A Table Laid on the Sea Floor
The thing was found by men who were looking for something else. In April a survey party working the Middle Deep off the Isle of Sheppey — grappling for a lost mooring chain, of all the unromantic errands — brought up a length of tarred leather hose with a bronze union at one end, and three casts later an oak stave eleven inches across, hooped, pitched, and pierced by a small arched window with the glass still in it. The stave was the wrong curve for a barrel. It was the wrong curve for anything that floats. It was bent the way a bell is bent, and somebody had gone to the trouble of putting a window in it.
What has come off the same eight acres of ground since is a service for fourteen. Pewter chargers. Gilt goblets, four of them standing upright in the silt exactly where they had been set down. Oyster shells in a quantity no wreck accounts for, because a wreck scatters and this did not: the shells lie in a ring, as though the diners had thrown them outwards from a table and the table had never moved. There is no hull. There is no hull ballast. There is a great deal of lead, but it is lead in the wrong arrangement — laid in a circle nineteen feet across, as if to hold something down rather than to keep something upright.
The Bureau has spent the spring with the papers that go with the find, and prints the substance of them below. The claim they make is not that a ship went down with a dinner party aboard. The claim is that the dinner party was the point; that the sea floor was the venue; and that for something like seven or eight years at the end of the sixteenth century a fashion ran through the English court for taking the sweet course three fathoms under.
Readers who wish to object are asked to hold the objection for one section. The bell is not the improbable part. The bell is the easy part.
What Holds the Water Out
Turn a tumbler upside down and push it into a bowl of water. The water does not fill it. A pocket of air is trapped under the base, the sea cannot get past it, and if the tumbler is large enough and the pocket deep enough a man can put his head inside and breathe. That is the whole of the diving bell, and it has been understood for as long as anybody has bothered to write it down: Aristotle describes the trick in the fourth century before Christ, a cauldron let down into the sea so that divers may draw breath out of it. Everything since is refinement, and most of the refinement was kept quiet. Leonardo covered sheets with breathing gear for divers around 1508 and published none of it. The first apparatus of the sort actually put to work went down into Lake Nemi in 1535, where Guglielmo de Lorena and Francesco de Marchi used an early bell to go over Caligula’s sunken ships — and deliberately kept the mechanism of it secret. A working bell, in other words, genuinely does predate the events described here. It predates them by half a century, and it is in the wrong country.
What the tumbler does not show you is the price. Push it deeper and the water starts to climb inside, because air can be squeezed and seawater cannot. The rule is a plain proportion — halve the room the air has and it pushes back twice as hard, and read the other way about, double the pressure on it and it takes half the room. Thirty-three feet of seawater lies on you as heavily as the whole atmosphere does. So at eighteen feet, three fathoms, you are carrying about half an atmosphere more than you had on deck, and the air in an empty bell shrinks to roughly two parts in three: a bell nine feet in the wall goes down with three feet of sea standing inside it. At thirty feet the load is very nearly doubled, the air is very nearly halved, and very nearly half the bell is water. At sixty feet you are down to about a third of your air, in a bell two parts flooded, and there is no dinner to be had in it.
A bell is therefore not a room. It is a bargain struck between height and depth, and the only way to win it is to put air back faster than the sea takes it. Every serious bell in history is an argument about how. Halley’s, a hundred years downstream of the events described here, did it with weighted casks sent down on lines and deliberately sunk below the bell, so that the greater pressure at the cask mouth drove the air uphill through a hose of its own accord; he claimed to keep men down at ten fathoms for an hour and a half by it, and he wrote up what the pressure did to his ears in the flat manner of a man who has decided to be scientific about his own suffering.
The Sheppey papers describe both methods and describe them a century early. The hose is leather, tarred, whipped at the splices, and fed from a pair of double bellows on a treading beam. The casks are oak, lead-hooped, and appear in the accounts under the heading the second service — which tells you precisely how the household paying for them had decided to think about air. It arrived between courses.
Pargiter of Deptford Strand
Every bell in the file was built by one man, and he was not a shipwright. Ambrose Pargiter kept a cooperage on Deptford Strand, a few hundred yards downriver of the yard Henry VIII laid out in 1513, and his trade was raising watertight vessels out of straight staves and iron hoops. A cooper does not build a hull. A cooper builds a curved wall that holds against pressure from one side, tightens the harder it is squeezed, and has no seam running the wrong way. Of every trade on that river, his was the only one already doing the job.
His accounts survive in a clerk’s copy and are the least romantic and most convincing part of the whole business. Oak by the load. Iron hoop by the hundredweight. Pitch, oakum, tallow, hemp, and lead at so much the pig — a quantity of lead that would embarrass a church roof, because the skirt of a bell has to be heavy enough to overcome the enormous buoyancy of the air trapped inside it. A nine-foot bell full of air wants to go up with the force of several tons, and everything Pargiter charged for was, in one way or another, an answer to that.
Then there is the line that dates him. Item, for glasse of the Crossed Fryers sorte, xij paines, bedded in lead. Clear glass fit to look through was not an English commodity for most of the century; it came in from Normandy and the Rhineland, and it came in green. The Venetian Giacomo Verzelini took over the glasshouse at Crutched Friars in 1575 and was granted a twenty-one-year monopoly on crystal glass made in the Venetian manner, and Pargiter’s first bell is dated twelve years after that. The whole conceit — a chamber you can see out of — waits on that patent. Before Verzelini a bell was a bucket you sat in the dark inside. After him it was a window.
The last thing to say for Pargiter is that he never once claimed to have invented anything. His covering letter to the Lord Chamberlain’s office describes the bell as a tunne sette upside downe, which is exactly what it was, and asks to be paid at the cooper’s rate.
The Launching
The bell goes out on a lighter with sheer legs stepped amidships and a windlass geared four to one, and it goes over the side empty, upright, and slowly, because a bell that is allowed to tip will spill its air in one belch and arrive on the bottom as a very expensive anchor. The papers are emphatic on this. Two guys fore and aft, four men on each, and the master calling the rate; the descent of a full bell from gunwale to bottom at three fathoms is set down as the space of a psalm.
The company did not go down with the bell. That is the detail that makes the rest of the arrangement credible, and it is the one the ballad versions always get wrong. The diners were rowed out and put aboard from underneath: the bell was set on its ring of lead, the water inside it allowed to find its level, and the guests brought down one at a time by a diver on a weighted line, ducking under the skirt and coming up inside into candlelight, on a ladder set against the wall. The Sheppey papers call this the entering, and give it four lines of instruction, three of which are about the hat.
A bell is dark, an evening dress is heavy, and the entering was, by every account in the file, the part nobody enjoyed and everybody afterwards described at length. One lady of the household is recorded as having been brought up again immediately and rowed ashore, and the clerk who wrote the entry has added, in a different ink, that she was right.
The Etiquette of Dining at Pressure
The great difficulty of a banquet at three fathoms is not drowning. It is precedence, and the Tudor court had opinions about precedence that no engineer was ever going to talk it out of.
In a hall, rank ran along the table from the salt. In a bell it ran vertically, because the floor of a bell is not a floor. It is the sea, standing at whatever height the air has been squeezed to, and it moves — up a hand’s breadth when the air cools, down again when the bellows are worked hard. Pargiter’s benches were therefore built in a stepped ring, and the seating plan in the file is a plan of heights above the water, in inches, with the names written against them. The place of honour is the driest. The gentleman at the bottom of the ring dined with his feet in the Thames estuary and knew exactly what the household thought of him.
Everything else about the meal is governed by two facts of the body, and both of them are real. The first is that the ears must be equalised on the way down, and nobody at the time knew why. The instruction as written is to gape, and swallow, and stop the nose and blow easily — which is, allowing for the spelling, precisely what a modern diver is taught. The second is that a man who eats and drinks at three fathoms as he would ashore will regret it, and that he does his regretting on the way up.
Half an atmosphere over is not enough to do a body any of the interesting harm the ballads have since given it. Nobody at that table was ever within reach of the bends: at eighteen feet, for forty minutes, there is nothing worth speaking of driven into a man, and so there is nothing to come out of him. What eighteen feet does is duller and entirely unavoidable. It squeezes every pocket of air he is already carrying — ears, sinuses, a bad tooth, the gut — flat on the way down, and hands every one of them back, larger, on the way up. The papers have a word for the consequence, and it is not a word about joints: the reckoning. It is placed in the ear, the jaw and the belly. It is described as beginning on the ladder and lasting the first quarter of an hour on deck, and the cure given for it is time, and walking about, and the open air. Forty minutes of oysters, jellies and small beer taken at half an atmosphere over yields a quantity of gas packed at half an atmosphere over, and on the way to the surface every bubble of it takes half as much room again as it had at table. The Bureau will only observe that the household blamed the shellfish; that this was nearer the truth than the household knew, the shellfish being the raw material if not the cause; and that the plain proportion accounting for the whole of it would not be printed anywhere for another seventy years.
Sealed things behave badly at both ends of the journey. A Tudor pie was baked in a thick standing crust called a coffin, and a coffin taken down whole arrives at the bottom cracked across the lid, because the air inside it has been squeezed to two thirds and the pastry is not flexible enough to follow. A leather bottle goes down flat and comes up drum-tight. Anything corked at the bottom must be opened before it is raised, and the file records one bottle that was not, and a steward who thereafter carried the scar.
And the voices go strange. Compressed air is denser air, and denser air rings differently in the mouth and the chest; the effect at three fathoms is very slight, but it is there, and it runs the wrong way from the squeak everyone expects. The company sounded, according to one guest, graver than they were.
The Bill of Fare
Here the file settles a question before it is asked, and the answer is better than the joke. In Tudor English a banquet is not the dinner. The banquet is the course that comes after it — the sugar course: marchpane, comfits, sucket, jellies, wafers, gingerbread, moulded conceits in hard sugar — and the whole fashion of the age was to take that course somewhere else, in a small room built for the occasion and for nothing else, on a roof, in a garden, at the end of a walk. Hardwick Hall, going up between 1590 and 1597, carries six banqueting houses on its tower tops. Seen against that, a chamber built for no purpose but the sweet course, reached by a short and slightly alarming journey, is not an aberration. It is the ordinary Elizabethan habit carried one step further than anybody had carried it before, in the one direction still unoccupied.
It should also be said that this was not a court easily embarrassed. Within living memory it had watched one king take six wives and take the whole country out of the Roman church, and it had spent two generations building on a scale — the Great Hall at Hampton Court went up around 1535, a hundred and six feet by forty, ninety-two to the gable — that treats a purpose-built room for one course as an entirely reasonable thing to commission. Against that record, a tub on the sea floor is a modest proposal.
The sea course that goes before it is, by the standards of the century, unremarkable. Oysters, which were the food of everybody in Tudor England and not a luxury at all. Lobster and crayfish. Whole fish baked and served cold. Oranges, pomegranates, dates and figs, all of them importable and all of them expensive. Wine of Gascony, sack from the Canaries, and Rhenish. What is remarkable is only that the whole service is cold: no fire is lit in a bell, because a fire in a sealed volume of air is a fire competing with the guests, and the household appears to have worked this out by the direct method.
Which brings us to the candles, and the candles are the most useful instrument anybody in this story ever had. A flame goes out well before a man is in danger, because it needs richer air than he does. Pargiter’s standing order is that the party rises when the second sconce fails — not the first, which may be a draught, and never the third, which nobody in the file ever waited for. He had, without any of the chemistry, built himself an alarm. It is the single most competent thing in the whole affair.
The Three Bells, As Rated in the Household Book
Ambrose Pargiter's own figures, converted where he gave them in fathoms and psalms.
| The Ewer, 1587 | The Great Panoramic, 1590 | The Cupola, 1594 | |
|---|---|---|---|
| Company at table | Six | Fourteen | Twenty, as designed |
| Height in the wall | 7 ft | 9 ft 6 in | 11 ft |
| Working depth | 2 fathoms | 3 fathoms | 5 fathoms |
| Sea standing inside | About 2 ft | About 3 ft 4 in | Very nearly half of it |
| Dwell at table | 22 minutes | 40 minutes | Never sat |
| Air | One hose, hand bellows | Twin bellows on the beam, and six casks | Twin bellows; casks refused by the master |
| Fate | Broken up 1594; hoops reused | Requisitioned for salvage 1596, lost off the Kentish flats | Left on the stocks; sold as staves |
The Queen’s Descent, Michaelmas 1591
The sitting the file is written around took place at Michaelmas 1591, in the Great Panoramic, on the flats inside the Swale. The Queen had turned fifty-eight three weeks before. Sir Walter Raleigh, who was then still in favour and had rather less than a year of it left, is named in the seating plan at her right hand, at a height above the water of thirty-one inches; the Queen sat at thirty-eight, on the top step, which is the highest place in the bell and, unavoidably, the one with the least headroom. The clerk notes that she stooped, and said nothing about it.
Raleigh is the one guest in the file who had any business being there. He had put his own money into ships, he had a hard practical interest in what could be got up off a wrecked hull, and the household book has him asking Pargiter two questions in the middle of the sweet course: what the bell would cost in quantity, and whether it could be worked in the open sea. Pargiter’s answers are recorded as eighty pounds, and not without a calmer bottom than we have here.
The second question was not idle. Every seaman at that table had grown up with the Mary Rose on the bottom of the Solent — Henry VIII’s ship, lost in 1545 within sight of the King, worked over by salvors who could not shift her and then left where she lay. She stayed there. It took until 1971 to find her again and until 11 October 1982 to lift her, with hydraulic jacks to break the seabed’s suction, a purpose-built cradle bolted through the hull, and a crane barge. A gentleman asking a Deptford cooper in 1591 whether his tub could be worked in open water was asking the most sensible question in the room.
Nothing else happened. That is worth saying plainly, because the ballads insist that something did. The company went down, ate a sugar course, watched the fish, said the things people say, and came up. The most exciting entry in the whole account is a charge for two candles more than were budgeted.
The Men on the Beam
The bell is the part everyone remembers and the part that matters least. A bell is a tub. What made the fashion possible for those seven or eight years was a leather hose and twelve men on a barge, and the twelve men are not named anywhere in the file.
They are counted, though, and paid, and the rate they worked at is set down because it had to be: the bellows were required to deliver a stroke every four seconds without a break, six men on the beam and six standing by, changed at the quarter. A stroke missed is not felt at once. That is the danger in the arrangement and the reason the tally-man exists — the bell has a pocket of air in it, so the failure of the pump is silent for several minutes and then abruptly is not.
The casks are the fallback, and they are the cleverest thing in the file, chiefly because they need nothing but arithmetic to work. A cask is filled with air on deck, stoppered, weighted with lead, and lowered past the bell until it hangs below the rim. Being deeper, it is under more pressure than the bell; open a hose between them and the air goes up into the bell without a pump, a man, or a prayer. Halley worked out the same thing a century later and is properly credited with it. Pargiter’s version is written in a cooper’s hand on the back of a bill for hoop iron.
What the Bureau notices, and prints without comment, is that the household book gives the name and the wage of the man who tuned the lute at the Michaelmas sitting, and does not give the name of any of the twelve men whose arms were the only reason anybody at that table was breathing.
What Came to the Glass
The entertainment, once you are down there and the sugar is finished, is the window, and the window worked far better than anybody had planned for. Light in the sea at night draws everything that swims, and a bell with a dozen candles in it is the brightest object on the estuary floor. The accounts describe silver shoals standing off the glass in a wall, a conger that came three nights running and was given a name, crabs that walked up the lead skirt and had to be knocked off with an oar because the guests found them upsetting, and something the clerk calls a jelly bell drifting past the port at eye level, which is a good description of a jellyfish written by somebody who had never seen one from the side.
The estuary is turbid and green, and the file is honest about it. The visibility is given at the length of a pike on the best evening and a hand before the face on the worst, and there is a note that on a bad night the household hung a lantern outside the port on a bracket, which made matters worse, because it lit the silt.
Nobody in this story ever saw a coral reef. What they saw was the bottom of the Thames Estuary in autumn, at close range, by candlelight, through six lines of Venetian glass. On the evidence of the accounts, it was more than enough.
The Night of the Fouled Hose
It went wrong once, in the way these things always go wrong, which is to say not dramatically and not where anybody was watching.
On an evening in the autumn of 1593 a hoy came down on the wind and drifted across the moorings, and her keel took the air hose under and pinched it against the barge’s own cable. Nothing happened. That is the important part. The bellows went on stroking against a closed pipe, the tally-man went on chalking, and below, seven people went on eating a sugar course in a pocket of air that had stopped being renewed.
They had between fifteen and twenty minutes of tolerable air in the bell and they used about eleven of them before the first sconce guttered. Pargiter’s order was for the second, and the household’s steward — whose name was Hugh Bellasis and who never worked for them again, for reasons that do not reflect well on the household — did not wait for it. He put the ladder into the water and started sending people up on the diver’s line at the first, one at a time, which is the only rate the arrangement allows and which took, by the clerk’s count, four minutes for seven.
Everybody came up. Two were bled, which will not have helped, and one was ill for a week. The Queen was not present; the party was a rehearsal for a sitting that never took place, which is the most Elizabethan detail in the entire file. The hoy’s master was fined and the fine was remitted on appeal, on the grounds — and the Bureau has read the judgment twice to be sure of it — that there was no law requiring a vessel under way to keep clear of a dinner.
Why the Fashion Stopped
It did not stop because of the fouled hose. Courts are not deterred by near misses; they are deterred by expense and by other people doing it. It stopped because of arithmetic, and the arithmetic was Pargiter’s own.
The Cupola was to have seated twenty at five fathoms, and Pargiter built it and then would not put company in it, and the reason he gives in his last letter is the only piece of applied science in the whole correspondence. At five fathoms the sea stands very nearly halfway up the wall, which halves the room. Twenty people is more than three times the company of the Ewer, and the pocket left to breathe is barely half of what the Cupola holds standing dry on the stocks: three times the company, in half a bell. And the bellows, he points out, do not get any better with depth — they get worse, because the deeper the bell the harder the air has to be pushed to enter it at all, and twelve men on a beam are twelve men on a beam. He puts the working dwell at not the fourth part of an houre, and that if all goe well, and declines the commission. The bell was sold for staves.
After that the money went elsewhere and so did the country. The war with Spain, which the Armada year of 1588 had opened rather than closed, was expensive and unglamorous by the middle of the nineties. Raleigh, the one man at that table with any real appetite for the thing, married without leave in 1592 and spent the following years first in the Tower and then in Guiana, and had no further use for a bell that could not be worked in open water. The Great Panoramic was taken up for salvage in 1596 and lost on the Kentish flats inside the year. Pargiter died in 1599, a cooper, and his inventory lists no bells at all.
The rest is chronology. A dynasty that had begun on a field near Market Bosworth in August 1485 had, by the time the Cupola was sold for staves, eight years left to run, and the fashions of its last decade did not outlive it.
What remains is the lead ring off Sheppey, nineteen feet across, and the oyster shells lying in a circle around a table that is no longer there. The Bureau has been asked several times this spring whether it believes any of it. Our position is that the physics is exactly right; that the accounting is exactly right; that the sea floor is precisely where a court which had run out of roofs to put its banqueting houses on would have gone next — and that we would very much like somebody to find the twelve names.
