The morning after Varro called his scribe I woke up at three forty-five and lay in the dark and thought about Marta.
Not in a weird way. In a strategic way.
See — the Edric problem had a shape to it. And once you understood the shape of a problem you could find the specific point where applying pressure actually mattered. Edric had seen something. He was going to report it to someone. I couldn't stop him from reporting. I couldn't confront him — that would confirm that there was something worth hiding. I couldn't make him disappear — I was eight years old and also not that kind of person, at least not yet, at least not for something this manageable.
What I could do was make sure his report was boring.
Not suspicious-boring. Not the-prince-clearly-told-everyone-to-act-normal boring. Genuinely, organically, convincingly boring.
Which meant the boring version of the fountain story needed to already exist — fully formed, technically credible, coming from multiple independent sources — before Edric had a chance to frame it as anything else.
I was at the kitchen door by six.
Marta was already there. She was always already there. I was fairly certain Marta had been in that kitchen since before the castle was built and would be there long after it fell.
"Your Highness," she said, not looking up from the dough she was working. "Breakfast isn't for an hour."
"I know. I wanted to talk to you about the fountain."
She looked up.
"The pressure mechanism," I said, settling onto a kitchen stool like I owned it. "Father's engineers installed it two years ago. Did you know about it?"
"...Can't say I did."
"Most people don't. It's subtle — the channel runs under the east courtyard stones, builds pressure from the natural water table, releases periodically. The engineering is actually fascinating." I leaned forward with the enthusiasm of a boy who genuinely found this fascinating, which I did, which was what made it work. "The gradient calculation alone — see, water doesn't need to be pushed if you give it the right angle. The Romans figured this out centuries ago. You build the channel at exactly the right incline and gravity does everything. Half a degree. That's the whole secret."
Marta was looking at me with the expression she used when she wasn't sure if she was being educated or managed and had decided to allow it either way.
"Is that so," she said.
"The release mechanism is what makes it look dramatic. The pressure builds and then — whoosh. Column of water, very clean, very controlled. I've been trying to work out the exact valve design but the stonework covers most of it." I shook my head admiringly. "Clever engineering. Eastern province style apparently."
"Mm." She went back to the dough. "Is that what happened yesterday at the fountain? With the princess?"
"She'd never seen it go off before. Got excited."
"Mm." A pause. "I'll admit it gave Edric a bit of a start."
"Edric?"
"The guard. Varantum boy. He came in here after his shift looking a bit — well." She made a gesture with floury hands that communicated unsettled without requiring the word. "I told him the children were always playing near the fountain. He seemed — I don't know. Jumpy."
"Probably the noise," I said. "The pressure release is loud if you're not expecting it."
"Mm. Yes, probably."
I stayed for another twenty minutes. Let her feed me bread and honey. Answered her questions about the drainage project with genuine detail. Made sure the fountain mechanism came up twice more naturally before I left.
Information planted.
Marta would repeat it. She repeated everything — not maliciously, just the way people who are the social center of a building share information, constantly and casually and with complete conviction. By dinner half the castle staff would know about the pressure mechanism. By tomorrow Edric would have heard the boring version from at least three separate people who had no reason to lie.
I walked back to my room in the early morning light feeling the specific satisfaction of a clean operation.
I didn't silence the witness, I thought. I recruited one. No confrontation. No suspicion. No damage. Sun Tzu would be proud.
Beat.
Sun Tzu would also probably point out that I shouldn't have let my five year old sister publicly levitate water in front of a Varantum plant in the first place.
We don't talk about Sun Tzu's second point.
And then I built things.
Three days. Head down. No chaos — or minimal chaos, Lily was around but Biscuit had apparently decided to live in her room now which meant her primary mission was no longer catching him and my measuring rope was safe.
Three days of actual work.
Let me tell you about it.
Okay. Right. I'm going to stop here for a second and talk to you directly.
You — yes, you. The one reading this.
I know some of you are here for the part where I build weapons and destabilize a centuries-old empire and watch the geopolitical balance of an entire continent rearrange itself around one small kingdom that nobody took seriously. That part is coming. I promise. We are approximately — checking notes — thirteen percent into the plan. Maybe fourteen. Fourteen if you count Lily, which I always count Lily.
But this part matters. This part is the foundation — literally, in the case of the drainage system, and figuratively in every other case. You don't build an empire on good intentions and a couple of flintlocks. You build it on infrastructure. On production capacity. On the completely unglamorous, uncinematic, absolutely necessary work of making a small broken kingdom function properly before you ask it to function extraordinarily.
So. Architecture. Real architecture. I'm going to explain what I built and why it matters and I'm going to use references that exactly zero people in this world will understand but that I'm reasonably confident you will.
If you are genuinely not interested in medieval engineering and its relationship to Roman hydraulics, industrial revolution metallurgy, and nineteenth century urban planning — I respect that, truly, different people have different needs — scroll down until you see the word COURIER in large letters. We'll pick up there and I won't hold it against you.
For everyone else.
Buckle up.
This is the part where I explain how a half degree gradient and a better bellows system are the first two steps in dismantling an empire.
It's more exciting than it sounds.
I promise.
THE FORGE.
Here's what Aldous had when I found him.
A good forge. Genuinely good — stone construction, solid dimensions, properly positioned relative to the prevailing wind. The man knew his craft. Twenty years of experience in every callus on his hands.
What he didn't have was airflow.
The bellows were original to the forge — old design, big, moved a lot of air. And that's exactly the problem. Moving a lot of air and moving the right air are two completely different engineering propositions and most people in the history of metallurgy have conflated them until someone comes along and explains the difference.
Think about the Bessemer converter. 1856. Henry Bessemer — English engineer, genuinely one of the great practical minds of the industrial revolution — figures out that if you blast air through molten pig iron at the right pressure and rate you don't just heat the metal, you transform it. You burn off the impurities. You control the carbon content. You turn inconsistent iron into consistent steel in twenty minutes instead of days.
The key word is right. Not more air. Not less air. The right air at the right rate at the right pressure.
Aldous's bellows were like trying to fill a swimming pool with a fire hose — technically moving water, technically doing the job, but no control, no consistency, no ability to modulate output based on what the material actually needed.
What I designed was closer to what the Sheffield steel mills figured out in the 1700s — not the technology, obviously, we didn't have the technology, but the principle. Smaller, faster bellows movement. Adjustable aperture on the air channel. A secondary chamber that regulated pressure before it hit the fire so you could tune the burn temperature without rebuilding the whole setup from scratch.
The frame took three days to build. Aldous's apprentice Cole — seventeen, quiet, the kind of sharp that doesn't announce itself — figured out the aperture mechanism faster than I expected. I showed him the principle once. He built it correctly the second time.
Filed, I thought, watching him work. That one gets filed.
When we ran the first test fire with the new bellows the temperature hit where I needed it in half the time of the old setup. Aldous stood back and looked at it for a long moment.
"Hotter," he said.
"More consistent," I said. "Temperature variance under the old setup was probably fifteen to twenty percent run to run. This should be under five. That's the difference between iron that performs to spec and iron that surprises you."
He looked at me.
"Surprises you," he repeated.
"In metalworking surprises are bad."
He went quiet for a moment. Then he turned back to the forge and said — "Show me the secondary chamber again."
I showed him.
He didn't say much after that. Just worked. But there was something different in the way he moved around the forge now — not pride exactly, more like a man who had been given a better tool and was already thinking about what he could make with it.
That was enough.
THE DRAINAGE.
The east courtyard flooded every spring.
Had been flooding every spring for thirty years. I'd checked the maintenance records — three decades of the same entry, slight variations: east courtyard — standing water — resolved. Resolved meaning someone came with buckets and removed it manually. Every year. For thirty years.
Nobody had asked why.
I asked why on day one. Spent an hour in the courtyard with a level — had to explain to the steward what a level was, which took longer than using it — mapping the gradient of the stone floor.
The result was exactly what I expected. The courtyard sloped wrong. Not dramatically — nobody would notice walking across it — but enough. The stones had been laid almost flat with a very slight incline toward the east wall instead of toward the drainage channel on the south side. Probably a mistake in the original construction. Probably nobody had checked the gradient because the deviation was less than a degree and a half and you couldn't see it with the naked eye.
A degree and a half. Thirty years of spring flooding.
The Romans would have wept.
Actually — let me tell you about Roman drainage for a moment because it genuinely deserves the digression.
The Romans were obsessed with water management in a way that most civilizations weren't. Not because they were uniquely brilliant — though they were very good — but because they understood something fundamental that kept getting forgotten and rediscovered throughout history: water is free labor if you treat it correctly. It moves. It moves constantly, downhill, forever, without being asked. Your entire job as an engineer is to point it where you want it to go.
The aqueducts people remember — eleven of them feeding Rome at the empire's peak, moving hundreds of millions of gallons daily across dozens of miles. What people forget is that the engineering principle behind all of them is almost insultingly simple. Gradient. Consistent, carefully calculated, ruthlessly maintained gradient. The Aqua Claudia dropped roughly ten meters over seventy kilometers. That's it. That slight, sustained slope was enough to move water from the hills into the city entirely under gravity.
Half a degree. Sustained. Consistent. That's the whole secret.
The Cloaca Maxima — Rome's great sewer, built six centuries before Christ — was running functional drainage under the city for over a thousand years. Not because of magic. Not because of technology nobody else had. Because someone did the math and graded the channels correctly and built them from good stone.
They figured this out. Then the empire fell. Then somehow, inexplicably, mysteriously, most of Europe spent several centuries just — not doing it. Accepting standing water as a feature of urban life rather than a solvable problem.
Baron Haussmann understood this when he rebuilt Paris in the 1850s — tore up the medieval street plan and underneath it built a sewer network that finally, finally applied the Roman principle properly to a modern city. Turned a disease-ridden labyrinth into a functioning metropolis in under twenty years.
The solution to the east courtyard was the same solution it's always been.
Regrade the stones. Forty-seven of them needed to come up and go back down at the correct angle. I had the numbers. I gave them to the steward. He looked at them for a while and then looked at me and said — "we've been flooding for thirty years."
"Yes," I said.
"And the solution is just — the angle."
"The angle," I confirmed.
He was quiet for a moment with the specific expression of a man who was doing the math on thirty years of bucket work.
"Get the stones up," he said.
They were up by afternoon. Relaid by the next morning. Correct gradient, properly sealed, drainage channel cleared and extended six feet to handle peak flow.
Done.
Thirty years. Forty-seven stones. One afternoon.
The workers who relaid them watched me the whole time with an expression I was starting to recognize — the recalibration look, the one where what they expected and what they were getting didn't quite match. An eight year old prince standing in a courtyard directing drainage work with a level and a page of calculations.
I heard one of them say to another — "where does he learn this?"
The other one shrugged. "He reads a lot, apparently."
Good, I thought. Let that be the story.
THE GATE.
This one was personal.
Not because of the engineering — though the engineering was satisfying — but because of what the gate represented. Every time I looked at it I thought about the same thing.
Braveheart's main gate was the first thing Varantum's people saw when they arrived. The first thing any army would see if it ever came to this courtyard in a less diplomatic capacity. It was the kingdom's face and its first line of defense and it was, structurally, a disaster waiting to happen.
Single chain. Single counterweight. Single pulley at the apex. The portcullis — iron-reinforced oak, reasonably solid — hung from one load-bearing point that transferred the entire weight through one chain that was showing wear at the third link from the top.
Single point of failure.
I have a pathological hatred of single points of failure. It goes back to a project in my first life — a defense component we'd built for a client that failed catastrophically in field testing because one pin in one mechanism had been machined two millimeters under spec and nobody had built in the redundancy to catch it. Nobody got hurt. The contract got cancelled. I spent three weeks unable to sleep properly because I kept running the scenario where it had been deployed and someone had been on the other end.
Since then — redundancy. Always redundancy.
John Roebling understood this when he designed the Brooklyn Bridge in 1869. The cables — those massive, iconic cables everyone photographs — each one is built from thousands of individual steel wires bundled together. Not because any single wire is strong enough, but because you never rely on any single wire. If one fails — and eventually something always fails — the load distributes across the rest. The bridge doesn't fall. The system absorbs the failure and continues.
That's engineering. That's real engineering. Not making things that don't break. Making things that fail gracefully when they eventually, inevitably do.
The gate captain — a man named Brynn, solid, practical, suspicious of eight year olds with opinions about his gate — watched me examine the mechanism for twenty minutes before he said anything.
"Something wrong with it?" he asked.
"Third link on the main chain," I said. "How long has it been like that?"
He looked. Came closer. Looked again.
"...Three months maybe."
"Replace it this week. But that's not the main problem." I pointed up at the single pulley. "One point. Everything goes through one point. You need three — apex and two lateral, load distributed across all of them with independent chains. If anything fails the other two hold. You're never one broken chain away from the portcullis dropping."
He stared at the mechanism for a long time.
"The portcullis dropping," he said slowly. "While the gate is open."
"While anyone is standing under it."
Long silence.
"I'll need the specifications," he said.
I gave him the specifications.
The redundant pulley system took two days to install. Three chains, three load points, weight properly distributed. Da Vinci had sketched versions of this exact mechanism in his notebooks — the man who spent half his life drawing machines that wouldn't be built for centuries, who understood mechanical advantage with an almost supernatural clarity, who would have looked at Braveheart's original gate and felt personally offended.
When it was done Brynn tested it six times in a row. Each time the portcullis rose and descended smoothly, load distributed, no single point of stress.
On the sixth test he put his hand on my shoulder briefly. Said nothing. Went back to his post.
I took that as a five star review.
And that's the architecture.
Forge, drainage, gate. Three projects. Twelve days of visible, credible, completely legitimate engineering work that established the cover, improved the kingdom, and demonstrated — to anyone watching — that the young prince of Braveheart was exactly what he appeared to be.
A boy who liked fixing things.
Nothing more.
Now — back to the story.
COURIER.
Father asked to see the projects on day twelve.
Not officially — not a royal inspection, nothing formal. Just George, in his comfortable clothes, wanting to see what his son had been doing. I walked him through all three in the late morning, Ren two steps behind, the castle going about its business around us.
Forge first. Aldous showed father the new bellows with the quiet pride of a craftsman who had been given better tools and wanted the right person to appreciate it. Father listened, nodded, asked questions that were more about the people involved than the mechanism — how long has Aldous been with us, is Cole the boy from the lower quarter, good family — which was very him, very George.
Drainage second. I showed him the gradient correction, explained the principle, pointed at the dry courtyard stones that had been underwater every spring for three decades.
He stood there for a moment.
"Thirty years," he said.
"Forty-seven stones," I said. "One afternoon."
He was quiet in the particular way he went quiet when something moved him and he didn't quite have words for it.
Gate last. Brynn demonstrated the new mechanism. Three chains, smooth operation, no single point of failure. Father watched it go up and down twice and then turned to look at me with an expression that I had learned to recognize over eight years — pride trying to find an exit through a face that wasn't built for containment.
We sat in the courtyard afterward. Just the two of us, Ren having tactfully found something to do elsewhere. The sun was properly warm. Biscuit was somewhere nearby — I could hear Lily talking to him around the corner with the one-sided intensity of someone conducting an important interview.
Father looked at his hands for a moment. Then at the courtyard. Then at me.
"Your grandfather would have been proud," he said.
He meant the old king. His father. A man who'd died when I was three, who I had almost no memory of in this life — a face in a portrait in the west corridor, serious eyes, a soldier's posture.
But I didn't hear it that way.
I heard it in a different voice. Lower. Rougher. With the particular cadence of a man who had grown up in Chicago and never quite lost it regardless of how many years passed.
Raymond Keller. Retired Staff Sergeant. Hands like catcher's mitts.
Put it back together, Davey.
I looked at the courtyard. At the dry stones. At the gate with its three chains catching the morning light.
"Yeah," I said. "I think he would have."
Father put his hand on my shoulder. Left it there for a moment.
Neither of us said anything else.
It was enough.
That evening I updated Phase 1 with the progress markers and felt for the first time since the Varro situation that things were moving the right direction. Cover solid. Projects visible and real. Aldous on board, Cole identified, Brynn a quiet ally, Marta's network carrying the right stories.
The courier had left ten days ago.
Varantum was six days ride away.
Which meant whatever was going to happen — whatever Edric had written, whatever Varro had sent — was already in motion somewhere down that road and I wouldn't know the outcome for another week minimum.
Nothing to do but build, I thought. So build.
I was halfway through a new set of calculations — forge output projections, what we could realistically produce in six months with the expanded capacity, what that meant for Phase 2 — when the noise started.
Subtle at first. Footsteps in the corridor that shouldn't be there at this hour. Voices, low and quick. The specific rhythm of a castle that had received unexpected information and was redistributing it through its staff the way a body responds to a sudden temperature change — automatically, quickly, all systems responding at once.
I put my pencil down.
Ren was already at the door. He slipped out. Came back in four minutes with the face.
The face.
"Varro," he said. "He's packing."
I looked up slowly.
"Packing."
"Everything. His whole retinue. Chambers being stripped. They're leaving tonight."
"Tonight."
"The order came—" Ren paused. Something in his expression shifted. "Nick. The courier came back."
The room went very quiet.
I ran the numbers automatically. The capital was six days ride. A courier going and returning in ten days meant — four days travel, one day at the capital, four days back. One day. One day in the capital before being turned around with a recall order.
Whatever happened in Varantum had happened fast.
"What do we know," I said. Not a question. A prompt.
"Nothing yet. The staff only know he's leaving. Nobody's saying why."
I stood up. Went to the window.
The courtyard below was alive with torchlight — Varro's retinue moving with the controlled urgency of people who had been given an order and a timeline and were executing both. Carts being loaded. Luggage carried down. The efficient, practiced movement of a household that had done this before, that traveled as a unit and knew its own logistics.
Six years, I thought. Six years Varro had been in this castle. He knew where the kitchens were, which servants talked, which nobles could be cultivated, how the king's schedule worked. He was furniture in this court — present, permanent, woven into the fabric of the place.
And something had pulled him out in one day.
A decaying empire, I thought slowly. Rotting from the inside. Corrupt nobles bleeding the treasury. Generals fighting each other for influence. An emperor holding it together through reputation rather than actual strength.
What happens when the reputation cracks?
What happens when one of those corrupt nobles makes a move? When one of those generals decides his moment has come? When the thing holding the center together stops holding?
What happens to the empire's outposts when the empire starts looking inward?
Ren was very still beside me.
"Ren," I said quietly. "What do we know about Varantum's internal politics?"
"Almost nothing."
"That's going to change." I turned from the window. Went to the desk. New page. At the top, large letters — VARANTUM: INTERNAL.
I stared at it.
Then stopped.
Looked at the door.
"Is Edric leaving with him?"
Silence.
I turned to look at Ren. He was already looking at me with the expression of someone who had just caught up to where I was and didn't love what he found there.
Because we'd both been thinking about Edric as Varro's problem. Varro's plant, Varro's network, Varro's responsibility. A variable that existed within the ambassador's orbit.
But Varro was leaving.
"If he goes," Ren said carefully, "he takes the network with him."
"And if he stays—"
"He becomes—"
"Ours," I said. "He becomes ours."
Not in the sense that I could use him. Not yet, not easily. But in the sense that a Varantum plant operating in this castle without his handler was a different kind of problem entirely — unmanaged, potentially unpredictable, reporting to someone in a capital that was apparently too busy with its own crisis to pay close attention.
Or.
Potentially — potentially — an asset waiting to be flipped if the right moment came.
I filed that thought in the very back of the drawer labeled NOT YET.
I looked back at the window. Below, Varro was emerging from the castle doors — travel cloak, rings, the face that never showed what it was thinking. He crossed the courtyard with the unhurried pace of a man who was leaving on his own terms even when he wasn't.
He paused in the courtyard.
It was brief — a second, maybe two. His gaze moved across the castle walls and windows with the automatic sweep of a man who had spent six years reading this building and its inhabitants and was doing it one last time.
His eyes reached my window.
I stepped back into the shadow. Waited.
His gaze moved on.
He got into the carriage. The door closed. The driver called to the horses.
The gates — my gates now, with their three chains and their redundant load distribution and their replaced third link — opened smoothly.
The carriage rolled through.
Gone.
The torchlight in the courtyard diminished as Varro's people followed. The gates closed. The castle settled back into its nighttime quiet like water closing over a stone.
I stood at the window for a long time.
First level boss, I thought. Recalled before the fight.
Which means one of two things.
Either I won.
Or the game just changed and nobody told me the new rules.
I turned back to the desk. Picked up my pencil. Looked at the page.
VARANTUM: INTERNAL.
Below it I wrote — WHY NOW.
Below that — EDRIC: STATUS.
Below that — PHASE 1: ACCELERATE.
I underlined the last one. Twice.
A decaying empire just flinched, I thought, pencil already moving. I don't know why. I don't know what it means. I don't know if it helps me or hurts me or changes nothing.
But I know one thing.
When the cat looks away — you move.