He did the accounting on a Thursday evening in May, sitting at his desk with the ledger sheet he kept in the inner pocket of his good jacket. He had maintained the ledger since Year 1 — a folded piece of heavyweight paper he replaced when it wore through, on which he tracked inflows, outflows, and running totals in the small-but-legible handwriting he used for documents he expected to read many times. The habit was from Hollowmere: his mother had kept a household ledger in a notch above the fireplace, and he had learned at about age six that the number in the bottom-right corner of the ledger sheet was the difference between a good month and a hard month.
He unfolded the sheet and looked at it. The running total at the bottom was twenty-one gold, four silver. He had updated it after the third Vesc commission cleared in February; he had not updated it since, because there had been nothing to update. The number had sat in his jacket pocket for three months as the number it was.
He took out a clean sheet of paper and wrote at the top: *Target: 50 gold Sigils.* He wrote below that: *Current: 21g 4s.* He wrote: *Shortfall: 28g 76s approximately.*
He looked at this for a moment. Twenty-eight gold and change. He thought: this is a solvable problem. He thought: I have been solving versions of this problem since I was six. He thought: I should be specific about how.
---
He drew a line across the paper and wrote: *Income sources.*
The first source was the Lir workshop. He wrote this down and put a zero next to it in the monetary column, because the arrangement was not monetary. He worked; he received instruction. The instruction was worth more than any reasonable wage he could command in Crooked Lane at his current standing — he thought about what Lir's sessions would cost if billed at a teaching rate, and arrived at an estimate that was larger than all his Vesc commissions combined. The workshop arrangement was not zeroed-out in any real sense. But it did not produce gold, which was what he needed, and he was not going to renegotiate it, because the instruction was worth more than the gold and he was not going to be foolish about that.
The second source was the Crooked Lane commissions. He wrote: *Vesc commissions.* He had done three so far — five gold, six gold, and the third which had cleared in February at a larger figure he had estimated at eight gold at the time (the actual payment had been seven gold, nine silver, which was close enough that his estimate had been reasonable). The fourth had been tentatively discussed. Vesc had mentioned, in the brief exchange at the commission handoff in February, that he had a larger project in development — sourcing components for a multi-unit filtration system for a customer whose scale Vesc had declined to specify. Larger than anything Kael had built for him before. No timeline confirmed.
He wrote: *Average per commission: ~6-7 gold. Rate: 1.5-2 per year.* He thought about the math. At two commissions per year at six to seven gold each, he was adding twelve to fourteen gold per year to his holdings. At that rate — and adjusting for the Doran business income on top — he was looking at approximately sixteen to eighteen months to close the gap to fifty gold.
The third source was the Doran business. He wrote: *Doran business: ~2s/week, growing.* Two silver per week was roughly ten silver per month, which was one gold per year with some adjustment for weeks when the business had a good run and the split ran higher. He had been averaging approximately one gold, two silver per semester from the business for the past two semesters. It was growing: Doran had been systematically expanding the supplier relationships and diversifying the product range, and the trajectory was upward, but not in a way that would dramatically accelerate his timeline. More like one to two gold per year if the growth continued as projected.
He thought about Doran's approach to the business with some appreciation. Doran ran it with what appeared to be effortless social facility but was actually careful operational thinking — he knew which supplier contacts to cultivate, which student customers were reliable, how to manage inventory through the parts of the year when demand shifted. Kael's contribution was the supply-side relationships and the Crooked Lane access; Doran's contribution was everything else. The business worked because both sides of it worked. He had told Doran this once, and Doran had said "yes, that's why I asked you specifically" without any particular modesty.
He looked at the numbers. Current rate: twelve to fourteen gold from Vesc, one to two gold from Doran business, no monetary contribution from Lir. Approximately thirteen to sixteen gold per year. Twenty-eight gold, seventy-six silver shortfall. He wrote: *Timeline at current rate: ~20 months. Range: 18-24 months.* He wrote below that, in smaller letters: *Assumes no major unexpected expenses. Assumes Vesc commissions continue at current rate. Both assumptions uncertain.*
He thought about the uncertainty specifically. The Vesc assumption was the most load-bearing: the commissions had been reliable for two years, but they were not formally contracted beyond the current scope. Vesc was a professional with specific standards, and Kael had met those standards consistently, and Vesc had mentioned ongoing work as a possibility. But "ongoing work as a possibility" was not the same as a confirmed fourth commission, and the difference between those two things was approximately the margin between twenty months and thirty. He had been treating the ongoing relationship as more certain than it strictly was. He made a note: *Confirm fourth commission status — do not plan on it until it is offered.*
The Doran business assumption was more stable — the business was running and growing and had established product lines and supplier relationships, and Doran was not going to stop running it as long as it was profitable, which it was. But the one-to-two gold per year figure was conservative even by cautious estimate: the recent trajectory was toward four silver per week, which if sustained was closer to two gold per semester, which was three to four gold per year. The conservative estimate served as planning floor. The actual outcome might be better. He made a note: *Doran business: plan on 1-2g/year; actual may be higher. Do not plan on higher until it arrives.*
He looked at the range. Twenty months was a long time, considering the other variable.
There was also the question of which ability. He had been sitting with this since the workshop session with Lir: the Salt required an Echo to be running at the moment of application. He needed to have a specific ability running, a choice made, before he could use the dose. He had nine Permanent Slots available. The first ability he sealed would be his for the rest of his life. He thought: that is not a choice I will make without knowing what I am choosing. He thought: I have time to think about it. He thought: I should not waste the time.
He thought about what was worth making permanent. The fabrication skill he had run the Echo of during Year 2 was gone — it had been his own skill doing the work since March, not an Echo. He could rebuild an Echo of fabrication capability if he found the right source and the right moment. But he was also conscious that fabrication was not the most strategically significant thing he could seal. He thought: I have nine slots. The first one should count. He thought: I should not rush the choice just because I am impatient with the money problem.
He set that aside. It was a problem for when he had the gold, and he did not have the gold yet.
---
He drew another line across the paper and wrote: *Conclave. Article 14.*
He had been tracking the Conclave schedule since Eilen had mentioned it in passing the previous autumn, in the context of a longer conversation about the regulatory history of Mirror Resonance practice. The language Eilen had used was cautious and academic — he had said "expected within two years" in the way that people said things they were not supposed to be speculating about but were speculating about anyway, because the information was available in the regulatory calendar and anyone who cared to look would find it. The Conclave met in structured sessions; Article 14 review had been deferred three times; there was a scheduled session within the next eighteen to twenty months that Eilen had understood to be the session where deferral was no longer likely.
He wrote: *Article 14 vote: ~18-20 months. Conclave session window.*
He did not know what the Conclave would decide. He was not certain it would decide against him — the reformist faction was active, and the Reformist letter in his jacket pocket was evidence that there were people within the political structure who were working toward a different outcome. He knew too little to predict.
But he thought about what the Salt ritual required. It required a witness. It required a formal record of the event — not filed publicly, but present, held by the witness. After Article 14 was voted on — however it went — the political environment around Mirror Resonance would shift. If the vote went the hard direction, a record of a formalized Mirror Resonance ritual, held by a Senior Workshop Master at the Academy, would become a different kind of liability than it currently was. Lir had offered to witness; Lir was not careless; Lir understood what he was offering. But the window in which the offer was relatively safe was narrower than the window Kael had assumed when he first heard "fifty gold."
He wrote: *Timeline problem: ~18-20 months to Conclave, ~20 months at current rate. No margin.*
He looked at this. He thought: I need to either accelerate the income or find a different path. He thought: there is one different path I have not taken.
---
He drew another line and wrote: *Loan.*
He sat with the word for a moment. He had grown up in a household that did not borrow money if there was any possible alternative. His father had been specific about this in a way that children in farming families were often made specific about things that mattered: a loan was a obligation, and an obligation had terms, and terms had failure states, and failure states in Hollowmere meant losing the equipment or the seed stock or, in the worst documented case he knew of from local history, the land itself. He did not have a theoretical resistance to borrowing. He had a practical understanding of what it cost.
He thought about the practical structure of a loan in this context. He needed approximately twenty-nine gold. He had one person in his immediate environment who had access to significant lending: Doran Drey, through House Drey's banking connections. House Drey was not the largest house in the Compact's financial structure, but Doran had mentioned, in at least three separate contexts over the past year, that the family maintained lines of credit and managed capital for several Compact-aligned trading interests. Doran was also his closest friend, and had been since the first semester. He did not think Doran would refuse to help.
He wrote: *Doran loan: possible. Risks.*
He wrote the risks in a column:
*1. What is the money for.* He could not tell Doran the full truth. He could not tell anyone the full truth except Lir, who already knew, and Mira, who he thought knew enough to draw the conclusion but had not been told directly. If he asked Doran for twenty-nine gold and Doran asked why — which Doran would, because Doran was not a person who lent money without asking why — Kael would have to answer in a way that was not a lie but was not the whole truth. He was capable of this in narrow contexts. He was not comfortable doing it to Doran.
*2. Debt obligation with terms.* Any lending from a House banking connection would come with terms. Interest, probably. A repayment schedule. The obligation would be real and tracked. He did not currently have a repayment timeline that would satisfy standard lending terms.
*3. If he can't repay.* He thought about this one carefully. He did not think he would be unable to repay it eventually — his income was growing, the Vesc work was building, the Doran business was real. But "eventually" was not a term that lending instruments used. Failure states.
He looked at the column of risks. He thought: none of these are disqualifying individually. He thought: they are all solvable with careful handling. He thought: I am not ready to handle them carefully enough yet.
There was also a fourth risk he had not written down, because it was harder to frame as a line item. Doran was his closest friend in a context where he had very few close friends, and the friendship was built on a specific kind of mutual usefulness — the practical partnership of the business on one side, and on the other, a genuine warmth that had developed through two years of daily proximity and specific exchange. He trusted Doran with a range of ordinary things. He did not want to test the edge of what the friendship could carry by putting it under financial obligation at the same time as he was carrying secrets he couldn't share. A business partnership and a close friendship and a loan agreement and a silence about what the loan was for — that was a lot of weight on a structure, and structures had failure points.
He did not write this down. He thought it, noted it, and filed it with the rest.
He thought: I need more information — specifically, whether I can get a fourth Vesc commission in time and how large it would be — before I decide whether the loan path is necessary.
He wrote: *Filed. Do not approach yet.*
He folded the sheet and put it in his jacket pocket beside the ledger. He sat for a moment. He thought: I have a plan. He thought: it is not a good plan in the sense of being guaranteed; it is a plan in the sense of being a sequence of actions with a directional outcome. He thought: that is as much as most plans are.
---
The message from Vesc arrived two days later, in the form of a folded note left with the Crooked Lane receiving window — the wood-paneled slot outside the door of the artisan's row where small-trade communications were dropped. Kael's name was on the outside in Vesc's compact script, which he had learned to recognize from two years of occasional correspondence.
He read it in the lane. *Commission opportunity, larger scope than previous. Components for a parallel-array filtration system, six units per array, multiple arrays. Prospective yield for your portion: 9-11 gold depending on final specifications. Interested? Session this week to discuss scope if so.* No greeting, no closing, just the information in Vesc's characteristic directness.
Kael folded it and put it in his pocket. He thought: nine to eleven gold. He thought: that closes the gap to approximately seventeen to nineteen gold remaining, depending on where the actual payment lands. He thought: that is closer than twenty-eight.
He went to the session two days later. Vesc was his usual self: economical with language, precise about what he needed, not interested in the social surface of a business conversation. He had the workshop ready when Kael arrived — specifications laid out across the bench in neat columns, the kind of organizational preparation that suggested he had been thinking about this commission for longer than the message indicated.
The scope was larger than anything Kael had built for him before. The parallel-array filtration system was designed for industrial-scale use: six units per array, the arrays running in parallel to process a volume of whatever was being filtered through them simultaneously rather than in sequence. The specifications called for resonance components that matched tolerances across units — they had to calibrate to each other, not just to a standard, which was a harder problem than building six identical pieces to the same spec. It meant that the sixth unit had to know what the first five were doing, in the sense of resonance alignment, and the only way to achieve that was to build them as a set from the start and tune them in the calibration phase as a group.
"Your previous work was unit-level accurate," Vesc said, looking at the specifications with Kael. "This requires array-level accuracy. Can you do that."
It was not quite a question. Kael looked at the specifications and thought about what array-level calibration involved — the process of building each component to spec and then running a collective tuning pass that adjusted the inter-unit resonance harmonics. He thought about the session work he had done with Lir — the dual-layer resonance problem, the node correction, the integrated feel that had been developing. He thought: I have not built at this scale before. He thought: I know the principles well enough to extend them.
"Yes," he said. "I'll need access to the full set during the calibration pass. I can't do it in units."
"I know," Vesc said. "The space is arranged for it."
They worked through the technical exchange for twenty minutes: Kael asking specific questions, Vesc answering with the directness of someone who respected competent questions and had no patience for anything else. The exchange was the most substantive conversation Kael had had with him in two years of commissions. He noted this — Vesc was expanding the scope because he had decided Kael was ready for it. That was its own kind of assessment.
He accepted the commission. The first delivery was due in six weeks. The payment would clear on delivery in segments — four gold at first delivery, the remainder on final delivery.
He walked back to the Academy with the commission specifications in his bag. He thought: nine gold minimum. He thought: at nine gold, his total holdings would be approximately thirty gold. He thought: twenty more to go. He thought: with another commission of similar scale — which was plausible, given that the customer relationship implied by this scope often meant ongoing work — he was within reach in perhaps twelve to fourteen months.
He thought: fourteen months was tighter than twenty. He thought: that is measurably better.
---
He had not been in the refectory much during the post-Trials period — the end of the Trials bracket had released the team from the structured meal timing they had been keeping during the bracket weeks, and he had been using the flexibility to eat earlier or later depending on his Crooked Lane schedule. On a Friday evening in late May, his schedule ran long in the workshop and he arrived at the refectory late enough that the main rush had cleared and the tables were at half-occupancy, which was an hour he normally avoided but did not specifically mind.
He collected his meal — pot-stew Friday, bread, a small dish of braised root vegetables he hadn't seen on the menu in weeks — and scanned the room. Tessa Marrow was at the far table with two of the other Trials team members — Prenna Aldas and Senn Lovar — and there was an open section across from them. He crossed the room and put his tray down and pulled out the bench.
Lyra was next to Tessa.
He had not registered her from the other side of the room — she was seated facing Tessa, turned slightly away from the entrance, and the lamp-fall in that section of the refectory was from the wall rather than overhead, which shifted the visibility angle. He sat down before he processed the full arrangement.
She looked up.
He said: "Sorry — I'll find another—"
Tessa said, already in the middle of something she was telling Prenna: "There's room, sit down."
He sat down.
The conversation at the table was about a third-year Solenne practitioner who had, apparently, gotten into an impressive quantity of administrative trouble during the Trials bracket for a reason that Tessa had only partial information about and was in the process of constructing from fragments. Prenna had the other half of the story. They were assembling it between them. Senn was contributing occasional skeptical corrections. It was the kind of conversation that needed no participation from Kael and received none, which left him free to eat and pay the appropriate partial attention.
Lyra was across from Tessa, which was the diagonal-across from Kael. She was eating and listening to the reconstruction of the administrative trouble story with the mild attentive expression she wore for things that she found moderately informative and not particularly important. She asked one clarifying question about the timeline of the incident — the question was precise and useful and moved Tessa's reconstruction forward by several minutes. She didn't look at Kael.
He didn't look at her. He ate. The food was better than average — the kitchen ran a good pot-stew on Fridays, which was one of the things the Academy refectory did well. The bread was fresh. He concentrated on the food with the portion of his attention that was available and ran the numbers in the portion that wasn't.
Twenty-one gold four silver, current. Fourth Vesc commission: nine gold at minimum. Running total after first delivery: twenty-five gold four silver; after full delivery: thirty gold four silver, approximately. Twenty more needed to reach fifty. Another commission after this one — which was plausible, given that Vesc had mentioned the parallel-array system implied a customer relationship that might extend to maintenance work — would put him in range within another twelve months. Fourteen months total, probably. Possibly twelve.
He watched Lyra ask Tessa a second question about the administrative incident. The question suggested she had derived the full picture of the incident from Tessa's partial information and was verifying the element she wasn't sure about. She had that quality in information contexts — she assembled toward conclusions before the information was complete, and the questions she asked were targeted at the remaining uncertainty rather than the parts she'd already resolved. He had seen it in the Trials planning work and in the analysis sessions after the match. He had been watching it for nearly two years and he still found it specific in the way particular things were specific — not generic competence, but this exact shape of thinking in this exact person.
Tessa was laughing about something. Prenna said something that extended the laugh. The table was ordinary in the way that refectory dinners were ordinary — noise, food, the social texture of a school in late spring when the academic pressure had released and people were in the unwound version of themselves.
He thought: thirty gold four silver. Twenty more to go. Twelve to fourteen months.
He thought: she is at the table and the table is fine and we had four training sessions and we'll have more and the lamp is on the shelf and she burned her father's letter and I'll see her at training tomorrow.
He thought: twelve months is not that long. He thought: it is long enough that I do not need to make any decisions tonight. He thought: that is a reasonable way to hold a thing.
He ate the last of the stew and looked at the refectory table in its ordinary late-Friday state — the fire banked against the cool of the evening, the overhead lamps on their standard setting, Tessa in the middle of her third incident about the Solenne third-year and Prenna now fully on her side of the story. He thought: the table has a quality that I did not expect to find here when I came to the Academy with four silver and no particular idea of what two years would build. He thought: I expected to be alone in this building in the specific way a person from Hollowmere is alone in an institution built for a different kind of person. He thought: I am not alone. He thought: the table is real, and the work is real, and the lamp is on the shelf, and the numbers are moving.
He finished his stew. He tore a piece of bread and ate it. Tessa had moved on to a different topic — something about a second-year student in Hall Aldemar who had, according to Tessa, managed to accidentally inscribe a resonance construct backwards and produce a result that was technically functional but did exactly the opposite of what was intended. Prenna was skeptical this was possible. Tessa was maintaining that she had seen the evidence herself.
Kael said: "It's possible. The inscription grammar has a sign-flip error pattern in the fourth standard form — there's a node sequence that looks symmetrical but isn't. If you approach it from the wrong direction it inverts the output polarity."
Everyone looked at him. He had not intended to be the one who settled this, but the technical point was specific enough that he could not leave the incorrect skepticism uncontested.
"You know this from experience?" Tessa said, with the tone of someone who suspected there was a story here.
"From one of Lir's assessments," he said. "It was a controlled error. He wanted to see if I'd catch it."
"Did you catch it?"
"After I'd already built half the construct," he said.
Tessa made a sound of vindicated satisfaction. Prenna conceded the point. Senn began asking a follow-up question about the fourth standard form. Kael answered it.
He did not look at Lyra. He could feel that she had not looked away from him while the exchange ran — not in the conspicuous way, but in the way he had learned to read her attention, which was quiet and precise and present in a way he would not have been able to articulate to anyone who hadn't learned to read the same person over two years. He knew what her attention felt like when it was directed at a problem and when it was directed at him and he knew the difference. He let it run.
He thought about what it meant that he could feel the difference — the specific texture of being attended to by Lyra, which was a different texture from Mira's attention or Doran's or Tessa's or Lir's. He had spent two years in the same building as her, accumulating a knowledge of her attentiveness the way he accumulated all knowledge of people he was in proximity to: not deliberately, but through repeated observation of a specific thing until the thing became readable. He thought: she has a quality of attention that is different when it is directed at someone she is tracking versus someone she is simply observing. He thought: I know what category I am in. He thought: I have known for a while. He thought: twelve months. The numbers and the other thing running in parallel, both real, both in their own timelines. He thought: both of them would resolve in their own sequence, and the sequence was already running, and running ahead of it was not the right approach. He thought: let the sequence run.
The Senn follow-up question took another four minutes to work through properly. Prenna had a correction on one of his points — a valid one; she knew the fourth standard form's sign-flip from a different angle than he did, and her correction refined the explanation. He acknowledged it. Tessa looked between them with the expression of someone who had expected to be the most technically informed person at this table and was discovering the terms of a more complicated situation.
"I don't know what any of that meant," she said, with good humor, "but I'm glad you both know it."
"It means Tessa was right," Kael said.
Tessa pointed at him. "See. That's what I like about you."
The table laughed. The conversation moved on.
He thought: thirty gold. Twenty more to go.
The table continued. Tessa had found a third incident involving the Solenne third-year and was reporting it with the storyteller's attention to pacing. He had not expected Tessa to be a storyteller — she was so precise in operational contexts that the social ease of her narration was a different register entirely, one that she apparently kept separate and deployed selectively. He had noticed this developing in the Thursday afternoon business reviews, where Doran's conversational warmth had created a context in which her own version of warmth became available. He thought: she is the same person in both registers; the business precision and the dinner-table storytelling are both accurate expressions of Tessa Marrow. He thought: it took Doran to make the second one visible, which was its own kind of data about what the right context produced. Prenna was fully invested now; even Senn had dropped the skeptical posture. The lamp above the table was one of the standard refectory types — warm, overhead, no particular quality. Ordinary light.
He thought: twelve months. He thought: the lamp is on the shelf cycling its amber and its dawn-light, and she burned her father's letter, and I said I'll see you at training and she said yes, and we are at the table and the table is real and the work is real and this is what the next twelve months looks like before the numbers reach fifty. He thought: this is not a bad thing to be looking at.
He thought: I'll see her at training.
---
*End of Chapter Eighteen*
Sign in to comment
No comments yet.