He bought a large paper map of Prague from a tourist shop near the Astronomical Clock. The kind tourists used fold-out, color-coded landmarks labeled in four languages. He had no use for the landmarks. What he wanted was the geography. The street grid. The river. The way the city sat in relation to itself, district by district, was the organic sprawl of something that had grown over centuries rather than been planned.
He spread it on his kitchen table and weighted the corners with his coffee mug, his keys, a spare lens cap, and the paperback he had been reading for three weeks without making meaningful progress. Then he got out his notebook, the narrow one, not the work log, and started transferring.
Every location where the hum had been present. Every location where it had intensified beyond baseline. He worked from memory first, then cross-referenced against the dated entries in his notebook, which went back to the first week after he arrived in Prague, when he had noted the sensation without yet having a name for it. Just written: strange acoustics in the basement of the Týn building, low vibration, possibly structural. Check foundations.
He had checked the foundations. They were fine.
That entry was seven months ago. There were forty-three subsequent entries, not all of them significant; in some locations, the hum had been faint, possibly incidental, not reproducible on a second visit. He filtered those out. What remained were nineteen confirmed locations where the response had been clear, consistent, and specific to the site rather than to him in transit.
He marked each one with a small red dot. Got the dots on the map. Then stepped back and looked at the whole thing from a distance, the way you look at a wall to see its overall condition rather than its individual stones.
The pattern wasn't obvious at first. Nineteen red dots are scattered across a city of a million people. Some clustered near the center, some out in the residential districts. He looked at it for a while without seeing anything useful, and then made more coffee and ate two biscuits from the packet on the counter, and looked again.
The river.
That was the first thing. Not all of the dots were close to the river, but the strongest ones were. He went back through the notebook, found his notes on intensity, and put a circle around the high-intensity locations. Seven of them. He drew a rough line through those seven dots, connecting them loosely, and then he looked at the line and where it ran in relation to the Vltava.
Parallel. Always within he measured with his thumb against the map's scale approximately two hundred to three hundred meters. On both banks. Never further. The river seemed to be, not a source exactly, but a boundary. An edge that the hum oriented itself around.
He sat down and thought about this.
Then he pulled out his laptop and opened the city's historical architecture database, which he had full access to through his restoration consultant credentials. He typed in the addresses of his seven high-intensity locations and ran them through the dating filters.
Pre-1600. All of them. Every single one.
He sat with this result for a minute. Then he ran the other twelve locations through the same filter. Nine of the twelve were also pre-1600. The remaining three were post-1600 buildings built on pre-1600 foundations. Which, if you were being precise about it, meant the stone beneath them was old even if the stone around them wasn't.
So, all confirmed locations. Pre-1600 stonework, or built on pre-1600 foundations. All within three hundred meters of the Vltava. He looked at the map with its red dots and its circled cluster and thought about what kind of thing that described.
A boundary. Or the remnants of one.
Something that had been built, not grown, not accumulated, but deliberately constructed along the river and around the historic center. Something that was still here, in some form, in the stone. Still present enough to produce a physical response in what, exactly? In him. In his body. Through his feet and his teeth and the bones behind his eyes.
He wrote in his notebook: Not random. Not acoustic. Something built into specific pre-1600 stonework, concentrated near the Vltava, is present across at least nineteen locations in the historic center. The pattern suggests a network rather than an isolated anomaly. Deliberate construction? Purpose unknown.
He underlined deliberately.
Then he wrote underneath: What was it for?
No question mark. He wrote it as a statement because it wasn't really a question yet. It was the shape that a question would eventually take. Right now, it was just the fact of the thing that something had been put here, into these stones, on purpose, and that he could feel it, and that as far as he could determine nobody else could, or nobody else was saying they could, which was not the same thing.
He went back to the map.
He had nineteen confirmed dots. He looked at the gaps between them, the spaces in the network where there was nothing marked. Some of those gaps were probably just places he hadn't been yet, or hadn't been in conditions where the hum was detectable. But some of them felt structural. Necessary spaces rather than absences. He thought about the way a wall's weight distribution worked the load paths through stone, the points where force concentrated, and the points where it dispersed, and applied that logic loosely to the map.
If this were a network, there would be nodes. Points where lines of whatever it was converged. He found three candidate locations by following the geometry of the marked dots outward to where the lines between them would intersect. Crossed them against the architectural database.
First candidate: a building on Malostranské náměstí, 16th century, documented. Check.
Second: the Old-New Synagogue in Josefov, one of the oldest continuously used buildings in Europe. Obviously pre-1600. Obviously, on the river's eastern bank. He had not been inside yet. He made a note.
The third candidate located the geometry of the network, if it were a network, suggested a point somewhere beneath the Nusle Bridge, which was not a pre-1600 structure. But the Nusle valley had been inhabited since the medieval period, and there were recorded foundation remains from several periods in the archaeological surveys. Something under the bridge. Possibly under the valley floor itself.
He circled all three candidates on the map.
Nineteen confirmed locations. Three probable nodes.
He started calling them nodes in his notes, because the word had a useful neutrality, it described a point in a network without committing to any particular theory about what the network was or what it did. He was careful about commitment in his notes. Commitment was premature. You described what you had, and you described it precisely, and you did not reach for the word that felt right when you hadn't yet earned it.
He took a photograph of the map with his phone and sent it to his own email address with the subject line: Working document. Do not delete.
Then he stood at the kitchen window with his second coffee and looked out at the street below. A woman walking a dog. A man on a phone call, pacing a small circle on the pavement. The ordinary infrastructure of a Tuesday morning in a city of a million people who had no idea they were living above or around, or inside something that Kael could not yet name but could feel every time he stood near certain stones.
The hum was there now. Not strong. It was almost always there to some degree in his apartment, faintly, which he had decided was because the building's foundations dated from the early seventeenth century, and whatever the thing was in the stonework, it was present here too, though dimly.
He counted the dots on the map again. Nineteen. He had been in Prague for eight months. He had covered maybe a third of the historic center systematically, and the rest in transit and chance. Which meant there were almost certainly more locations he hadn't found yet.
He wrote the number nineteen in the margin of the map, circled it, and then wrote a number below it with a question mark.
The question mark number was forty.
He didn't know where forty came from. It wasn't a calculation. It was an instinct, the kind that came from years of reading buildings, the sense of scale that a thing had, the scope of the intention behind it. Whatever had been built here, it was large. Larger than nineteen points. Larger, maybe than he could find in a year of systematic walking.
He folded the map carefully, along its original creases, and put it in the drawer with his notebooks.
He had seven high-intensity locations confirmed. Three probable nodes. One partial translation of an inscription that someone had not wanted widely read. One retired detective was sitting on information he hadn't shared yet. One bookshop owner who had known his name before he gave it to her.
And a crack in a bridge stone that had been warm at seven forty-five on a November morning when it had no business being warm at all.
He wrote at the bottom of the page, below everything else, in letters smaller than his normal handwriting:
He counts seven. There are more. Some are already gone.
He didn't know where that came from either. He looked at it for a moment. Then he closed the notebook and went to put his boots on because he had a documentation site to get to, and the morning was not going to wait for him to figure out what he knew and didn't know. That was a job for the evenings. The mornings were for the work that paid for all the rest of it.
He picked up his camera bag and his keys and went out into Prague, and the hum went with him.
It always went with him now.
End of Chapter 6