#!/usr/bin/env python3 """ Generate a test EPUB that exercises edge cases in spine/TOC relationships. Edge cases covered: 1. Multiple TOC entries → single spine item (via fragment anchors) - frontmatter.xhtml: Dedication (#dedication), Epigraph (#epigraph), Foreword (#foreword) — three TOC entries, one spine item - chapter3.xhtml: Chapter 3 heading + three sub-sections (#the-reef, #the-current, #the-depths) — four TOC entries, one spine item - chapter5.xhtml: Chapter 5 heading + three sub-sections (#isle-of-echoes, #compass-rose-atoll, #whirlpool-narrows) — four TOC entries - appendix.xhtml: Appendix A (#appendix-a), B (#appendix-b), C (#appendix-c), D (#appendix-d), E (#appendix-e) — five TOC entries, one spine item (D and E are deliberately tiny) 2. Single TOC entry → multiple spine items (chapter spans files) - Chapter 2: chapter2_part1.xhtml + chapter2_part2.xhtml — one TOC entry, two spine items - Chapter 4: chapter4_part1.xhtml + chapter4_part2.xhtml + chapter4_part3.xhtml — one TOC entry, three spine items 3. Spine item with no TOC entry - interlude.xhtml: present in spine order, absent from TOC nav 4. TOC entry pointing to mid-file anchor (not file start) - backmatter.xhtml#colophon: the file starts with an Author's Note, but only the mid-file Colophon anchor appears in the TOC 5. Nested TOC hierarchy - Chapter 3 and Chapter 5 use nested

Front Matter

Dedication

For every reader whose bookmark landed on the wrong page, and every navigator whose chart placed the lighthouse on the wrong shore.

Epigraph

“A captain who trusts a single table of contents has never sailed past the edge of a spine.”

— Admiral Fragmenta, On the Perils of Pagination

Foreword

The document you hold in your hands — or rather, the document your e-reader is attempting to reassemble from its constituent parts — is deliberately, almost aggressively, tangled.

Some chapters sprawl across multiple files. Others cram several table-of-contents entries into a single page. At least one section appears in the reading order yet refuses to show up in the table of contents at all. And if you look carefully, you will find a table of contents entry that points not to the beginning of its file, but to a spot squarely in the middle.

This is all by design. If your reader survives, it can survive anything.

Chapter 1

Chapter 1
Setting Sail

Anchora had packed three things for the voyage: a sextant with a cracked lens, a notebook whose pages were already curling from salt air, and an unshakeable conviction that every archipelago deserved a proper atlas.

The sextant had belonged to her grandmother, a woman who had charted the entire Leeward Passage in an open dinghy using nothing but dead reckoning and a vocabulary of profanity that could strip barnacles at forty paces. The crack in the lens was the result of an encounter with a boom during an unexpected jibe — not Anchora’s grandmother’s boom, but rather the boom of a racing yacht that had cut across her bow in the harbor approaches at a speed the grandmother had described, with characteristic understatement, as “imprudent.”

The notebook was new, purchased from a stationer in the port town who specialized in paper for maritime use. It was bound in oilskin and stitched with waxed thread, and the pages were a heavy cream stock that the stationer claimed would resist salt spray, coffee spills, and the tears of frustrated navigators in roughly equal measure. Anchora had tested the first two claims. She hoped not to test the third.

“The trouble with charts,” she told the harbour-master as he untied her bowline, “is that they assume the world sits still long enough to be drawn.”

The harbour-master, whose name was Gideon and whose experience of the sea was limited to watching it from the end of the municipal pier, nodded politely and threw her the rope. He had heard variations of this sentiment from every cartographer, surveyor, and marine geologist who had passed through his harbour in thirty-one years of service, and he had learned that the most efficient response was agreement.

“Tides shift, sandbars migrate, and new rocks appear where no rocks have any business being,” Anchora continued, warming to her theme. She was coiling the bowline with the automatic precision of someone who has coiled ten thousand ropes. “By the time you publish an atlas, half the coastline has wandered off. You might as well try to draw a portrait of a cloud.”

“And yet you keep drawing them,” Gideon observed.

“And yet I keep drawing them,” Anchora agreed. She stowed the coiled rope in the lazarette and checked the halyards. The main was neatly furled on the boom; the jib was hanked on and ready. Everything was in order. Everything was always in order aboard the Pagination. Anchora tolerated chaos in the natural world because she had no choice, but she would not tolerate it aboard her own boat.

The Pagination was a twenty-eight-foot sloop of a design that had been fashionable forty years ago and was now merely practical, which suited Anchora considerably better. She had a long waterline, a shallow keel for poking into the kind of anchorages that deeper boats could only look at wistfully, and a rig that one person could handle in all but the most theatrical weather. Her hull was white, her boot-stripe was navy, and her name was painted across the transom in gold leaf that was beginning, after twelve seasons, to develop the kind of distinguished patina that a less generous observer might have called peeling.

Anchora had bought her for a sum that the previous owner had described as “a bargain” and that Anchora’s bank manager had described as “concerning.” That had been eight years ago. In the intervening time, the Pagination had carried Anchora to thirty-seven islands, through four named storms, across two time zones, and into one regrettable encounter with a shipping container that had fallen off a freighter in heavy weather and was drifting, unmarked, at exactly the height of a sloop’s waterline. The encounter had cost Anchora a starboard stanchion and three weeks in a boatyard. The container, so far as she knew, was still drifting.

She cast off the stern line and let the Pagination drift clear of the dock. The wind was from the northwest at eight knots, which was enough to move but not enough to excite. She unfurled the jib, sheeted it in, and felt the boat come alive beneath her feet — that small, particular pleasure of a hull finding its groove in the water.

The harbour fell away behind her: the stone breakwater, the red and green channel markers, the row of chandleries and fish restaurants that constituted the town’s entire economy. Gideon was still standing on the dock, growing smaller. He raised a hand. Anchora raised hers in return.

Beyond the breakwater, the sea opened up. It was mid-morning and the light was the flat, even grey of an overcast day — not dramatic, not beautiful, but excellent for navigation. Dramatic light made for good paintings but unreliable bearings. Anchora preferred her horizons unambiguous.

She unrolled the chart she had prepared the night before and pinned it to the table in the companionway. It was mostly blank. A few coastlines sketched from Admiralty data. A scattering of depth soundings copied from a survey that was either thirty or sixty years old, depending on which edition you believed. A compass rose in the corner, drawn with the care of someone who understood that a compass rose is both a tool and a promise.

The rest was empty space. White paper waiting for ink. Terra incognita, or rather aqua incognita, which sounded less impressive in Latin but was considerably more relevant to someone in a boat.

She spread a fresh sheet of vellum on the chart table, uncapped her pen, and wrote at the top in careful block letters: ATLAS OF THE UNCHARTED REACH.

Below this she added, in smaller letters: Compiled from original observations by A. Vellum, Master Cartographer, aboard the sloop Pagination. This was perhaps optimistic — the atlas currently contained no observations at all — but Anchora believed in stating one’s intentions clearly. A blank atlas was not an admission of ignorance; it was a declaration of ambition.

She set the pen down, took a bearing on the harbour entrance — now two miles astern and shrinking — and plotted her first position fix on the chart. A small cross, precisely drawn, with the time noted beside it: 09:17, departure.

The first mark on a new chart. It never got old.

She adjusted course to the southeast, trimmed the jib, and settled into the cockpit with her notebook on her knee. The Pagination heeled gently to port and began to make way. The wake streamed out behind in a narrow V, the only mark she would leave on the water. By evening it would be gone. By tomorrow the sea would have no memory of her passage at all.

That was the fundamental asymmetry of cartography. The cartographer remembers the sea; the sea does not remember the cartographer. You could spend a lifetime recording every depth, every current, every contour of a coastline, and the ocean would regard you with precisely the same indifference it had shown the first person who ever put to sea on a log.

Anchora found this oddly comforting. It meant the work was never finished. It meant there was always another chart to draw.

She uncapped her pen again and began to sketch the headland that was passing to starboard: a blunt promontory of dark rock, topped with scrub grass and a navigation light that blinked every four seconds. She noted the light’s characteristics in the margin — Fl(1) 4s, 12m, 8M — and added a small annotation about the rocks that extended from the headland’s base in a series of jagged shelves. These rocks did not appear on the Admiralty chart. They were, she suspected, a relatively recent arrival, deposited by a landslip that no one had bothered to report to the hydrographic office.

This was exactly the sort of thing that justified the voyage. The official charts were not wrong, exactly. They were merely incomplete. And in navigation, incomplete and wrong amounted to the same thing when your keel was the instrument of discovery.

By noon, the headland was well astern and the coast had receded to a thin grey line on the port beam. Anchora made a sandwich from the provisions she had stowed in the ice-box — hard cheese, cured ham, and bread that would be stale by tomorrow — and ate it in the cockpit with one hand on the tiller. The wind was backing toward the west and strengthening. She would need to reef before evening.

She washed the sandwich down with coffee from the thermos, plotted another position fix, and wrote in her notebook: Day 1, 12:30. Twenty miles offshore. Wind NW 12 kn, backing. Seas moderate. One headland charted, with previously unreported rocks. Atlas begun.

She had no idea how uncharted things were about to become.

Chapter 2 (Part 1)

Chapter 2
The Twin Harbors

The first landmark on Anchora’s route was a pair of harbors so close together that early cartographers had drawn them as one. Later cartographers, attempting to correct the error, had split the entry in two — but neglected to update the name on the second sheet. The result was a pair of harbors that existed on some charts as a single entity and on others as two separate places with the same name, separated by a narrow cut of water that was either an inlet, a channel, or “the bit in the middle,” depending on whom you asked.

Anchora first spotted the harbors from three miles out. They presented themselves as a low, dark line of wharves and warehouses, punctuated by the vertical strokes of mast-heads and the occasional plume of woodsmoke from a chandler’s stove. Behind the waterfront, the town rose in terraces: whitewashed houses with terracotta roofs, a church tower, and what appeared to be either a lighthouse or a very ambitious chimney.

She consulted her chart. According to the Admiralty edition, this was Port Gemini — a single harbor with a single entrance. According to the pilot book, it was East Harbor and West Harbor, two separate harbors sharing a common breakwater. According to the local fishing cooperative’s newsletter, which Anchora had found pinned to a notice board during her last provisioning stop, it was “the Twins,” and had been called that for longer than anyone could remember, which is to say at least forty years.

A pilot boat emerged from behind the breakwater as Anchora rounded the outer buoy. It was a stubby craft, painted orange, with a cabin that looked like it had been designed by someone who believed windows were a sign of weakness. A man stood on the foredeck, holding a coiled line and wearing an expression of professional neutrality.

“So which is East Harbor and which is West?” Anchora called across the narrowing gap.

“Depends which chart you’re reading,” the pilot replied cheerfully. “On mine, you’re entering East. On the harbourmaster’s, this is still West. We gave up arguing about it years ago.”

“And the Admiralty?”

“The Admiralty says there’s only one harbor, which will come as a surprise to the three hundred boats moored in the other one. We sent them a letter about it. Twice. They sent back a form asking us to confirm the coordinates of the harbor we were claiming didn’t exist, which rather missed the point.”

The pilot threw his line to Anchora, who caught it, made it fast to the midship cleat, and allowed herself to be towed through the entrance at a stately two knots. The breakwater slid past: massive blocks of granite, green with weed below the tide line, topped with a rusting iron railing and a row of bollards that had seen better centuries.

Inside the breakwater, the harbor opened up. It was larger than Anchora had expected — a broad basin ringed with pontoons, moorings, and the kind of stone quay walls that speak of an era when civic infrastructure was built to impress rather than merely to function. Fishing boats occupied most of the moorings: trawlers with rust-streaked hulls, lobster boats bristling with pots, and the occasional sleek yacht that looked profoundly uncomfortable in such working company.

“This is East Harbor, then?” Anchora asked as the pilot pointed her toward an empty berth on the visitors’ pontoon.

“If you like,” the pilot said agreeably. “The post office calls it East. The tax office calls it West. The pub calls it ‘here.’ I find that last one the most accurate.”

Anchora secured the Pagination to the pontoon with a bowline fore and aft, adjusted the fenders, and stepped ashore. The pontoon swayed under her feet in a way that she found pleasantly familiar. Solid ground, by contrast, always felt slightly unreliable.

She spent the morning exploring the first harbor on foot, making notes and sketches. The quay wall was three hundred and twelve paces long — she measured it, because that was the kind of person she was. The depth alongside ranged from two to four metres, depending on the state of the tide, which was semi-diurnal with a range of about two metres. There were fourteen pontoons, each capable of berthing eight to ten boats. The harbor entrance was fifty metres wide, oriented to the southwest, and partially sheltered by the breakwater from the prevailing swell.

All of this she recorded in her notebook, then transferred to her chart in the careful, precise hand that was her professional signature. She drew the quay walls in solid black lines, the pontoons in dashed lines, the depth contours in fine blue, and added a small compass rose in the corner of the inset. It was satisfying work. By lunchtime she had a complete survey of what she was, for the sake of her sanity, calling East Harbor.

In the afternoon she walked through the narrow cut that connected the two basins. It was barely thirty metres wide — a slot in the rock through which the tide funneled with surprising force. The walls of the cut were sheer, rising six or seven metres above the waterline, and covered with the layered evidence of centuries of marine growth: barnacles, mussels, a fur of green algae, and the occasional optimistic sea anemone.

Anchora recorded the discrepancy in her notebook — one harbor or two? — and sailed the Pagination through the narrow cut under engine. The water changed from grey-green to a deep cobalt as the bottom dropped away. The echo of the engine bounced off the cut walls and returned with a slight delay, as though the passage were thinking about what it had heard before deciding to repeat it.

Chapter 2 (Part 2)

On the far side of the cut, the second harbor opened like a cupped hand. Fishing boats bobbed in neat rows, their hulls painted in fading primaries: cadmium red, cerulean blue, chrome yellow. This harbor was slightly smaller than the first but considerably busier: a fish market was in full cry along the north quay, and the air smelled of brine, diesel, and the particular sharp tang of fresh-caught mackerel.

Anchora nosed the Pagination into a gap between two lobster boats whose skippers were engaged in a conversation conducted entirely in hand gestures and appeared to concern the ownership of a crate of bait. She made fast, fore and aft, and went ashore with her notebook.

The second harbor was, if anything, more confusing than the first. It had its own harbourmaster — a different one from the East Harbor harbourmaster, which suggested a degree of administrative independence that was hard to reconcile with the Admiralty’s insistence that this was all one place. It also had its own set of charts, posted on a notice board outside the harbourmaster’s office, and these charts were not merely different from the Admiralty charts; they were different from the charts posted in the East Harbor office.

“We draw our own,” the West Harbor harbourmaster explained when Anchora asked about this. Her name was Constance and she had the weathered complexion and steady gaze of someone who had spent forty years watching boats do foolish things. “The Admiralty charts are… aspirational. We prefer something grounded in what we can actually see.”

“And East Harbor?”

“They draw their own as well. Different from ours, naturally. We disagree about the depth in the cut — they say three metres at low water, we say two and a half. We also disagree about the location of the inner reef — they put it forty metres south of where we put it. We have been disagreeing about this for, I believe, about seventy years.”

“Has anyone thought of measuring?” Anchora ventured.

Constance gave her a look that suggested this question had been asked before and that the answer was both yes and complicated.

“Three times,” Constance said. “The first survey found three metres. The second found two and a half. The third found two point seven but was conducted by the East Harbor harbourmaster’s nephew and is therefore inadmissible.”

Anchora spent two days surveying the second harbor with the same methodical care she had applied to the first. She measured the quay walls, sounded the depths, timed the tides, and charted the moorings. She also, because she was thorough, measured the cut from both ends and found it to be two point six metres at low water springs, which satisfied nobody but at least had the virtue of being her own measurement.

On the second evening, as the sun dropped below the breakwater and turned the water in the harbor to molten copper, the pilot boat came alongside the Pagination. The pilot — whose name, Anchora had learned, was Rufus — was sitting on the gunwale eating an apple.

“You’ll want to note,” he called across, “that neither harbor appears on the Admiralty chart. Officially, this is all open water.”

“I had noticed,” Anchora said.

“The chart shows a straight coastline where we’re sitting. No indentation, no breakwater, no harbors. According to the Royal Hydrographic Office, we are currently bobbing about in the open sea. I sometimes think about this when I’m tying up to the pontoon.”

Anchora drew both harbors on a single sheet — one chart, two basins, no ambiguity — and labeled it with coordinates she trusted rather than names she did not. She drew the cut at its correct width, marked the depth as 2.6 m LAT, and added the inner reef in the position she had determined by triangulation from three fixed points on shore. She used blue for the water, black for the structures, and a particularly assertive shade of red for the reef, because reefs that are subject to a seventy-year argument deserve to be drawn in a colour that commands attention.

She pinned the completed chart to the wall above her berth, where the varnished wood held brass tacks with the grip of long practice. The chart looked good. Clean lines, clear labels, unambiguous depth figures. For a moment — a brief, perfect moment — the world felt like a place that could be made orderly.

She would not feel that way again for some time.

Chapter 3

Chapter 3
Charting the Depths

Beyond the Twin Harbors, the continental shelf dropped away in three distinct terraces. Anchora’s fathom-line, weighted with lead and tallow, told the story in numbers: ten, forty, two hundred. Each number represented a different world — a different temperature, a different colour, a different set of hazards — and she intended to chart them all.

She had sailed out of the second harbor at dawn, motoring through the cut in the grey half-light while the fishing boats were still sorting their nets. Constance had waved from the harbourmaster’s office window. Rufus had been asleep on his pilot boat and had not waved at all.

Each terrace had its own character, its own light, its own hazards. She gave each a name and — because she was, above all, a cartographer — a sub-heading.

The Reef

The first terrace was a broad shelf of coral, alive with colour and peril in roughly equal measure. It extended for perhaps two miles from the outer edge of the harbor approaches, a gradually shoaling platform of limestone and living coral that the chart described as “foul ground” and the local fishermen described with a word that Anchora did not record in her notebook but committed to memory for future use.

Staghorn formations reached toward the hull like bony fingers. Fan corals swayed in the current with the slow grace of a metronome. Brain corals squatted on the seabed like boulders, their surfaces etched with grooves that looked, from above, like the contour lines on a particularly complicated chart. Anchora appreciated the irony. Even the ocean floor was trying to be a map.

The water over the reef was extraordinarily clear. She could see the bottom in seven metres of water as though it were behind glass: every coral head, every sandy patch, every dark crevice where a moray eel might be contemplating its options. Fish moved through this landscape in schools and squadrons, their colours too vivid for the grey day — electric blue, acid yellow, a hot pink that no chart convention had a symbol for.

Anchora worked her way across the reef under reduced sail, taking soundings every hundred metres. The depth varied unpredictably: three metres here, seven metres there, and in one alarming spot, barely one and a half metres over a coral head that lurked just below the surface like a geological ambush. She marked each sounding on her chart with the compulsive precision of someone who knows that a single missed hazard can ruin a hull and a reputation in the same instant.

“Beautiful,” she murmured, leaning over the rail to watch a school of parrotfish graze on the coral below. The fish were vivid turquoise, the size of dinner plates, and they ate the coral with an audible crunching sound that was both fascinating and faintly disturbing. Then her keel scraped limestone — a sound like a giant clearing its throat — and she revised her opinion to “beautiful but inconvenient.”

She started the engine and motored off the coral head, checking the bilge for leaks (none, thank goodness; the Pagination’s hull was tougher than it looked) and the keel for damage (a scrape, nothing more, but a scrape that would need antifouling paint before the season was out). The parrotfish watched her departure with the sublime indifference of creatures that do not have keels.

By late morning she had completed her survey of the reef. It covered an area of roughly four square miles and contained, by her count, twenty-three significant coral heads within two metres of the surface at low water. She inked the reef in hatched lines on her chart, shading the shallowest areas in a warning blue, and adding a marginal note: Draft exceeding 1.5 m: proceed with caution and strong language.

She added, after a moment’s thought: Draft exceeding 2.0 m: do not proceed at all. Go around.

The reef chart was one of the most detailed pieces of work she had produced. She allowed herself a moment of satisfaction, then turned her attention to the second terrace.

The Current

Below the reef shelf, the seabed fell away sharply — not gradually, not gently, but with the decisiveness of a cliff edge. One moment the depth sounder was reading eight metres; the next it was reading forty, and still dropping. The colour of the water changed too, from the pale turquoise of the reef to a deep, serious blue that suggested the ocean was no longer in a playful mood.

And the water began to move. Not the slow, companionable drift of tidal flow, not the rhythmic surge of swell, but a purposeful lateral current that pushed the Pagination sideways at a rate Anchora found personally offensive. She had been sailing a straight course toward a landmark on the far shore; within ten minutes, the landmark had migrated from dead ahead to thirty degrees off the starboard bow, and the Pagination was crabbing sideways like a startled dog on a polished floor.

Anchora dropped the jib, started the engine, and pointed the bow directly into the current. The Pagination held position, but only just. The engine was working at three-quarters throttle to make zero progress over the ground, which was the nautical equivalent of running on a treadmill.

She set a kedge anchor — a small, light anchor attached to a long line that she rowed out in the dinghy and dropped on the seabed upstream of the current. With the boat held in place by the anchor, she could take bearings without the distraction of being pushed steadily toward a different postcode.

The boat swung in lazy arcs at the end of her anchor line, like a weathervane that could not quite decide where the wind was coming from. Anchora took bearings on three fixed points on shore every fifteen minutes, plotted them on her chart, and calculated the current’s speed and direction from the resulting drift vectors.

The current, she calculated, ran at two and a quarter knots on the ebb tide, swinging to three and a half on the flood. It flowed roughly parallel to the coast, with a slight onshore component during the flood that would push an unwary vessel toward the reef. At springs — when the tidal range was at its greatest — she estimated the current would exceed four knots, which was faster than many boats could motor against and would certainly be faster than any sailboat could point into.

It was the kind of data that looked innocent in a table and lethal on a lee shore. Anchora had a healthy respect for currents. They were the ocean’s way of reminding you that it was bigger than you and had somewhere to be.

She spent the rest of the day anchored in the current zone, taking readings at half-hour intervals. By evening she had enough data to draw a current chart: a pattern of arrows showing direction and speed at various states of the tide, annotated with times referenced to high water at the Twin Harbors. She added a caution box in the margin: Spring rates may exceed 4 kn. Passage against the flood inadvisable for vessels under power of 15 hp. Passage under sail alone is a matter for the individual conscience.

She noted it all down, adding small arrows to indicate direction and strength. The chart was filling up nicely. She had coral, she had current, and she had the beginnings of a bathymetric profile that would make a hydrographer weep with something that might have been either admiration or envy.

The Depths

Past the current, the water turned from blue to black. It was a gradual transition — not a line, not a boundary, but a slow deepening of colour that spoke of depths where sunlight was a rumour and pressure was a fact of life. The Pagination sailed into this darker water on a light breeze, and Anchora felt the temperature drop a degree as the deep water exhaled its chill.

Anchora’s lead-line ran out at two hundred fathoms and found nothing. She tied on an extension — fifty fathoms of spare line she kept coiled in the lazarette for exactly this purpose. At three hundred fathoms: nothing. The line hung straight down into the black water, trembling slightly with the current, offering no information except that the bottom was not yet here.

She tied on a second extension. At four hundred fathoms — nearly half a mile of line, its weight now enough to make her arms ache from the hauling — she felt a distant, uncertain bump. The lead had found something. Whether it was bedrock, sediment, a drowned mountain, or the roof of something she preferred not to think about, the line could not tell her. It only knew that it had stopped going down.

“Well,” she said to the empty cockpit, “there is apparently a bottom.”

She hauled the line in, arm over arm, for what felt like an hour but was probably fifteen minutes. The lead came up coated in a fine grey clay that told her the bottom was soft sediment — the accumulated drift of millennia, particles of sand and silt and the ground-down remains of creatures that had lived and died in the water column above and settled, with infinite slowness, to the floor.

She examined the clay sample with the professional interest of someone who understands that even mud has a story to tell. It was fine-grained, slightly sticky, with no visible shell fragments or organic material. Deep-water sediment. The kind of bottom that offers poor holding for an anchor but excellent preservation for anything that sinks to it.

She took three more soundings over the course of the afternoon, sailing a line perpendicular to the coast. The depths were remarkably consistent: three hundred and eighty fathoms, four hundred and ten, three hundred and ninety-five. The seabed here was a flat plain, featureless and unchanging, stretching into the darkness in every direction.

She drew a single contour line on her chart — the four-hundred fathom line, traced with a steady hand despite the gentle rolling of the boat — and wrote 400+ fm beside it. Below the line she left blank space: the cartographer’s admission that some things are simply too deep to record, too far from light and air and the concerns of surface-dwellers to justify the ink.

But she kept the clay sample, sealed in a small glass jar and labeled with the date, position, and depth. A cartographer records what can be recorded. The blank space on the chart was not ignorance. It was honesty.

Interlude

The Phantom Island

Between Chapter Three and Chapter Four, Anchora sailed past an island that did not exist.

She spotted it on the morning of her third day out from the Twin Harbors, in a stretch of sea where her charts showed nothing but open water and a scattering of depth soundings that ranged from “deep” to “very deep.” It appeared first as a dark smudge on the horizon, which she initially took for a low cloud or a particularly ambitious wave. But clouds do not stay in one place, and waves do not have palm trees, and this thing had both.

She altered course to investigate. A cartographer who sails past an uncharted island without investigating is not, in Anchora’s professional opinion, a cartographer at all, but merely someone with a boat and a collection of pens.

The island resolved itself as she approached. It was small — perhaps a quarter of a mile across, roughly circular, fringed with a beach of white sand that was so bright in the morning sun it made her squint. Behind the beach, a stand of coconut palms rose to a modest height, their fronds rustling in a breeze that Anchora could not feel from the water. At the island’s centre, a low hill rose perhaps fifty feet above sea level, covered in scrub grass and what appeared to be a single, determined frangipani tree in full bloom.

It had, in short, all the attributes of an island. It had mass, it had volume, it had geographic coordinates that Anchora carefully noted from her sextant readings. It had a beach of white sand that squeaked underfoot when she rowed the dinghy ashore and stepped out, which is a very specific and tactile quality that imaginary places tend not to possess.

It had a small but enthusiastic colony of penguins, which was geographically improbable but zoologically undeniable. They were small penguins — about eighteen inches tall, with blue-grey backs and white fronts and an air of having been expecting her. They stood in a loose group near the tree line, regarding her with the intensity of creatures who have opinions about visitors but lack the vocal apparatus to express them clearly.

“You’re not supposed to be here,” Anchora told them. Penguins, she was fairly certain, belonged in the southern hemisphere, on rocky coasts and ice shelves, in places where the water was cold and the fish were plentiful and the nearest palm tree was several thousand miles away. These penguins appeared not to have read the relevant literature.

She walked the perimeter of the island, which took twenty-three minutes at a brisk pace. The beach was continuous — fine white sand, no rocks, gently shelving into water that was a shade of turquoise that postcards aspire to but rarely achieve. The interior was equally unremarkable: the palm grove, the scrub hill, the frangipani, and an absence of anything that might explain why this island was here. No volcanic cone. No coral atoll. No geological justification whatsoever.

She took careful measurements. She recorded the latitude and longitude from three separate sextant shots, all of which agreed. She sounded the water around the island’s shore and found a uniform depth of four fathoms, dropping sharply to twenty fathoms fifty metres out. She noted the compass bearing to the nearest known landmark — the headland beyond the Twin Harbors, barely visible on the horizon — and calculated the distance.

All the data said the island was real. The numbers were precise and consistent. The sand was tangible. The penguins were audible (they had begun a low, conversational muttering among themselves that suggested they were discussing her survey technique).

She reached for her pen to add it to the atlas, then paused.

The island was not in the table of contents. She checked — carefully, methodically, the way she checked everything. She ran her finger down the list of chapters, sections, and appendices. The Foreword was there. The five chapters were there. The appendices were there. But between Chapter Three and Chapter Four, where this island chronologically belonged, there was no entry. No heading, no sub-heading, no footnote.

It was not listed in the index. It appeared in the reading order — she was, after all, reading about it right now, and so were you — but no navigation entry pointed to it. If you were using the table of contents to move through this book, you would skip from Chapter Three directly to Chapter Four and never know this island was here. It existed in the spine but not in the navigation. It was structurally present but navigationally invisible.

“If a place has no entry,” she asked one of the penguins, “does it really exist?”

The penguin regarded her with the weary patience of a creature that has heard this question before — perhaps from other cartographers who had stumbled upon the island and grappled with the same ontological difficulty. It tilted its head to one side, blinked once with each eye in sequence, and then waddled back to the surf with the unhurried dignity of a creature whose existence does not depend on being listed in a table of contents.

Anchora sat on the beach for a while, thinking about this. The penguins went about their business around her: waddling, swimming, standing in small groups and staring at the horizon with the focused attention of creatures who are waiting for something but have forgotten what. The waves lapped at the sand. The palm fronds rustled. The frangipani released its scent into the warm air.

It was, she had to admit, a very pleasant island. Peaceful. Unhurried. The kind of place where you could sit for an afternoon and forget that you had charts to draw and depths to sound and harbors to argue about. Perhaps that was the point. Perhaps some places are better off uncharted — hidden in the reading order, reachable only by those patient enough to read every page rather than jumping from heading to heading.

She considered this philosophical position for approximately four minutes, then rejected it. She was a cartographer. Uncharted places were not romantic; they were errors.

And yet.

She closed her notebook. She rowed back to the Pagination. She hauled up the anchor, set the jib, and sailed away from the island on a reaching course that would bring her to the coordinates she had been given for the beginning of the Long Voyage.

She did not add the island to her chart. There was no heading for it in the atlas, no entry in the navigation, no place for it in the structure she had so carefully planned. An island without a table of contents entry was, from a cartographic perspective, an orphan — and Anchora’s atlas did not have room for orphans.

She did, however, add penguins to her growing list of things that were not supposed to be there but were. The list, she reflected, was getting longer with every day of the voyage. She was not sure whether this said something about the world or about her list.

Behind her, the island shimmered in the heat haze and slowly sank below the horizon, taking its penguins and its palm trees and its existential questions with it. By evening it was gone. By morning she would not be entirely sure she had seen it at all.

But the sand between her toes — fine, white, faintly squeaky — suggested otherwise.

Chapter 4 (Day 1)

Chapter 4
The Long Voyage

Day One — Departure

The passage to the outer archipelago was, by all accounts, a three-day sail. This estimate came from Rufus the pilot, who had never made the passage himself but had spoken to several people who had, and from Constance the harbourmaster, who had made it once, thirty years ago, in a boat considerably faster than the Pagination and in weather considerably better than the forecast was promising.

Anchora provisioned accordingly: three days of hard-tack, three days of tinned sardines, and three days of the sort of instant coffee that dissolves under protest. She also packed a fourth day’s supplies, because she had learned through experience that three-day passages have a way of becoming four-day passages, and a sailor without coffee on day four is a sailor without hope.

She reviewed her charts one more time before casting off. The outer archipelago was marked on the Admiralty chart as a cluster of small islands approximately seventy miles to the southeast — close enough to reach in three days of moderate sailing, far enough to require planning, provisioning, and the acceptance that the nearest help, should help be needed, was a day’s sail behind her. The islands themselves were drawn as rough outlines, their coastlines sketched from satellite imagery and the reports of passing ships. No one had surveyed them properly. That was why Anchora was going.

She cast off at first light, motoring through the harbor and out past the breakwater into open water. The dawn was a thin line of amber on the eastern horizon, and the sea was a flat expanse of grey that merged, at the edges, with a sky of precisely the same shade. The wind was light from the southeast — a soldier’s wind, steady and reliable, the kind of wind that does not require constant attention to the sheets.

Day one was, in the main, uneventful. The wind held steady, the seas were moderate — a long, low swell from the south that the Pagination rode with an easy, rocking motion — and the boat made six knots on a beam reach without complaint. This was her best point of sail: the wind on the beam, the sails drawing well, the hull slicing through the water with the minimum of fuss and the maximum of speed.

Anchora spent the daylight hours sketching coastline profiles from the deck. The coast was still visible to the northwest, a low grey line that rose occasionally into headlands and fell back into bays. She drew each headland as she passed it, noting its shape, its height, its distinguishing features. A lighthouse here. A radio mast there. A cliff face that had collapsed into a jumble of rocks that would, she suspected, not appear on any chart for another decade.

By mid-morning the coast had begun to recede, and by early afternoon it was little more than a suggestion — a slight thickening of the horizon, a shade of grey that was fractionally darker than the grey of the sea. Anchora took a last bearing on the highest headland, plotted it on her chart, and acknowledged that she was, for practical purposes, out of sight of land.

This was a moment that some sailors dreaded and others relished. Anchora was in the second category. Out of sight of land, the world simplified itself. There was the boat, there was the sea, and there was the compass. Everything else — the harbors and their arguments, the charts and their disagreements, the penguins and their inexplicable geography — fell away. What remained was navigation in its purest form: a vessel, a heading, and the conviction that the maths would work out.

She spent the evening hours transferring her coastline sketches to the master chart, working by the light of the oil lamp that swung from a hook above the chart table. The lamp threw a warm, yellow circle of light that moved gently with the boat’s motion, and the shadows it cast gave the chart a depth and texture that the flat light of day did not. She added her soundings from the reef and current zones, drew the four-hundred-fathom contour line she had found over the deep water, and penciled in a tentative route to the archipelago.

By sunset, the Twin Harbors had sunk below the horizon and there was nothing in any direction but water and the thin, bright line where it met the sky. The sky turned from grey to amber to a deep blue-violet that held the first stars in its upper reaches. The wind eased slightly as the land breeze died and the sea breeze had not yet arrived. The Pagination slowed to four knots, then three, and the water around her hull changed from a hiss to a whisper.

Anchora set the self-steering gear — a wind-vane device of her own construction that held the boat on course without human intervention, provided the wind did not shift by more than fifteen degrees, which it did roughly once an hour — and went below to make dinner. Tinned sardines on hard-tack, with a cup of instant coffee that tasted, as always, of ambition unfulfilled.

She marked her position on the chart with a small cross and the time: Day 1, 18:47, all well. 42 miles made good. Barometer steady. Wind SE 8 kn. Coffee: adequate.

Chapter 4 (Day 2)

Day Two — Becalmed

The wind died at dawn and did not return. It did not die gradually, the way winds usually die — fading through a series of diminishing puffs until the last breath of air gives up and the sails go slack. It died all at once, as though someone had closed a door. One moment the Pagination was sailing; the next she was not.

The Pagination sat on water so flat it reflected the clouds like a tray of mercury. Every ripple, every swell, every hint of motion had been smoothed away. The sea was a mirror. The sky was reflected in it so perfectly that the horizon vanished, and Anchora had the unsettling sensation of floating in the centre of a sphere, with identical views in every direction, up and down included.

The sails hung in dispirited loops. The main sagged against the shrouds. The jib, unfurled and hopeful an hour ago, now hung from the forestay like a bedsheet on a clothesline. Even the tell-tales — small strips of yarn attached to the shrouds to indicate wind direction — drooped vertically, pointing at the deck with the resigned finality of arrows that have given up looking for a target.

Anchora tried whistling for wind (maritime tradition). She stood in the cockpit and whistled a jaunty tune toward each of the four cardinal points in turn, which was supposed to summon a breeze from the direction you whistled at. The tradition did not specify what to do when all four directions were equally windless. Anchora whistled at all of them anyway, on the principle that comprehensive failure was preferable to targeted failure.

She tried tapping the barometer (maritime superstition). The barometer was a fine brass instrument mounted on the bulkhead in the companionway, and it showed 1024 millibars, which was high and stable and thoroughly uninterested in change. Tapping it — a gentle, respectful tap with the knuckle of the index finger, as tradition demanded — produced no change. She tapped it again, harder. Still nothing. The barometer regarded her with the mechanical smugness of a device that knows it is right.

She tried simply sitting in the cockpit and staring at the horizon (maritime realism). This was, in many ways, the most honest response to a calm: the acceptance that the wind would return when it chose and not before, and that no amount of whistling, tapping, or staring would hasten its arrival. Sailors who raged against calms accomplished nothing except to make themselves hot and irritable. Sailors who accepted them at least stayed cool.

The morning passed in a haze of heat and stillness. Anchora read a chapter of the pilot book, made a cup of coffee (the second-to-last packet of instant — she was rationing now), and wrote a long entry in her notebook about the geology of the continental shelf, drawing on the soundings she had taken over the past few days. The writing helped. It gave her hands something to do and her mind something to chew on that was not the infuriating absence of wind.

By noon the heat was ferocious. The sun stood directly overhead, pouring its energy into the flat sea and the flat boat and the flat cartographer who was beginning to feel like a piece of hard-tack left on a windowsill. The deck was too hot to touch barefoot. The cabin was an oven. The cockpit, shaded by the boom, was merely oppressive rather than unbearable, which was the best that could be said for it.

She rigged a sun-shade from the spare jib, draping the heavy sailcloth over the boom and tying it to the shrouds on either side. This created a rectangle of shade over the cockpit that was eight feet by six feet and reduced the temperature from “punitive” to “merely unpleasant.” She sat in this shade and spent the afternoon measuring the depth with her lead-line, which was at least productive even if it was not comfortable.

The bottom here was sixty fathoms of soft mud — the same fine grey clay she had found in the deep water, but closer to the surface. The mud yielded to the lead with a reluctant sucking sound that Anchora could feel through the line, a hundred and twenty metres away. She noted this on the chart, adding a small sounding figure at her current position, though she suspected no one would ever care. Sixty fathoms of mud in the middle of nowhere was not the kind of information that made it into the Admiralty Notices to Mariners.

She took several more soundings as the afternoon wore on, drifting imperceptibly — the current was weak here, barely a quarter of a knot — and recording each depth. Fifty-eight fathoms. Sixty-two. Fifty-five. The seabed was gently undulating, rising and falling by a few fathoms over a distance of perhaps half a mile. She drew these contours on her chart, and the result looked like the gentle hills of a drowned landscape, which is precisely what it was.

At sunset, a catspaw rippled the surface to the north — a small, dark patch of ruffled water that raced across the mirror-flat sea like a whispered secret. Anchora leaped to her feet, freed the jib sheet, and hauled in the main, her hands moving with the desperate speed of someone who has been waiting all day for this exact moment. By the time she had the sails trimmed, the catspaw had arrived, touched the Pagination’s sails with the lightest of fingers, and moved on. The sails filled for a moment — a single, heartbreaking moment — and then fell slack again.

The catspaw raced on to the south and disappeared. The sea returned to its mirror state. The tell-tales drooped.

Anchora sat down in the cockpit and regarded the empty horizon with an expression that, in a less disciplined cartographer, might have been called despair. In Anchora it was merely a very thorough species of disappointment.

She marked her position: Day 2, 19:02, becalmed. 3 miles made good by drift. Barometer 1024, steady. Wind: none. Morale stable but coffee supply critical. Lead-line soundings taken; seabed contour added to chart. If the wind does not return by morning I shall motor, which is an admission of defeat but at least it is defeat with forward progress.

Chapter 4 (Day 3)

Day Three — Arrival

The wind returned overnight with the subtlety of a cannon shot. One moment the Pagination was drifting, her hull barely creasing the water, her sails hanging like theatre curtains after the final act. The next moment she was heeled over at twenty degrees, spray flying from the bow in white sheets, the rigging singing a chord in B-flat minor that rose in pitch as the gusts built and fell again as they passed.

Anchora, who had been asleep in the quarterberth with her head on a folded chart and her feet on a coil of rope, arrived on deck wearing one boot and an expression of startled competence. The wind was from the northwest — exactly the direction she needed — and blowing at what she estimated was twenty knots, gusting twenty-five. This was more wind than the Pagination wanted with full sail set, and considerably more wind than Anchora wanted at three o’clock in the morning with one boot on.

She reefed the main in the dark, working by feel and muscle memory. The reef points were where her fingers expected them to be; the lines ran through the blocks without snagging; the sail came down to its first reef in under two minutes. She trimmed the jib to match, easing the sheet until the sail stopped flogging and settled into the shape it wanted, and pointed the bow toward the coordinates she had been given for the outer archipelago.

The Pagination responded immediately. With the reef in, she was balanced, manageable, and making seven knots through the water — her best speed of the voyage. The bow wave hissed along the hull. The wake stretched out behind in a long, white tail. The self- steering gear held the course without complaint, and Anchora, who was now fully awake and beginning to enjoy herself, went below to find the other boot.

Dawn came slowly, the way it does at sea: a gradual lightening of the eastern sky, from black to charcoal to a pale grey that spread upward like water rising in a glass. The horizon emerged from the darkness, sharp and clean, and Anchora scanned it with the practiced eye of someone who has been looking for land from the deck of a small boat for most of her adult life.

Nothing. Not yet. But the chart said the archipelago should be visible by mid-morning at this speed, and Anchora trusted her chart because she had drawn it herself and she knew exactly how much trust it deserved, which was a moderate amount tempered by professional humility.

She made coffee — the last packet, which she regarded as both a sacrifice and an incentive — and drank it in the cockpit while the Pagination sailed herself. The wind was settling down to a steady eighteen knots, the gusts becoming less frequent, the sea developing a regular pattern of waves that the boat cut through with a satisfying, rhythmic motion. This was good sailing. The kind of sailing you remember long after the voyage is over.

At ten o’clock she saw something on the horizon that was not cloud and not sea. It was a faint, dark irregularity in the line where sky met water — a slight bump, barely perceptible, that could have been a distant ship or a trick of the light but wasn’t. She knew what it was. She had been looking for it for three days.

Land appeared at noon: a low, dark smudge that resolved slowly into individual islands, then individual trees, then individual birds perched in the individual trees. She counted seven major islands and an uncountable number of rocks, shoals, and ambiguous features that might have been either. The islands were spread across perhaps ten miles of sea, arranged in a rough arc that opened toward the northwest — toward her, as it happened, which felt like a welcome even though she knew it was merely geography.

She sailed closer, reducing speed as the water shallowed. The depth sounder, which had been reading “deep” for two days, suddenly began producing numbers: 80 metres, 60, 40, 30. She could see the bottom now — a sandy seabed dappled with shadows from the islands above. Fish scattered as the Pagination’s shadow passed over them.

She rounded the northern tip of the largest island, staying well clear of a line of rocks that extended from the shore like a broken jetty, and found herself in a sheltered anchorage on the island’s lee side. The water was calm here, protected from the northwest wind by the island’s bulk. She dropped anchor in four fathoms of sand, felt it bite, and paid out enough chain to hold in a gale.

Then she sat in the cockpit and looked at the islands spread before her. Seven islands. Uncharted, unsurveyed, waiting. This was why she had come. Not for the harbors with their naming disputes, not for the reefs with their lurking coral heads, not for the currents with their treacherous pull. She had come for this: blank space on the chart, waiting to be filled.

She uncapped her pen. This, at last, was what she had come for.

She marked her position one final time: Day 3, 12:15, landfall. Anchored in the lee of the largest island, 4 fm sand. Seven islands visible. Wind NW 18 kn. Coffee: exhausted. The atlas begins in earnest.

Chapter 5

Chapter 5
Archipelago of Wonders

The outer archipelago comprised seven islands, but Anchora quickly learned that three of them demanded most of her attention. The other four were low, sandy, and profoundly uninteresting — the kind of islands that exist primarily to give seabirds somewhere to argue. She surveyed them anyway, because a cartographer who skips the boring bits produces an atlas with holes in it, and an atlas with holes is just a collection of maps pretending to be complete.

The four minor islands took two days to chart. They were, as expected, unremarkable: flat coral platforms, none more than ten feet above sea level, covered in coarse grass and nesting terns. Each island had a fringing reef, a scattering of rocks, and a beach that was pleasant enough in theory but occupied in practice by large numbers of territorial seabirds that regarded Anchora’s survey equipment with undisguised hostility.

She drew them quickly and moved on to the three islands that mattered.

Isle of Echoes

The first notable island was ringed by basalt cliffs so sheer they formed a natural amphitheatre. The cliffs rose two hundred feet from the waterline, sheer and dark, their faces carved into columns by millennia of cooling and cracking. At the base, where the waves struck, the rock had been hollowed into caves and arches that amplified every sound into a symphony of percussion.

Every sound bounced from wall to wall in diminishing repetitions: the crash of a wave became a series of claps; a shout became a conversation with oneself; the rattle of an anchor chain became a cascading diminuendo that lasted for eight or nine seconds before finally fading into the ambient hiss of the sea.

Anchora discovered the echoes accidentally. She was approaching the island from the northeast, looking for an anchorage, when she called out “Hello!” to see if anyone was ashore. The island replied: Hello … hello … hello … ello … lo … It was, she reflected, the most polite island she had ever visited. Most islands ignored you entirely.

She tested the acoustics systematically, because that was her nature. She stood in the cockpit, positioned the Pagination at measured distances from the cliff face, and shouted a series of test words: single syllables first, then multi-syllable words, then complete sentences. She timed the echoes with her watch and recorded the results in her notebook.

She tested the acoustics further by reading her coordinates aloud. “Fourteen degrees, thirty-seven minutes south,” she announced in a clear, carrying voice. The cliffs repeated it back, each echo slightly garbled by the complex geometry of the rock face, until by the seventh repetition the island appeared to be declaring itself at forty-seven degrees north. She noted this as a navigational hazard of a kind not typically covered in the pilot books.

The anchorage, when she found it, was on the island’s sheltered western side, where the cliffs gave way to a small bay with a sandy bottom in three fathoms. She anchored here and spent two days surveying the island from the dinghy, rowing along the cliff face with a notebook in one hand and an oar in the other, measuring the height of the cliffs by trigonometry and the depth of the water by lead-line. The caves at the base of the cliffs were deep enough to row into — dark, dripping spaces where the sound of the oars echoed in unsettling ways and the water was so still it reflected the cave roof like polished obsidian.

She mapped each cave entrance, noting its position, width, height, and the depth of water at its threshold. There were eleven caves in total, ranging from barely large enough for the dinghy to a cathedral- sized chamber that could have sheltered a small fishing fleet. In the largest cave she found the remains of an old mooring ring, bolted into the rock at waterline level and rusted to a deep orange-brown. Someone had been here before. Someone had anchored in this cave, in this darkness, and had thought it worthwhile to drill a hole in the basalt and hammer in a ring.

She added this to the chart with a small symbol and the annotation: Old mooring ring, condition poor. Cave depth 14 m, width 8 m, height 5 m at entrance. Adequate shelter in westerly weather.

The chart grew a detailed inset of the Isle of Echoes, complete with soundings, anchorage notes, cave positions, and a warning about acoustic anomalies that she phrased with care: Echoes from basalt cliffs may distort voice communications. Coordinates heard from echo may not correspond to coordinates spoken. Fog signals unreliable within 0.5 nm of cliff face.

Compass Rose Atoll

The second island was not, strictly speaking, an island at all. It was an atoll: a ring of coral enclosing a shallow lagoon, with four narrow passes at the cardinal points. From the air — or from a sufficiently tall mast — it looked exactly like a compass rose, its four passes aligned so precisely with north, south, east, and west that Anchora suspected the coincidence was too perfect to be entirely natural and too natural to be entirely coincidence.

The atoll was roughly circular, perhaps a mile across, its rim barely six feet above the water at its highest point. The coral was old and dense, covered with a thin layer of sand and scrub vegetation that had taken hold wherever the salt spray was not too fierce. On the wider sections of the rim, coconut palms grew in sparse lines, their trunks curved by years of prevailing wind into identical question-mark shapes.

Anchora sailed through the north pass and anchored in the lagoon. The pass was narrow — barely thirty metres wide, with coral walls on either side that rose steeply from a bottom of clean white sand — and the current through it was brisk on the flood tide, perhaps two knots, which required attention but not alarm. She entered on the slack, when the current was negligible, and dropped anchor in the centre of the lagoon.

The water was so clear she could see her anchor chain lying on the bottom in gentle curves, like cursive script. The bottom was four fathoms of white sand, undisturbed by current, and the anchor had dug in with the easy confidence of a hook in soft ground. Fish moved through the water column above the sand: small, bright, purposeful fish that paid no attention to the boat or its occupant.

She spent a full day surveying the atoll, measuring each pass and sounding the lagoon. The symmetry was remarkable: each pass was within a boat-length of the same width, and the lagoon was uniformly four fathoms deep. The north and south passes were slightly wider than the east and west — thirty-two metres versus twenty-eight — but the difference was so small it might have been within the margin of error of her measuring method, which involved rowing the dinghy across the pass and counting oar-strokes.

She sounded the lagoon on a grid pattern, rowing the dinghy in north-south lines spaced fifty metres apart and dropping the lead every twenty metres. The result was a bathymetric chart of extraordinary regularity: a flat, featureless bowl of sand, uniformly four fathoms deep, with no coral heads, no rocks, no obstacles of any kind. It was, she reflected, the most boring piece of underwater terrain she had ever surveyed, and also one of the most useful: a lagoon with a flat bottom and predictable depth was a sailor’s dream, a place where you could anchor anywhere with equal confidence.

Nature, it seemed, had a taste for geometry. Anchora, who also had a taste for geometry, appreciated this more than most.

She drew the atoll with particular care, using a compass to lay out the circular rim and placing each pass at its precisely measured bearing. The chart of Compass Rose Atoll was, she thought, the most beautiful piece of cartography she had produced on the voyage: clean lines, perfect symmetry, four passes opening like the petals of a flower. She added a small compass rose in the corner of the inset chart, which created the pleasing recursion of a compass rose drawn inside a compass rose.

The Whirlpool Narrows

Between the second and third islands, the tidal flow compressed through a gap barely a cable’s length wide. The two islands were close here — so close that their reefs almost touched, leaving only a narrow channel of deep water between them. Through this channel, twice a day, the entire volume of water that filled and emptied the lagoons and bays of the archipelago had to pass, and it did so with considerable energy and very little patience.

The result was a whirlpool that spun with metronomic regularity: clockwise on the ebb, counterclockwise on the flood, and in a state of churning indecision at slack water. The whirlpool was not large — perhaps twenty metres across at its widest — but it was vigorous, and the water within it moved with a purposeful circular motion that was both mesmerizing and slightly alarming.

Anchora observed the whirlpool from a safe distance, anchored in the lee of the eastern island, and made detailed notes. The clockwise rotation during the ebb was faster than the counterclockwise rotation during the flood, which she attributed to the shape of the channel: slightly wider at the southern end, which gave the ebb current — flowing from north to south — more room to accelerate before hitting the narrows. She estimated the rotational speed at three revolutions per minute during the peak ebb and two during the peak flood.

The whirlpool was accompanied by an array of secondary effects that were equally interesting and equally dangerous. Standing waves formed at the edges of the narrows, where the moving water met the still water beyond, creating a line of breaking crests that would have been at home on a surfing beach. Eddies spun off the main whirlpool like sparks from a wheel, racing downstream in tight spirals before dissipating in the calmer water beyond. And the sound — a low, continuous roar, punctuated by the slap of breaking waves — carried for half a mile in every direction.

Anchora timed the cycles, measured the diameter, and estimated the rotational speed. She also measured the current through the narrows at various states of the tide, using a timing float — a sealed bottle with a small flag attached — that she dropped into the current and tracked with bearings. The maximum current was four and a half knots during the spring ebb, which was faster than the Pagination could motor against and considerably faster than anyone would want to sail through.

The slack water between the ebb and flood was brief — eight minutes by her measurement, during which the whirlpool lost its coherence, the standing waves subsided, and the channel was briefly, deceptively calm. This was the transit window: eight minutes in which a careful boat could pass through the narrows without being spun, swamped, or swept sideways into the reef.

Eight minutes was not a generous margin. It was, however, sufficient for a boat that was ready and a skipper who was decisive. Anchora filed this information under “useful” and drew the whirlpool on her chart as a neat spiral with arrows, then added the note: Transit at slack water only. Allow margin for error. Do not bring the good sextant.

She had now been in the archipelago for four days, and her chart was beginning to look like something a real navigator might trust. The blank spaces were filling in. The coastlines were taking shape. The soundings and current data were building a picture of an archipelago that was complex, challenging, and deeply satisfying to chart.

This pleased her more than she would have admitted to the penguins.

Appendices

Appendix A
Knot Types Employed

The following knots were tied at various points during the voyage and are reproduced here for the edification of the reader. Each knot is described in terms of its construction, its application aboard the Pagination, and Anchora’s personal assessment of its character, because all knots have character, and some have more character than their tyers would prefer.

Bowline. The king of knots. Used to secure the Pagination to docks, bollards, and on one occasion a surprisingly cooperative palm tree on the Isle of Echoes. The bowline forms a fixed loop at the end of a line that will not slip under load and will not jam when you want to untie it, which is a combination of virtues that very few knots can claim. Anchora tied bowlines instinctively; her fingers would form the loop, pass the tail around the standing part, and thread it back through the loop in a single fluid motion that took approximately one and a half seconds. She had once won a bowline-tying competition at a sailing club in the Leeward Islands, beating a boatswain with forty years’ experience and very large hands. The boatswain had been gracious in defeat and had bought her a rum, which is the traditional currency of nautical respect. Advantages: does not slip, does not jam, can be tied under load if necessary. Disadvantages: requires two hands, which is one more than a sailor holding a coffee cup has available.

Cleat hitch. Used at every marina, every pontoon, and every dock that the Pagination visited. The cleat hitch is not a glamorous knot — it is the workhorse of the marina, the knot equivalent of a reliable sedan — but it is satisfyingly quick to tie and untie, and it holds with absolute reliability in any conditions. Anchora could do it in four seconds. She timed herself, because she was the sort of person who timed herself tying knots, and she regarded this as a perfectly normal hobby. The cleat hitch involves one full turn around the base of the cleat, followed by two figure-eight turns over the horns, finished with a locking hitch. The locking hitch is optional but recommended by Anchora, who had once seen an unlocked cleat hitch work loose in a gale and release a thirty-foot yacht into the fairway, where it drifted gently into a row of dinghies and caused what the insurance industry would later describe as a “multi-vessel incident.”

Figure-eight. Used as a stopper knot on every sheet and halyard aboard the Pagination. The figure-eight prevents the line from running out through its block or fairlead, which is the kind of thing that happens at the worst possible moment — during a tack, in heavy weather, when both hands are already occupied with something else that is going wrong. The figure-eight is the kind of knot that, once learned, the fingers tie without consulting the brain. Anchora tied figure-eights the way other people blink: automatically, unconsciously, and with a frequency that suggested a deep-seated neurological commitment to the prevention of runaway lines.

Round turn and two half-hitches. Used for securing the dinghy painter to rocks, trees, and the occasional bollard that was too large or too oddly shaped for a bowline. This knot is not elegant, but it is secure, adjustable, and can be tied around objects of any shape, which makes it the knot of last resort — the one you reach for when nothing else will work. Anchora used it approximately once per week and thought about it approximately never, which is the highest compliment a sailor can pay a knot.

Reef knot. Used for tying reef points when shortening sail. The reef knot is perhaps the most misunderstood knot in the nautical lexicon: it is often taught as a general-purpose binding knot, which it is not. It will slip if loaded unevenly, it will jam if loaded heavily, and it will capsize if looked at sternly. For tying reef points, however, it is perfect: quick to tie, quick to untie, and adequate for the modest loads involved. Anchora tied reef knots in the dark, in the rain, on a heaving deck, with numb fingers, and the reef knot never once let her down, because she never once asked it to do anything it was not designed to do.

Sheet bend. Used for joining two lines of different diameter. Anchora used this knot exactly once during the voyage, when she needed to extend her lead-line with a lighter-weight extension over the deep water. The sheet bend held, the sounding was taken, and the knot was untied. It was, she reflected, the knot equivalent of a specialist consultant: expensive to learn, rarely needed, but invaluable on the one occasion you called upon it.

Appendix B
Signal Flags Observed

During the voyage, the following International Code of Signals flags were observed flying from other vessels. The International Code is a system of flag signals that allows ships of different nationalities to communicate without a common language, which is an admirable ambition that works considerably better in theory than in practice, because the system assumes that all parties can identify forty flags at a distance and under conditions that typically include spray, glare, and a rolling deck.

Alpha. A white and blue swallowtail flag. “I have a diver down; keep well clear at slow speed.” Observed near the reef, flying from a battered workboat whose diver was, presumably, somewhere below, doing whatever it is that divers do in two metres of water over a coral reef. Anchora kept well clear, as instructed, and made a note of the workboat’s position so that she could add a “diving operations” annotation to her chart of the reef. She later removed the annotation on the grounds that the diving was presumably temporary, whereas the chart was intended to be permanent, or at least as permanent as a chart of a reef can be when the reef is still growing.

Bravo. A red swallowtail flag. “I am taking in, or discharging, or carrying dangerous goods.” Observed at the Twin Harbors, flying from a rusty coaster that was moored alongside the north quay in East Harbor (or possibly West Harbor; the question, as always, remained unresolved). The dangerous goods turned out to be a crate of live chickens, which the harbourmaster’s regulations classified as “livestock” but which the chickens’ behaviour — aggressive, noisy, and apparently attempting to escape — suggested might more accurately have been classified as “hazardous materials.”

Hotel. A white flag with a red vertical stripe. “I have a pilot on board.” Observed on Rufus’s pilot boat, naturally, and on several vessels entering and leaving the Twin Harbors under his guidance. The Hotel flag is one of the more useful signals in the code, as it tells other vessels that the flagged vessel is being navigated by someone who knows the local waters, and that any unusual manoeuvres it performs are probably intentional rather than accidental. Anchora flew the Hotel flag exactly once, when Rufus came aboard to guide her through the cut between the two harbors, and she removed it the moment he left, because flying a pilot flag without a pilot aboard is both improper and illegal, and Anchora was punctilious about such things.

November. A chequered blue and white flag. “No” or “negative.” Observed on a yacht that was being hailed by the coast guard and appeared to be declining to stop. Anchora watched this exchange with professional interest and noted in her log that the yacht was making nine knots downwind, which was probably insufficient to outrun the coast guard cutter but was certainly a spirited attempt. The outcome of the pursuit was not recorded, as the Pagination had by then sailed out of visual range.

Quebec. A solid yellow flag. “My vessel is healthy and I request free pratique.” Flown by the Pagination herself upon arrival at each port, mostly out of optimism. Free pratique is permission to make contact with the shore, granted by the port health authority after they have satisfied themselves that your vessel is not carrying plague, cholera, or any other communicable disease. In modern times, the flag is largely ceremonial — most ports grant pratique automatically — but Anchora flew it anyway, because traditions are traditions, and because a yellow flag on a small sloop entering a foreign harbor has a certain jaunty charm that she privately enjoyed.

Appendix C
Tidal Observations

Tidal data collected during the voyage is summarized below. All times are approximate, all heights are measured from chart datum, and all predictions should be treated with the same confidence one extends to a weather forecast — which is to say, they are probably correct, but you should not bet your keel on it.

Tides in the region are predominantly semi-diurnal, with two high waters and two low waters in each lunar day. The tidal range varies with the phase of the moon: greatest at springs (when the sun and moon are aligned) and smallest at neaps (when they are at right angles). Anchora recorded tidal observations at every anchorage, using a graduated staff driven into the seabed at the waterline, which she read at hourly intervals whenever she was aboard and awake, and at less regular intervals when she was not.

Twin Harbors. Semi-diurnal, range 1.8–2.4 m at springs, 0.8–1.2 m at neaps. High water approximately coincides with lunar transit, with a lag of roughly forty minutes that Anchora attributed to the constriction of the harbor entrance and the friction of the water flowing over the shallow bar at the approach channel. The two basins exhibit a 12-minute phase lag — high water in East Harbor occurs twelve minutes before high water in West Harbor — which the locals blame on a submerged rock formation in the cut between the two basins and Anchora blames on insufficient data. She measured the phase lag on three consecutive tidal cycles and got three different results: 10 minutes, 14 minutes, and 12 minutes. The average was 12 minutes. She reported 12 minutes. The truth, she suspected, was considerably more complicated than a single number could express, but a single number was what the chart demanded, and the chart was, for better or worse, what people would rely on.

The Reef. Tidal range similar to the Twin Harbors. The critical observation for the reef is not the range but the height of low water at springs, which determines how much of the coral is exposed. At LAT (Lowest Astronomical Tide), several of the larger coral heads break the surface, creating visible markers that are useful for daylight navigation but invisible — and therefore lethal — at night. Anchora marked these drying heights on her chart with the standard symbol: a depth figure with a line underneath, indicating that the number represents elevation above chart datum rather than depth below it.

Whirlpool Narrows. Tidal streams reach 4.5 kn at springs, 2.8 kn at neaps. The flood stream sets northward; the ebb sets southward. Slack water occurs approximately 15 minutes before local high and low water and lasts approximately 8 minutes at springs, 12 minutes at neaps. This is not a generous margin in either case, but it is sufficient for a transit if the vessel is prepared and the skipper decisive. Anchora recommended approaching the narrows under power rather than sail, on the grounds that an engine provides consistent thrust regardless of wind angle, whereas a sail provides consistent anxiety regardless of everything.

Compass Rose Atoll. Negligible tidal range inside the lagoon (0.3 m at springs, barely measurable at neaps). This is because the four passes, while wide enough for navigation, are narrow enough to attenuate the tidal wave as it enters the lagoon, spreading the rise and fall over a much longer period and reducing its amplitude to almost nothing. The practical consequence is that the lagoon maintains a nearly constant depth, which makes it an ideal anchorage in any conditions — a rare and valuable quality in an archipelago where most of the water is trying to go somewhere in a hurry. Currents through the passes reach 2 kn on springs but are predictable and well-behaved. Anchora described them in her notes as “the only polite water in the archipelago.”

Isle of Echoes. Semi-diurnal, range 1.6–2.0 m. The tidal stream along the cliff face sets to the north on the flood and to the south on the ebb, reaching 1.5 kn at springs. The stream is strongest at the headlands and weakest in the embayments, which is the normal pattern for tidal flow around a rocky island and was entirely predictable from the chart. What was not predictable was the acoustic effect of the tide on the cave systems at the base of the cliffs: at certain states of the tide, the incoming water compressed the air inside the caves and produced a low, resonant boom that could be heard from the anchorage, half a mile away. Anchora noted this phenomenon in her log but did not add it to the chart, on the grounds that “makes spooky noises at half tide” was not a recognised chart annotation.

Appendix D
Errata

No errors have been found. This is, in itself, suspicious.

Appendix E
Acknowledgments

The author wishes to thank the penguins.

Back Matter

Author’s Note

This section has no entry in the table of contents. If you have arrived here by paging forward from the appendices, congratulations: you are reading in spine order, which is the only reliable way to find content that the navigation has chosen to ignore.

The Author’s Note exists to test a specific edge case: a spine item whose table-of-contents entry points not to the beginning of the file, but to an anchor partway through it. Everything above the anchor is “dark matter” — present in the document, reachable by paging, but invisible to the TOC.

Consider, for a moment, what this means for a reading application. When a user taps “Colophon” in the table of contents, the application must navigate not to the start of this file, but to the colophon anchor partway through it. The content above the anchor — the text you are reading right now — exists in the spine, occupies space on the page, and can be reached by turning pages forward from the appendices. But it cannot be reached by any navigation element. It is, in the language of the EPUB specification, part of the spine order but outside the navigation document.

This creates an interesting problem for any reader that attempts to map TOC entries to page ranges. If the TOC says the Colophon begins at anchor #colophon, what is the page range of the content before that anchor? Does it belong to the previous TOC entry (Appendix C)? Does it belong to the Colophon? Does it belong to no entry at all? These are the kinds of questions that this test fixture is designed to provoke.

Anchora would have had opinions about this. Cartographers do not approve of places that exist but cannot be navigated to. An island without a name on the chart is an island that will, sooner or later, catch a keel.

But perhaps some content is meant to be found only by those who read sequentially, page by page, without jumping ahead. Perhaps the Author’s Note is a reward for patience. Or perhaps it is simply a test fixture dressed up in a narrative voice.

Either way, you have found it. Well done.

Colophon

This EPUB was generated by a Python script as a test fixture for the Crosspoint Reader project. It is not a real book, though it contains real sentences arranged in a real order, which is more than can be said for some publications that aspire to the title.

The spine of this EPUB contains fourteen items. Its table of contents contains eighteen entries. The relationship between the two is, by design, entertainingly non-trivial. Specifically:

Three TOC entries point to anchors within frontmatter.xhtml, making it a single spine item with multiple navigation targets. Four TOC entries (one parent and three children) point to anchors within chapter3.xhtml, and four more point into chapter5.xhtml, testing nested TOC hierarchies within a single spine item. Three TOC entries point to anchors within appendix.xhtml, the same pattern without nesting. Two of these appendix entries (D and E) are deliberately tiny — a single sentence each — to test how the reader handles TOC sections too small to fill a screen.

Chapter 2 occupies two spine items (chapter2_part1.xhtml and chapter2_part2.xhtml) but appears as a single entry in the TOC. Chapter 4 occupies three spine items and also appears as a single TOC entry. These test the case where a reader must determine that consecutive spine items belong to the same logical chapter.

The interlude (interlude.xhtml) appears in the spine between Chapter 3 and Chapter 4 but has no TOC entry whatsoever, testing the case where spine items exist outside the navigation hierarchy.

And this very file (backmatter.xhtml) has a TOC entry that points to the #colophon anchor, which is not at the beginning of the file. The Author’s Note above occupies the first portion of the file but is invisible to the TOC, testing the case where a navigation target lands in the middle of a document.

The cover image was generated programmatically using the Bookerly typeface at 536×800 pixels, rendered on a deep teal background (RGB 15, 55, 65) with light beige text (RGB 225, 220, 205).

No penguins were harmed in the making of this book. Several were mildly inconvenienced, but they bore it with their customary dignity.

Cover

urn:uuid:{BOOK_UUID} {TITLE} {AUTHOR} en {DATE} {DATE}T00:00:00Z

Table of Contents

Dedication

Epigraph

Foreword

Chapter 1Setting Sail

Chapter 2The Twin Harbors

Chapter 3Charting the Depths