"He who hath slept in the cold knoweth the warmth of morning better than one who hath not.
And he who hath wept in the dark knoweth the length of night.
Neither knowledge is wasted, though neither is sought."
No attributed source
I woke up with my face in the pillow and my left hand clenched around the edge of the mattress hard enough that my fingers ached when I uncurled them. The dormitory was dark. Thomas was breathing. Somewhere in the castle, deep enough in the stone that it was almost nothing, water moved through a pipe or a channel. My own breathing was the loudest thing in the room. I lay on my back and looked at the dark ceiling and did not move for a long time, tears silently streaming down my cheeks.
The dream had been very clear. Most bad dreams dissolve the moment you open your eyes, leaving only an impression, a residue of fear without specific content. This one did not disappear but faded to a dull recollection but I could still recall the general details and the headless infant.
After a while I got up. The washroom was cold and the water from the basin was colder. I dressed in the dark and brought my notebook down to the common room, which was empty at that hour, the fire banked to coals that gave enough light to see by if you sat close. I did not try to add wood. I sat cross-legged on the hearthrug with my notebook open across my knees and stayed there until the coals went pale and the first gray light appeared at the narrow window above the bookshelves.
By the time Thomas came down the stairs and found me there, the sky outside had gone properly light.
"Thou art up early," he said.
"Couldn't sleep."
He looked at me a moment longer, then went to see about the fire without asking anything further. Thomas had an instinct for when conversation was not wanted, which was one of his better qualities and not one I had fully appreciated until now. Though I doubt I was being subtle about my current mood or willingness to talk.
The rune work that week was focused on solving the problem I had named last month: how to use the muscle circuit safely over time.
The circuit itself I had already built and tested. It was a small series of carved runes on a strip of cured leather, intended to press against the forearm. At its simplest, it worked like this: the first rune, Fehu, drew in ambient magical energy and sent it moving through the carved channels. The second rune, Mannaz, which governs anything to do with the human body, recognized that the circuit was in contact with living tissue. The third rune, Tiwaz, a directed force rune, took that energy and shaped it into a specific push. The last rune, Thurisaz, released that push outward into the muscle itself, producing a controlled contraction. Removed thumb, contraction stopped. Clean and repeatable.
The problem was simple enough. If someone fell asleep wearing the thing, nothing would stop it from running continuously. A muscle held in contraction without rest does not recover. Long enough and the damage becomes real. I had no desire to wake with a dead arm, or worse.
The fix required two additional runes working in sequence. The first was Mora, a delay element, which I thought of as a pause built into the circuit. After each activation it would hold the pathway closed for a fixed interval before allowing another. The second was Isa, the static or dormant state rune, which I placed at the point where Mora worked. Isa's channel I carved narrower than the rest, which slowed the flow of magic moving through that section to a trickle. The practical effect was that Mora's delay became consistent regardless of how much ambient magic happened to be present in the room at any given moment. Without that narrowing, the delay would run fast in a magic-rich environment and slow elsewhere, which made it unpredictable and therefore useless as a safety measure.
The tested result was roughly ninety seconds between activations. Not a serious rest period by any exercise standard I had known in a previous life, but enforced, and consistent, and sufficient to prevent continuous contraction. I re-carved the leather patch with the revised layout and tested it twice. The delay held each time. I wrote ninety-second delay confirmed in my notebook and left it there.
The larger ambition, what I had taken to calling the Perpetual Magical Workout in my private notes, was harder and had run into a wall of a different kind.
The reasoning behind it went back to what Professor Ashford had explained about magical capacity: it was not a fuel that ran out, but a capacity that could be trained like muscle, by pushing it to its limit and allowing recovery. If that was true, the most efficient training would be to push the limit as many times as possible in a controlled way. My idea was to build an external storage array out of carved runes, drain my own magic into it through deliberate expenditure, then draw the stored magic back into myself once I had recovered. Repeat. Stress and recovery on a tight cycle, the same principle as any physical conditioning program.
The outward part was straightforward. Runes already moved magic from a caster into an array. That was what they were for. The inward part was where the vocabulary ran out. I read through the runic text twice looking for any rune that described an intake pathway, a mechanism for pulling stored magic back into a person rather than pushing it further out. There was nothing. Jera, the cycle rune, accumulated energy within an array over time, but that accumulation was internal to the array. The loop stayed inside the circuit. It did not bridge back to the caster.
The rune I needed either did not exist in the sources I had access to or had not been developed yet. I wrote: the problem is vocabulary, not logic, and left it on the desk. The caravan had brought a new text on ward construction that I had not finished yet. The answer might be in there. If it was not, working out the symbol from first principles was a different kind of problem entirely, one I was not sure I was equipped to approach.
What I did have, in the meantime, was the MAW. One arm at a time for ninety-second intervals. It was not the elegant solution I wanted but it was a working one, and in the first two weeks of March I was not too proud to use it.
The physical effects were noticeable within a week. The morning work assignments, which had been hard going in January, stopped being hard in the same way. I hauled a full barrow of debris from the east courtyard without stopping and still had enough left to stack the timber alongside. Hugh, the second-year Hufflepuff I had worked beside in January, remarked on it without particular ceremony: "Thou art stronger than thou wert." He sounded mildly impressed, which from Hugh was approximately equivalent to a standing ovation from anyone else.
It was true. I was. The rune circuit was doing something. How much of the gain was the circuit and how much was simply three months of physical work and proper feeding was impossible to say, but the circuit was contributing. I could feel the difference in the arm I had been running it on compared to the other.
What I had not fully accounted for was the cost.
Magic was exertion, not fuel, which I had understood in theory since Ashford's January lecture. But understanding a principle and running yourself up against it are different things. By the end of the second week I was finishing each evening with less to give than I had started it with, in a way that went beyond normal tiredness. Ashford's lecture, the warming charm exercise, the concealment work in the library, the rune sessions in the classroom, three activations of the sensing circuit I had built in February to test its behavior with the new text's observations in mind; all of it was expenditure, and the cumulative weight of it did not disappear overnight the way physical tiredness from labor usually did.
I started the third week slower than I had started the second. Not dramatically. Not enough that Thomas noticed, or if he noticed said anything. I was completing the assigned work and attending every class and keeping up with the reading. But by the evening I was sitting at the desk in the classroom staring at a notebook page and finding that the thought I had started forming was simply not there anymore, which had not happened before.
The first Quidditch match of the March term was a Saturday.
I knew three days in advance because it was the only topic of conversation in the Hufflepuff common room from Wednesday onward. First years were not ordinarily included in the main spectator arrangement, which occupied a tiered wooden stand on the south side of the pitch used by second years and up, by unspoken convention if not formal rule. There was a smaller stand on the east side, barely more than three rows of benches on an elevated platform, and first years were directed there by the older students in the same way that younger siblings are directed to the children's table at a family gathering. Not cruelly. Simply as the established order of things.
It was cold. March in Scotland did not carry the warmth the name suggested, and sitting on an open platform for two hours without the option of moving around meant that by the end of the first half I could no longer reliably feel my toes and had stopped trying.
Thomas did not appear to notice the cold. He watched the match with the focused attention of someone following a complicated argument and determined not to miss the resolution. He knew the players on both teams, Gryffindor against Ravenclaw, by name and position and broom model, all of which he had obtained from the sixth years by the simple method of asking questions in a direct and unself-conscious way until people had told him everything they knew simply to be done with him.
"That is the Ravenclaw Keeper," he said, pointing. "He did stop fourteen of seventeen goal attempts in the autumn. He doth ride a Cleansweep. It doth turn faster than a Comet but doth not accelerate so quickly in the straight."
"How do you know that?" Eleanor asked.
"Roderick in fifth year did tell me."
"How did Roderick know?"
Thomas considered this. "I did not think to ask."
The pitch was larger than it looked from the castle windows. Seven players to a side. Three Chasers moving the Quaffle between them in passing patterns that were sometimes graceful and sometimes a collision. Two Beaters managing the Bludgers with what I was fairly certain involved actual violence. A Keeper at each end defending three ring-shaped goals set on tall poles. The two Seekers ranging high and separate from all of it, watching.
The Golden Snitch was small and fast. I had read about it in the context of the Snidget lesson, the enchanted ball that had replaced the living bird two and a half centuries prior, but watching it in actual flight was different from any description. It moved in ways the eye could not quite track, appearing at one end of the pitch and occupying another position before the intermediate movement had fully registered. The Gryffindor Seeker, a third year whose name Thomas had mentioned and I had not retained, tracked it with a patience I found genuinely interesting. She was not following its movement directly. She was watching the pattern of it, waiting for the pattern to repeat.
She caught it forty minutes in. The Gryffindor stand erupted at a volume that carried easily across the pitch to our smaller bench.
We walked back through mud that had thawed enough to be properly disagreeable. Thomas talked about the match for the entire return. Eleanor listened with the expression of someone who is listening because they have no reason not to. Margaret said nothing at all, which was her standard mode after anything requiring sustained attention.
My boot had developed a failing seam at the left side, which I discovered at supper when I removed it under the table to assess the damage and found the interior was damp. Eleanor noticed what I was doing and said nothing. Thomas noticed Eleanor not saying anything and asked what was wrong, and by the end of the meal everyone at our corner of the table knew about the boot.
"The cobbler in Hogsmeade might see to it ere the seam doth give entire," Margaret said.
"Yes, however I'd rather use the connecto spell. There's no telling when we might be able to visit Hogsmeade next."
The Reparo work started as a theoretical exercise on a Tuesday evening when I had finished the assigned reading and had nothing else scheduled.
Connecto, the binding thread charm from Ashford's class, accomplished something specific: it gathered the existing threads of a fibrous material and wove them back against each other at the break. The mechanism was physical. The charm moved existing material, created no new material, altered no properties. The limitation followed directly: fibrous materials only. You could not gather threads in ceramic. Stone had none. Wood's grain ran continuously and a crack in it was a fracture through the grain, not a break in threads, which was different.
Reparo, as I knew it from my previous life, was general. Cloth, ceramic, wood, glass, metal. The mechanism had to differ from Connecto in some fundamental way. Professor Crane's framework for Transfiguration was the closest model I had: pattern substitution, reimposing a structure that had been present before damage occurred. Where Connecto moved existing material, a general mending spell would need to restore a configuration. To show the object what it had been before the break, and ask it to be that again. There was even speculation it tapped into time magic but that remains to be seen. Though, I guess with the right movements and such, any effect could be achieved.
I tried it on a small clay disc with a crack across it, taken from the bin of damaged practice pieces in the Transfiguration room. The motion I chose started at the undamaged rim and drew inward toward the crack, the same directional logic as Connecto but applied to shape rather than thread. The incantation I settled on was Reparo, from the Latin for to restore. It seemed appropriate and had the advantage of being memorable.
The first five attempts produced nothing visible. On the sixth, running my thumbnail along the crack, the edges felt marginally smoother. I tested the same movement on an untreated disc and confirmed the change was real. Small. Insufficient to close the crack. But real.
I wrote: the surviving geometry principle may be sound. The crack edge softens but does not close. Try starting the motion simultaneously from both sides of the break, meeting at the center, as though closing a seam from two ends rather than pressing from one.
It was a poor result by most measurements and an adequate one by the only measurement that mattered: the principle appeared to be sound. A wrong principle would have produced nothing at all.
The library held a small collection of texts on historical spell practice, shelved in the restricted section. Mistress Forrest, the head librarian, regarded requests for it with the specific suspicion of someone who assumes bad intentions until shown otherwise.
I had asked twice in February and been told the texts were not for first-year study. In March I brought a note from Professor Ashford. Mistress Forrest read it, read it again, told me she would require twenty-four hours to verify the request, and the following day granted access for forty-five minutes at a time, twice per week, to three specific texts. She would be present.
She was present but occupied with cataloguing. I had the reading table to myself in any practical sense. The text I wanted described the Patronus charm, not as instruction but as an account of how it actually worked, aimed at someone who already knew the incantation and was trying to understand the mechanism.
The description was careful and unusual. The spell did not operate on standard magical will directed outward. It operated on the caster's emotional history. Specifically on the quality and accessibility of a sustained, genuine positive experience brought into active mental use at the moment of casting. The magic did not come from the emotion itself. The emotion provided the structural template through which the magic was shaped. A clear and genuine experience would produce a Patronus of corresponding strength. An experience that was vague, or performed, or only observed from the outside rather than inhabited from within, would produce nothing, or something that dispersed within seconds.
The text used conviction rather than happiness, which I noticed because it was specific. A happy memory was not sufficient if it was only remembered. The caster had to briefly inhabit the experience, be present inside it as though it were occurring, not observe it at a distance.
I considered what I had. My previous life held genuinely happy memories but they belonged to someone else's experience. My present life was shorter and had spent a substantial portion of its length being unpleasant by most objective standards. What I settled on was an evening earlier in the year, the common room, the fire, Thomas and Eleanor and Margaret there, nobody requiring anything from me except to show them something I had been working on, and Thomas saying it was rather impressive in the tone that from Thomas amounted to elaborate praise. A small moment. Ordinary in every measurable respect. I had been genuinely present in it.
I tried the incantation at the reading table with Mistress Forrest eight feet away, on the grounds that witnessing a failure was less consequential than not attempting at all.
She looked up from the catalogue. She did not speak.
What came from my wand was a silver-gray mist that dispersed within two seconds without forming any recognizable shape. The text had described this result precisely: insufficient conviction, or genuine experience held too briefly. I packed up and left with two minutes remaining in the session.
"I will be back Thursday," I said on my way out.
"Thy leave doth permit Tuesday and Thursday," Mistress Forrest confirmed, without looking up.
The next task I had set for myself spell-wise was the disillusionment charm. Oddly enough, this was probably the easiest charm. I kept in mind what invisibility was and wasn't. I didn't want to camouflage like a chameleon nor to bend light around me but to simply be completely see through. After having found a relevant wand motion in the library, I ended up nailing the charm down in just a couple of days. I could now move around without a glimmer or shimmer, which would have been the case for the described effects in the library book, which sounded like a mirage in a desert. With that down, I'm one step closer to the legendary stealth build. As everyone knows, the most overpowered build early-stage in any setting is stealth. If you can't see me, you definitely won't see the knife in your back.
And then things started to go downhill. By the end of the third week I was not recovering overnight the way I had been.
The physical gains held. The labor assignments that had been genuinely taxing in January were now manageable without much thought, and there were now bands for each limb, core, back, shoulders, and a lighter intensity one for the chest. I didn't want to mess with my heart, which was also a muscle that could have been forcefully contracted. I imagine it was quite a sight to see: an eleven year old twitching and flexing all over his body while reading. Gotta grind, you know?
The magical side was a different account.
What I had been doing across the past three weeks amounted to this: the PMW circuit running several sessions per evening, the Reparo attempts each requiring sustained conceptual effort, the Patronus sessions twice per week each of which left a specific kind of flatness afterward, the concealment work, the sensing circuit from February which I was still running occasional tests on, and the standard class requirements on top of all of it. Each of these was a draw. None of them were catastrophic individually. Accumulated over three weeks without adequate recovery, the total was becoming noticeable. The PMW was active at any time of the day unless I was practicing spellwork, even while sleeping. Unfortunately, I could not measure my capacity except for the time per cycle that increased over the weeks.
I fell asleep at my desk in the classroom one evening and woke up with my face against the notebook, the candle burned nearly to the socket. I had been working on a rune diagram, or had been trying to, and the pencil had rolled to the edge of the desk. I could not remember where the thought had been going.
At breakfast the following morning I was hungry in a way that the standard serving did not address, and I went back for a second portion, which Thomas observed with the particular attention he gave to anything outside normal behavior.
"Thou art eating more than common," he said.
"I know."
He looked at me a moment. "Art thou well?"
"Fine. Just tired."
He accepted this, apparently satisfied, and returned to his bread. Margaret, across the table, did not comment. She was watching the same thing Thomas had noticed and drawing her own conclusions, which she would share with me eventually and not before she had decided they were correct.
In Charms that afternoon I managed the assigned exercise with adequate results and then sat through the remaining quarter-hour doing nothing in particular, which I did not normally do. Professor Ashford glanced at me twice. He did not ask anything.
On Saturday I slept past the morning and woke to an empty dormitory and the sound of rain. My arms were sore. Both of them, which was odd, since the soreness in the circuit arm I had come to expect. The other arm being sore told me something I should have already concluded: the magical fatigue was expressing itself physically in ways that were not specific to the limb being trained. Everything was simply more tired. Not injured. Just thoroughly used.
I lay there for a while and then got up and ate and felt somewhat better and went to the classroom and looked at the rune diagram I had fallen asleep on and found that the thought was back, clear enough to work with. So I worked on it for two hours and then stopped when the thought began to go soft again, which was the signal I had been ignoring for three weeks and was now paying attention to.
I put the notebook down and did nothing for the rest of the afternoon.
The caravan arrived in the last week of March, when the ground had thawed enough to be deeply unpleasant underfoot without committing to full mud, and the snow had retreated to the north faces of the castle towers where it had packed against the stone during weeks when the sun had never quite reached that angle.
They came up the road from Hogsmeade in the early morning. Four large wagons with canvas covers and two smaller carts, the whole column moving slowly enough that I had time to watch from the third-floor classroom window while meant to be reviewing Charms notes. Heavy-breed horses built for load. Roughly six adults traveling with the wagons, three of whom carried wands openly at their belts.
Owen Thatcher met them at the outer gate. The conversation lasted several minutes and involved gestures toward the inner courtyard. They came in.
A sixth-year student had explained the arrangement to me earlier in the year: the school kept agreements with several traveling merchants who came annually or twice yearly. The merchants brought what the castle could not produce or grow locally, preserved ingredients, texts outside the library's standard holdings, materials for the workrooms, and occasionally living creatures for the Magical Creatures lessons. The castle provided food, shelter for wagons and animals, gold from the school accounts, and sometimes trade goods: firewood, stored provisions, items produced by older students in craft-skills instruction. It was how an institution in a remote location fed itself through the parts of the year when roads were difficult and supply from the south was unreliable.
Students were permitted to examine the merchant wagons during the midday hours on the two days the caravan was present. I went, along with most of the school population with free time.
The creature wagon held ventilated crates along its interior walls. Professor Wren was already there, in conversation with the merchant who managed the animal cargo. In the visible crates were three creatures resembling hedgehogs with longer legs and independently moving eyes, one creature resembling a small elk with blue-gray antlers and clearly juvenile, and something in a covered crate that made a sound at irregular intervals I did not attempt to identify. In the corner, in a crate with reinforced iron fittings, what appeared to be a small dragon. Dull scales in the low light, roughly two feet from snout to tail, and dormant.
"What species?" I asked the merchant woman.
She looked where I was pointing. "A Hebridean Black, young yet. Full grown doth reach eight to thirty feet. This one hath five years ere it reacheth any size of concern." She studied me. "Thou art a first year."
"Yes."
"Then thou art not permitted to purchase a dragon."
"I was not trying to." Totally was, not that I would have enough money. But wasn't that eye opening: you could buy dragons to raise as pets. Something else to add to the list.
"Good," she said, and returned to Wren.
The book wagon was more interesting. Three shelves of volumes, some in ordinary condition, some worn, some wrapped in waxed cloth against the damp, and a merchant who allowed students to handle them carefully but watched each one with the attention of someone who has learned from experience what careless students do to books. I found two texts I wanted: one on charm theory with a section on optical effects not in the library's holdings, and a thin Latin volume that appeared to be a practical text on runic combinations specific to ward construction.
I had two shillings and a worn halfpenny. The optical text was four shillings. The runic text was seven.
The rains came back in the last days of March and settled in for the kind of stay that needed no announcing.
It was not the sharp, intermittent rain of early spring but the low, gray, persistent kind that Scotland seemed to produce by default when the mood was upon it, arriving sometime in the night and still present by the following supper without once having raised its voice. The castle absorbed it the way stone absorbs most things: steadily, without protest, growing a degree or two cooler in the passages away from the fires.
On the Wednesday of that week Potions was cancelled, Professor Thorne having a meeting that ran long, and the afternoon opened without assigned work. I ended up in the common room at midafternoon with the new Latin text and nothing pressing, because the rain offered no argument for going anywhere.
Margaret was at the table near the window with her Charms notes spread before her. She had arranged the papers with the same precision she applied to everything, the draft she was working from on the left, the fresh parchment on the right, her inkpot positioned precisely at the upper corner where it would not be knocked by her writing hand. She looked up when I came in, registered that I was carrying the Latin text, and looked back down at her notes with the expression that meant she had filed the information and deemed it unremarkable.
The chair situation in the common room was adequate but not particularly comfortable for extended reading. There were two armchairs near the fire, both occupied by second years who had been there since morning and showed no intention of moving, a bench along the far wall that was long enough to lie on but too hard for any duration, and the table chairs that Margaret and two other students were already using. I stood in the middle of the room for a moment looking at the options.
I had been practicing object-to-object Transfiguration for two months. A chair was, structurally, not substantially different from the bench: four legs, a seating surface, a back, all of it wood. The bench was long enough and solid enough that there was material to work with.
I cut off approximately one-third of the bench's length, in the Transfiguration sense of designating a portion and leaving the rest unchanged, which required about four minutes of concentration and produced a rough block of wood that bore very little resemblance to a chair. Then I worked the block. The legs came first, which was the part I was most practiced at from the lead-to-copper disc exercises. The seat surface followed, thicker at the back than the front. The back was the hardest, because I wanted it to curve, and curve in Transfiguration required more precise mental geometry than flat surfaces. It took three attempts and the result was not quite the curve I had intended, more of a soft tilt than a true arc.
The runners I added last. A rocking chair required two curved pieces underneath the legs, and the principle was the same as the back: a defined curve held in the mind and imposed on the material. These came out better than the back, possibly because a runner's curve is shallower and more forgiving than a seat back's, or possibly because by that point I had the shape more firmly in my head.
The finished object was recognizable as a rocking chair. It was not beautiful. The proportions were slightly off, the back a degree or two past vertical, the seat perhaps an inch lower than ideal. But it rocked correctly when I sat in it, smooth and even, and the wood was solid, which mattered more than the proportions.
Margaret had been watching the last third of this process from the corner of her eye while continuing to write. She said nothing until I had settled into the chair and opened the Latin text across my knees.
"Thou hast made a chair," she said.
"Out of the bench."
She looked at the bench, which was now shorter by about a third. She looked at the chair. "The bench doth look well enough."
"It does."
She returned to her notes. The fire was steady. Outside, the rain moved across the windows in slow gray sheets, and the sound of it against the glass was the steadiest thing in the room. I opened the Latin text to the third chapter and began reading, the chair rocking in a small slow arc, and the afternoon settled around us without requiring anything further from either of us.