lorena and Francesco de Marchi certainly deserve credit for being the first underwater archaeologists to use a diving apparatus and de Lorena probably deserves the credit for inventing Halley’s method two hundred years before the recog­nized inventor. To close the loop, another amateur archaeologist, Annesio Fusconi, explored the ships of Lake Nemi in a diving bell almost three hundred years later, in 1827. Fusconi used a Halley diving bell.

Notes on contributor

Joseph Eliav, after more than forty-five years’ experience of R&D, system engineering and technical management, mostly in aerospace and defence systems, has returned to university in Israel to study for higher degrees. He has published papers on sixteenth- and seventeenth-century galleys and is currently a reseacher at the Leon Recanati Institute of Maritime Studies, University of Haifa.

Notes

Regular readers of this Journal will recall Kenneth A. Kroos’ paper on ‘Central Heating for Caligula’s Pleasure Ship’, in Int. J. for the History of Eng. & Tech., 81.2 (2011), 291–99. This mentions the archaeological ‘digs’ and the destruction of the hulls in the Second World War.

1 Flavio Biondo, Roma ristaurata et Italia illustrata (Venice: Domenico Giglio, 1558), pp. 110–11.

2 G. C. Speziale, ‘The Roman Galleys in the Lake Nemi’, Mariner’s Mirror, 15.4 (1929), 333–46.

3 Leon Battista Alberti, I dieci libri de L’ Architettura, (Venice: Vicenzo Vagaris, 1546), Book V, Chapter XII.

4 Francesco de Marchi, Architettura militare (Rome: Romanis, 1810), pp. 356–66.

5 Ibid., pp. 370–74.

6 Guido Ucelli, Le Navi di Nemi (Instituto Poligrafico e Zecca dello Stato, 1950).

7 Respiratory Protection Standard 29CFR 1910.134 (Occupational Safety and Health Administration), para. d.2.

8 Anon, Diving Physiology (National Oceanic and Atmospheric Administration, US Department of Commerce), figs 3–10.

9 Anon., Human Integration Design Handbook, NASA/SP-2010-3407, Appendix B: Examples for Anthropometry, Biomechanics and Strength.

10 Speziale, p. 338.

11 Robert F.Marx, The History of Underwater Exploration (Dover, 1990), pp. 30–35.