Chapter 34: "The Night Nothing Happened and That Was Worse"

Archive entry, Volume 24 (of 23). Field note 1,411. Compiled from direct observation, subsequent physical evidence review, and one episode of what I will classify, for lack of a better taxonomic framework, as dread.

The tower was Thrain's idea. I note this not to assign blame — though blame will, as always, find its correct address before this entry concludes — but to establish the decision chain with the precision future scholars will require when they attempt to understand how two experienced travelers slept inside someone else's observation post and called it shelter.

We had been walking north from the collapsed waystation on Crab-Tooth Ridge for eleven hours. Thrain had not spoken since hour two. This was not unusual. What was unusual was the quality of the silence, which had shifted from his standard operational muteness into something denser, something that suggested internal processing. I had noted the waystation's condition in Chapter 33. Thrain had noted it by walking away from it. We were, by that point, two individuals moving through open terrain with forty-two confirmed hostile institutional responses behind us, an unspecified number of underground operatives ahead or beneath us, and no allied faction within any distance I cared to calculate.

The tower appeared at dusk. Three leagues north of the burned waystation, set against a ridge of shale that the Silt Marches use as a boundary marker for the territory nobody wants. Round. Stone. One door. No windows below the third story. The roof had partially collapsed inward but the walls held.

—There, Thrain said.

I assessed the structure from forty paces. No light. No smoke. No tracks visible at this distance, though the shale would not hold tracks well. Single point of entry, as Thrain had identified. Isolated from patrol routes — the nearest Tidal Council checkpoint was nine leagues southeast and had likely been abandoned after Chapter 33's jurisdictional consequences rendered its authority ambiguous.

Methodological observation: Thrain's tactical reasoning was, on paper, correct. Isolation. Defensibility. Jurisdictional irrelevance. I recorded these criteria in my notebook at the time. I have since drawn a line through them.

We entered at dusk. The ground floor was bare stone, swept clean by wind through the open door. A spiral staircase, partially eroded, led upward. Thrain checked the second and third floors with the hammer drawn. I remained below, listening. He descended after four minutes.

—Empty.

—You verified all rooms?

—Two rooms. Both empty.

—And the roof?

—Open sky. No one.

I recorded: Tower entered 19:40 approx. Structure cleared by subject. Two habitable floors. Roof compromised. Single ground-level access point. Assessment: adequate for overnight defensive rest.

We settled on the second floor. Thrain positioned himself against the wall adjacent to the stairwell, hammer across his knees. I set my notebook on a stone ledge that might once have been a window seat, now sealed with rubble. We ate cold rations. Thrain drank from his flask. I declined, as I have declined for four years.

I slept from approximately the third hour past midnight until the fifth. Thrain, I believe, did not sleep at all, though he would not confirm this. Dawn arrived grey and gradual, filtering through the gap in the roof above us.

I descended to the ground floor to relieve myself outside and found the first note.

It was pinned beneath a stone just inside the doorway. Not hidden. Placed. The paper was of a quality I did not recognize — neither the pressed reed-stock of Tidal Council administrative supply nor the treated cotton of Aureate Dominion correspondence. The ink was uniform. Machine-applied or written with extraordinary steadiness.

The text read:

Day 1. Subjects depart Crab-Tooth Ridge waystation, 07:12. Northwest bearing. Dwarf leads. Gnome follows at four paces, adjusting to three when terrain narrows. Rest stop at 11:40 beside split boulder. Duration: fourteen minutes. Gnome looks east twice during rest. Dwarf does not eat. Flask accessed once.

I read it twice. Then I turned the paper over.

Day 2. Bearing holds northwest. Pace decreases after hour six — gnome's stride shortens by approximately two inches, consistent with fatigue pattern observed Day 1. Water source accessed at 15:22. Forty-seven steps from trail to stream bank. Gnome fills one container. Dwarf fills flask. Departure 15:31.

Day 3. Subjects arrive at tower, 19:40. Dwarf clears interior. Gnome remains at ground level for four minutes, six seconds. Second floor selected. Dwarf does not sleep.

I set the paper down. I picked it up again. I counted the details. The number of steps to the water source. I had not counted them myself. Forty-seven. I did not know if this was correct. I did not know how to feel about not knowing.

I walked outside. The morning air was flat and cold. Twenty paces from the door, behind a low shelf of shale, I found the second document. This one was longer. It contained a diagram of our two previous rest locations — the overhang where we had sheltered during the brief rain on Day 1, and the depression beside the dry creek bed on Day 2. Both were marked with sightlines. Angles drawn in precise arcs from elevated positions to our exact resting spots. Observation posts. Two per site. Fields of view overlapping to eliminate blind spots.

I returned to the second floor. Thrain was standing. He had heard my pace change on the stairs.

—Problem, he said. Not a question.

I handed him both documents. He cannot read gnomish, but these were not in gnomish. They were not in dwarvish, common trade script, or the Aureate Dominion's formal register. I did not recognize the language. The numerals, however, were universal.

—We were watched, I said. Three days minimum. Two observation posts at each of our previous stops. They documented our movements to the step count. They knew when you accessed your flask. They noted the direction I looked when I was startled by that bird on Day 1.

—I wasn't followed.

—No. You were not followed. You were preceded. Whoever placed these had people at locations we had not yet chosen to stop at, or they repositioned faster than we traveled, or — and this is the possibility I find least comfortable — they knew where we would stop before we decided to stop there.

Thrain looked at the documents for a long time. His hands did not move. His jaw worked once beneath the split beard, a single clench that traveled from the hinge to the chin and stopped.

—For the record, I said, because the question was necessary even though the timing was, as always, wrong, —at what point during your tactical assessment of this tower did you consider the possibility that its suitability was itself the reason to avoid it?

He did not answer. He set the papers on the floor. He picked up his hammer. Then he set it down again.

I had not seen Thrain set the hammer down twice in succession before. I noted it. Subject placed weapon down, retrieved it, and replaced it within a span of eight seconds. First recorded instance of indecisive weapon handling in four years of observation. Possible interpretation: the absence of a target to strike has produced a motor-planning disruption.

—Which direction, I asked.

Thrain stood at the top of the stairwell and did not descend. His boots did not move. His weight stayed centered. I have documented Thrain in thirty-two chapters of catastrophe, and in every one of them the failure mode was action — too fast, too committed, too irreversible. This was different. This was Thrain understanding, perhaps for the first time in the period of our association, that the next step he took would be observed, anticipated, and recorded by someone whose system he could not reach with a hammer.

Five minutes passed.

—They want us to find this, I said.

Nothing.

—The placement inside the door. The diagram left in the open, twenty paces out. These are not mistakes. This is communication. They are telling us they were here. They are telling us they chose to let us know.

—Why.

One word. I wrote it down. It was, I realized, the first genuinely investigative question Thrain had asked me in four years.

—I don't know. I logged this answer with the same notation I use for all gaps in my data: insufficient information, collection method unavailable.

The silence lasted another three minutes. Outside, the wind moved across the shale with the sound of something patient.

—We go south, Thrain said finally.

South was wrong. South was back toward Crab-Tooth Ridge, toward the jurisdictions now hostile, toward the burned waystation and every consequence we had been walking away from. South was the one direction inconsistent with every movement pattern we had established over three days. It was, in the logic of evasion, the only bearing that broke the model.

I understood this. I did not say so. Thrain did not need agreement. He needed to move in a direction no one had written down yet.

We descended. We left the documents where they lay. Thrain paused at the door, looked north — the direction we should have continued — and then turned south and walked. I followed at four paces. Then I adjusted to five.

Marginal note: the adjustment to five paces was deliberate. If the pattern they recorded was four, I would give them four no longer. This is not strategy. This is the behavioral equivalent of slamming a door. I am aware it changes nothing.

Official register, Chapter 34. Confirmed casualties: zero. Structures destroyed: zero. Factions offended: zero, to my knowledge, though my knowledge has been demonstrated to contain significant gaps. New data: an organized entity of unknown affiliation, unknown size, and unknown intent has been tracking our movements with institutional-grade precision for a minimum of three days. Their documentation was thorough, accurate, and positioned for discovery. Thrain's response was immobility followed by a course reversal. My response was to change the distance between us by one pace. Both responses are, by any rigorous analysis, inadequate. I recorded them anyway. The pencil does not care whether the data is useful. Neither, at this point, do I.