The survey shelter was a low stone structure built into the lee of the ridge, its roof reinforced with iron struts and its single door braced against the wind by a heavy oak beam. Inside, a cast-iron stove glowed low and orange, throwing uneven light across the walls where old survey maps were pinned in chronological rows.
Vane sat at the wooden table, his ledger open before him, the 412/449 discrepancy written in crisp black ink. He had been staring at it for the better part of an hour.
Brandt stood by the stove, his heavy coat draped over a crate, his thick forearms resting on his knees. He had already suggested checking the chain twice and recalibrating the link counts once. Vane had agreed both times. The chain was straight. The links were true.
"That stone didn't move," Brandt said, not for the first time. "I drove the pin myself. Four inches into solid bedrock. If the marker shifted, we'd see it at the base."
"We don't see it," Vane agreed.
"Then the ledger's wrong."
Vane didn't answer immediately. He turned the ledger toward the stove light and ran his thumb down the column of entries. Marker E-16 to Marker E-17, 412 yards. It was dated six years ago, signed by the Chief Registrar's office. The ink was clean, the handwriting precise. There was no sign of erasure, no marginal correction, no appended note.
"It's a clean entry," Vane said. "No marks. No cross-references. It's the only distance recorded for that leg."
Brandt grunted. "Then who measured it?"
Vane looked at the name on the bottom of the page. The signature was illegible, a compressed scrawl that could have belonged to any one of a dozen surveyors who had rotated through the Eastern Spur over the past decade. He would have to pull the original filing to confirm the hand.
"The secondary bearing," Vane said, tapping the margin of the ledger. "The map was drawn from two separate triangulation points. If the marker had been placed incorrectly, the secondary bearing would have been off by at least three degrees."
"And was it?"
Vane turned the page. He had already checked. The secondary bearing was recorded as consistent with the primary measurement—a neat, mathematical confirmation that E-17 was exactly where it was supposed to be.
The problem, he realized, was that the ledger's secondary bearing matched the ledger's primary measurement. The 412-yard mark. It was an internally consistent record.
It just didn't match the stone.
"The marker's been there for twenty years," Brandt said, sitting back heavily. "If the Academy measured it at 412 and the stone says 449, then someone made a mistake. That's the simplest answer."
Vane didn't argue. He agreed with Brandt. The simplest answer was a measurement error. But the chain was calibrated. The anchor was straight. The secondary bearing was consistent with the 449-yard measurement Vale had recorded, not the ledger's 412.
If someone had made a mistake, Vane couldn't find where.
He stood, walked to the map wall, and pulled down the oldest survey sheet he could find—the original academy boundary map from the Eastern Expansion, dated forty-seven years ago. The parchment was yellowed and fragile, the ink faded to brown. He laid it flat on the table and weighted the corners with his ledger and a stone inkwell.
E-17 was marked on the map, but the distance label was missing—obscured by a dark, age-worn stain that had bled through from the reverse. The second bearing point was similarly faded. The only legible detail was the marker's position relative to the ridge line.
"The old map doesn't have the measurement," Vane said. "It only shows the marker's position."
"Then we don't know what it was supposed to be," Brandt said.
"We don't know what the surveyors measured," Vane corrected. "But the marker itself is the primary record. The measurement is just a note."
Brandt frowned. "You're saying the ledger is the mistake, not the stone."
"I'm saying we don't know which is the mistake," Vane said. "And we don't know what caused it."
He gathered the papers and began the slow work of compiling a formal discrepancy report. He wrote the date, the location, the names of the survey team, and the archival reference numbers. He recorded the archived distance and the current verified distance. He noted the instrument calibration, the anchor displacement check, the terrain inspection, and the secondary bearing comparison.
Then he paused over the classification field.
There was no appropriate box.
This wasn't a failed measurement. It wasn't a damaged anchor. It wasn't a recorded data entry error that could be corrected by cross-referencing another source. The discrepancy was a genuine difference between the official record and the physical world.
Vane wrote "Pending Review" and set the report aside.
Brandt had gone out to check the horses. The shelter was quiet save for the crackling stove and the distant wind against the stone walls. Vale sat on a crate near the door, his brass optical frame folded across his knees, polishing a smudge from the lens with the edge of his sleeve.
Vane watched him for a moment. Vale had been quiet throughout the evening, but Vane had noticed him studying the old map with unusual attention—tracing the faded ink lines with his finger, comparing them to the current survey drawings without being asked.
"You noticed the reflection before the others," Vane said.
Vale looked up. "The light was wrong."
"Wrong how?"
Vale didn't answer immediately. He folded the lens cloth and set it aside, his gray eyes calm and distant.
"When I looked at the marker from the ridge, the reflection didn't line up with the surface it was bouncing off of," he said. "The stone is polished. The face catches the northern sky and throws a pale glare onto the rock wall behind it."
Vane nodded. He had seen the glare himself.
"The glare should be centered on the stone face," Vale continued. "It's not. It's shifted six yards east. The light is coming from somewhere else."
Vane sat forward. "Where?"
Vale didn't look at him. He was staring at the old map again, his finger tracing the faded ink line of the ridge.
"I don't know where it's coming from," Vale said. "I know where it looks like it's coming from."
Vane waited. Vale didn't elaborate.
"Which marker are we measuring?" Vale asked.
Vane blinked. "E-17."
"Yes," Vale said. "The stone we're measuring now, or the point the original surveyors were measuring?"
Vane frowned. "It's the same marker."
Vale looked at the old map, his expression unchanged.
"Yes."
He didn't press the point. He didn't offer a theory. He simply folded the optical frame under his arm and stood, pulling his coat tight.
"I'll check the northern pin before the frost sets," he said. "The cold will shift the ground by morning."
He stepped out into the wind, leaving Vane alone at the table.
Vane looked down at the report. The classification box stared back at him, empty.
He signed his name at the bottom, slid the parchment into a leather folder, and placed it in the outgoing records tray. The folder sat there, ordinary and unremarkable, ready to be carried back to the Academy with the morning courier.
Vane didn't know what category of problem this was. He didn't know whether the discrepancy was a measurement error, a mapping error, or something else entirely. He only knew that he had recorded what he had found, and that somewhere, eventually, someone would read it.
The wind picked up outside, rattling the iron struts. Vane stood, banked the stove, and prepared to wait for Brandt's return. The report sat in the tray, unresolved, waiting for a classification that didn't yet exist.