The report arrived in three pieces.
The first came from the physical analysis team: what they had recovered from beneath Vhal-Dorim was divine matter, roughly seventy percent intact. The skull showed marks consistent with deliberate extraction — clean, purposeful, not the kind of damage sustained in combat or in the process of dying. The team had documented this with the precision of people whose instruments could identify the fact and whose training could not fully account for what the fact meant. They had not reached beyond what their tools confirmed. He had appreciated that.
The second piece was a passage from a restricted Circle archive, a document he had pulled three days ago and not finished processing. He had stopped on one line near the middle of the third page: the reasons attached to the Solar God's actions in the surviving accounts do not fully align with the original sources. He had read past it. He had come back to it.
In forty years, do not fully align had meant one of two things. Either the sources were unreliable — the earlier records incomplete, or corrupted, or simply wrong. Or the sources were solid and what had been written over them was not.
The third piece was Aldric's operational summary: the remains had been maintained by the Children of Medusa through a ritual connection, renewed at fixed intervals, for two hundred years without detection. The summary used the word continuous. He had read it twice.
The first reading gave him the shape of what they were dealing with. The second reading gave him the question that shape was pointing at: if the connection had held for two hundred years without being found, the failure to find it was not a failure of the Church's intelligence apparatus. Something had been permitting it to remain hidden. Those were different problems.
He set the third piece down. He picked up the second.
The line sat where he had left it. He read it again. Then he placed the two documents side by side and looked at what they had in common.
Aldric sat across from him with the quality of someone whose stillness was not the absence of motion but its product — the body having been released from movement because the mind no longer needed it to think. Armandel had watched enough people over four decades to know the difference between a man at rest and a man who had already finished the calculation and was waiting for the conversation to reach the same conclusion.
He had been applying a version of that observation to Aldric for several weeks now. Not suspicion, precisely. Assessment — the kind he sustained toward anything he could not fully place. Whether what occupied that chair was what it presented itself as was a question he had not answered, and had not stopped asking. He had not stopped using what sat there in the meantime. That was not contradiction. That was the work.
"The connection," Armandel said. "You described it as sustained."
"The documentation described it as sustained," Aldric said. "The maintenance intervals were fixed. The records specified what a missed interval would produce — not a catastrophic failure, but a slow degradation. Something that could be caught early and corrected, if the right person was watching."
"Caught and corrected," Armandel said.
"That was the language used."
He looked at the reports on the table. Not at Aldric.
"The three who carried out the maintenance."
"Aldrel, Orath, and Sueven, from the records we recovered. There may be others — the documentation described the mechanism, not a complete roster. Those three appear most consistently."
He marked this without writing it down. "Sueven."
"Yes."
"He presented himself to the active investigation and provided information about the network."
"Information that was accurate," Aldric said. "About operations that had already ended. Nothing he said opened anything that was still running."
Armandel looked at him. "Your reading of why."
Two breaths. Aldric's hands did not move.
"A handler who presents to an active investigation with accurate information about a dead operation is either genuinely cooperating, performing cooperation in order to redirect attention elsewhere, or was directed to perform it by someone who assessed the presentation as the most useful move available given what the investigation was looking for." He stopped. "The information Sueven provided was accurate. It was accurate about operations with no current value. It gave the investigation something to confirm without giving it anything it hadn't already found."
"You lean toward the performance."
"I lean toward the performance. Whether that conclusion holds depends on details the current evidence doesn't establish. The alternative — that the timing and content were coincidental — requires accepting a degree of luck I find difficult to accommodate."
Armandel held this for a moment. Then he pulled a separate folder from the right side of the table.
"His file will be reviewed. Through the Circle's secondary channel, not the standard investigation path."
Aldric did not ask why. He understood why. Eleven years of positioning within the Church's institutional structure was eleven years of knowing which channels ran where. Armandel was not going to route a review of Sueven through infrastructure that Sueven might have touched.
He set the folder aside.
"The body," he said. "The physical analysis team worked from the skull."
"Yes."
"Their estimates of the overall condition were extrapolated from what they had in front of them."
"Yes."
He looked at the report. "You flagged a detection during the operation. A weight signature at distance — registered initially as consistent with a living presence."
"Registered that way before I had identified the source," Aldric said. "The signature was directional, not diffuse. The quality of weight a living thing produces, not the quality remains typically produce at that age and condition. Before I understood what I was reading, the initial response was inconsistent with what I later concluded."
"And you concluded it was the remainder of the body."
"Yes."
"In a different location from the skull."
"A significant distance."
Armandel looked at his hands on the table. He had been adding pieces for three days. This was the one that had not fit correctly until now, and it fit now only because of what it implied.
The skull was in Church custody. The marks on the skull, as the analysis team had documented them, pointed to a deliberate removal — something extracted from within while the subject was still alive, the extraction itself being the event that had produced death. The signature Aldric had registered at distance, the one he had initially identified as a living presence rather than remains, was consistent with what the rest of the body would produce if whatever had been removed from the skull continued to exert some form of sustained influence on it across the distance.
Something had been taken out of Medusa before she died. The taking had killed her. What had been taken was somewhere. What remained of her was somewhere else — not here, not in the vault beneath Vhal-Dorim, and carrying a signature that did not behave like remains were supposed to behave.
"The authorization request," Armandel said.
Aldric looked at him.
"For processing the remains. Submitted through the Pontiff two weeks ago." He was not asking. "No response has returned."
A pause.
"Two hundred years ago," Armandel said, "the Solar God authorized the action against Medusa. The action was carried out. It produced what is now in our possession. Two weeks ago, we submitted the formal request to authorize the next step." He looked at the blank space at the bottom of the page where the authorization should have appeared. "The response has not come back."
He let that sit in the room without attempting to fill it.
There were three explanations he could construct. A delay rather than a refusal — the process was ongoing, the response would arrive. A deliberate hold — the remains were more valuable in their current condition, being observed or used for purposes he did not yet have the information to name. Or the remains contained something that processing would destroy, something that an entity with two centuries of patience and planning was still accounting for.
He did not know which was correct. He might not be able to determine which was correct from the information he could currently obtain. That was an uncomfortable position to occupy, and he occupied it without adjusting his expression.
He turned back to Aldric.
"What I called a victory six weeks ago," Armandel said. He said it the way he said most things he found genuinely difficult — flatly, without cushion. "The connection to the remains was still active. The remains were in two locations, not one, and we had custody of one. The three principal handlers were not among those we disrupted." He looked at the table. "The network's internal infrastructure within the Church was intact when I called it done."
Aldric said nothing. There was nothing to say that would improve the accuracy of the assessment.
"I need a full review," Armandel said. "Everything we have on Aldrel and Orath. Last confirmed locations, any financial movement, communication traces — anything that has shifted in the past six weeks. I want it before the week ends."
"You'll have it before the week ends," Aldric said.
Armandel nodded once.
After Aldric left, he stayed at the table.
The lamp burned without any draft to move it. The room had the quality it held whenever the building settled into quiet around it — borrowed space, functional, without the residue of anything that had happened there being retained in the walls. He had worked in rooms like this for most of his career. He had stopped needing them to be otherwise.
He looked at the three reports.
The connection to the body had been live when he called the operation concluded. The remains existed in two locations, and he had information about one. The three handlers who had maintained the ritual connection for two centuries were unaccounted for. The authorization to dissolve what the Church held had not returned in two weeks. And the reasons historically attached to the Solar God's original actions, two hundred years prior, did not fully align with what the sources from before that period contained.
Five facts. He did not try to force them into a single structure yet. He had learned, at a cost he did not revisit often, that the worst error available to someone in his position was concluding before the information had finished arriving. The shape he could see now was incomplete. The complete shape would be worse.
He pulled the next item from the table. He turned to it and kept working.