March came in cold and stayed that way. The castle grounds were grey with lingering frost, the lake dark and still, and Rowan spent most of his free hours in the Room of Requirement with his journal open and copper plates spread across the workbench.
The luminaire was proving more difficult than he'd expected.
The runic array itself worked beautifully. Three tiers of inscribed runes on a copper disk the size of his palm, activated with a single touch, producing a warm steady glow that lasted anywhere from six hours to three days depending on how precisely he'd managed the spacing. He'd built a dozen prototypes now.
The best of them had been glowing continuously on his workbench for four days, bright enough to read by from across the room.
The problem was the copper.
He discovered it on the fifth day, when the first prototype's light began to flicker. The surface had tarnished where moisture from the air had settled overnight, and the oxidation had eaten into the fine lines of the cycling runes. One small point of corrosion in the wrong place and the perpetual flow pattern broke. The light stuttered, dimmed, and died.
He scrubbed the disk clean, re-inscribed the damaged rune, and reactivated the array. It worked again. But two days later, the same section tarnished and the same rune failed.
Copper wouldn't hold. He needed a substrate that resisted degradation, something that would keep the rune lines sharp for years rather than days. Silver was the obvious answer. It held enchantments well, barely tarnished, and its Lunar association with illumination made it a natural fit for a lighting device.
But silver cost eleven Sickles an ounce, and each disk required roughly four ounces. At that price, materials alone pushed the unit cost above two Galleons.
The whole point was to build something everyone could use.
He brought the problem to Iris on a Tuesday evening in the common room, sitting across from her at their usual table near the fire.
"I want to know if you think this is actually worth building," he said, and set one of the functioning copper prototypes on the table between them. The amber glow lit the underside of Iris's chin as she leaned forward to examine it.
"What is it?"
"Permanent magical lighting. A runic array that produces steady light from a single activation, no wand needed, no maintenance. I'm calling it a luminaire."
Iris studied it, turning the disk in her hands. "It's warm. Brighter than I expected."
"That's the low setting. Ambient light, for replacing candles and torches. But the same array at a steeper inscription angle produces far more than this." He paused. "Bright enough to make a room look like midday. At night."
Iris looked up from the disk. "Daylight."
"Every home, every shop, every corridor in this castle goes dark when the sun sets. We light candles and torches and cast Lumos and accept that nighttime means dim. This changes that entirely. Actual daylight, whenever you want it, in any room."
Iris was quiet for a moment, turning the disk over again. "You're asking me whether witches and wizards would spend money on this."
"I'm asking whether you think the wizarding world is ready for it. Whether they'd see it as useful or just strange."
She set the disk on the table and sat back, considering. "Useful is what sells. And this is useful. Think about a witch brewing potions at midnight, squinting by candlelight. A shopkeeper whose merchandise looks dull the moment the sun goes down. Parents worrying about their children reading by torchlight." She ran her finger along the edge of the disk. "But trusting it is another matter. It's new, it's made by a Muggleborn, and half the wizarding world treats innovation like a personal insult."
"Which is why the product has to be flawless before it reaches a customer. If even one luminaire fails in someone's home, the whole enterprise fails with it."
"Then make sure they don't fail." Iris pushed the disk back toward him. "Yes, I think it will sell. Make it reliable and affordable and the rest will follow."
Reliable and affordable. The two things that mattered most, and both pointed back to the same pair of problems. The substrate had to hold up indefinitely, which meant silver or an alloy with similar properties. And the production had to be fast enough to bring the price down, which meant he couldn't inscribe every disk by hand.
He couldn't solve either problem alone. Iris's practical instincts were invaluable, but the technical work, the runic geometry and inscription engineering and the transmutation process, needed someone who could sit beside him at the bench. Lawrence was the only real choice. They'd been discussing artificing since first year, and Lawrence's mind worked in exactly the complementary direction Rowan needed, mechanical where Rowan was theoretical, hands-on where Rowan was abstract.
But Lawrence didn't know runes. He'd never studied the Flamels' framework, the approach that treated runic geometry as engineering rather than mysticism. Without that foundation, he couldn't contribute to the array design or the inscription process.
Rowan pulled the Flamels' unpublished manuscript from his shelf that evening and brought it to the dormitory.
"I need you to read this," he said, setting the book on Lawrence's desk. Lawrence picked it up and examined the cover. "Foundations of Runic Theory and Practice. Nicholas and Perenelle Flamel." He looked up. "This isn't published."
"They wrote it forty years ago. They gave me access to it last summer. It's the foundation for everything I've been building in the Room of Requirement, and I need a second person who understands it."
Lawrence opened to the first chapter and started reading. After a few pages, he went still. Rowan recognized the look. He'd worn it himself the first time he'd read Nicholas's explanation of how a rune's geometric structure created its energy flow pattern rather than carrying some arbitrarily assigned magical meaning.
"This changes everything about runes," Lawrence said quietly. "They're engineering. Actual engineering, with geometric principles governing energy flow."
"Read the chapters on sympathetic geometries and runic arrays. Those are the ones that matter most for what I'm working on. And the section on inscription methods. Take your time with it. I need you to actually understand it."
Lawrence spent the next two weeks with the book. He read during meals, during free periods, in bed after lights out with a Lumos charm tucked under his blankets. He asked questions constantly, and they were good questions, the kind that showed he was engaging with the material rather than just absorbing it. Why did Isa's geometry create stillness rather than stagnation? How did Ingwaz's containment function differ from Thurisaz's protective barrier when both seemed to restrict energy flow? What happened when you combined runes whose flow patterns were neither compatible nor incompatible but orthogonal?
The last question was one Rowan didn't have an answer for, which meant Lawrence was already thinking beyond the text.
By the end of the second week, Rowan brought him to the Room of Requirement.
Lawrence stopped in the doorway. The workbench, the copper prototypes in various states of completion, the journal pages pinned to the walls.
"How long have you been working on this?" he asked, walking to the bench and picking up one of the luminaires. The amber glow lit his face as he turned the disk over, examining the rune inscriptions.
"Since February. Everything you've been reading is the theory behind what's on this workbench."
Lawrence studied the array for a long time, tracing the geometric relationships between runes with his finger. "Kenaz core function, Sowilo for intensity, Jera cycling between them. Isa regulation on the outer ring, Eihwaz binding at the cardinal points." He looked up. "The spacing between the Jera and Ingwaz runes is tighter than the book recommends."
"I compressed it to fit the disk. It still cycles cleanly as long as the lines stay sharp."
"But copper doesn't stay sharp."
"No. Which is the problem."
Rowan walked him through the rest. The substrate issue, the silver cost, the transmutation path he was exploring: lead to silver through the planetary hierarchy the Flamels had taught him, Saturn to Moon, fewer stages than lead to gold, potentially achievable in days rather than weeks. And then the production bottleneck, forty minutes per disk by hand, the need for some kind of mechanical inscription process.
"A device that inscribes runes automatically," Lawrence said, and there it was, the look Rowan had seen a hundred times during their discussions about layered enchantments and self-stirring cauldrons. A mind catching on a problem it wouldn't release. "Consistent depth and consistent spacing."
"I keep thinking about a printing press. A stone base, a mechanical arm following a predetermined path across the disk, carving each rune at the correct depth and angle. But gears and levers can't achieve the precision I need. Fractions of a millimetre matter."
"So the movement itself has to be governed by runes." Lawrence pulled Rowan's journal toward him and reached for a quill. "Raido for controlled, repeatable motion. Its bidirectional flow pattern could guide the arm along the same path every time." He was already sketching, his lines quick and certain. "And if the carving tip had a Sowilo rune built into it, it could inject magical energy into each rune as it carves. Activate the array during manufacturing instead of requiring a separate step."
"You've been reading."
"You told me to actually understand it." Lawrence tapped the sketch. "But the arm needs stabilization. Any vibration during inscription degrades precision. You'd need Isa in the base structure, dampening resonance."
"And a power source. I was thinking a Jera-Ingwaz stored charge in the base. Same cycling-containment principle as the luminaire itself. Charge it in the morning, let it run."
They spent three hours redesigning Rowan's original concept. Lawrence identified a problem with the carving arm's range of motion that Rowan had missed entirely, a geometric constraint that prevented the arm from reaching the outer runes without repositioning the disk. His solution was a pivot mechanism using steel, the pivot point governed by a secondary Raido rune that allowed smooth angular adjustment.
By the time they finished, the journal page was covered in overlapping sketches and annotations in both their handwriting.
"This is what artificing is supposed to be," Lawrence said, leaning back. "Everything we've been talking about since first year."
"I want to build it this summer. The press, the transmutation setup, the luminaire production. All of it. In Diagon Alley."
Lawrence looked at him. "You're opening a shop."
"Yes. I'll need Weasley to sign off as my guardian for the registration."
Lawrence was quiet for a moment, already rearranging his summer in his head. "You're going to need help. Someone who knows the runic systems, who can work the press and troubleshoot the arrays when they break."
"I know. That's why I spent two weeks making sure you could read the book."
Lawrence understood the implication immediately. "You want me to come to Diagon Alley over summer?"
"If you can. You'd be away from home for weeks, working long hours on a venture that might not even turn a profit."
"I'll need to ask my mother." But he was already smiling. "She'll worry. She always worries. But she also knows I've been talking about artificing since I got to Hogwarts, and this is the first time someone's actually offered me the chance to do it."
"Tell her I'll pay you."
"Tell her yourself when she inevitably writes you a twelve-page letter full of questions about safety and living arrangements and whether we'll be eating properly."