The cracked practice plate did nothing.
Arin spent the first evening inspecting it.
He compared its channels with the Cleanse model.
He sketched the central mark, measured the distance between the lines, and noted where the outer channel had been severed.
Then he tried to sense its spiritual structure.
Nothing happened.
The plate remained cold.
He placed it beneath the academy lamp.
Nothing.
He moved it near his meditation space.
Nothing.
He held it beside the Cleanse model.
The Cleanse model warmed faintly.
The cracked plate remained dead.
Tomas watched from his bed.
“Maybe it is just broken.” “That is possible.” “You spent your money on it.” “Yes.” “So it is broken.” “That is not proof.” “It seems like proof.” Arin ignored him.
He drew the plate on a piece of paper and marked the surviving channels.
The lines were too precise to be decoration.
Even damaged, the model retained a clear structure.
A spell model was not a picture.
It was a physical pattern capable of guiding mana.
If the plate no longer responded, one of three things had happened.
The damaged channel interrupted the circuit.
The model had been deliberately disabled.
Or the plate required a specific input Arin did not know how to provide.
He wrote the possibilities.
Then he added a fourth.
The plate was never functional.
That possibility was the most irritating because it could not be tested easily.
The next morning, Arin carried the plate to spell class.
He kept it wrapped beneath his robe.
Instructor Valeor was explaining the difference between a spell model and a rune.
“A spell model provides a prepared structure for spiritual force and mana,” he said.
“A rune is a symbol or array that may produce an effect when supplied with the correct conditions.
The two are related, but they are not identical.” Arin’s attention sharpened.
“A spell model usually requires a living caster,” Valeor continued.
“A rune may be activated by a person, a stored charge, a trigger condition, or another rune.” That explained the academy lamps.
Their runes operated continuously without a visible caster.
It also explained the stabilizing array beneath the meditation hall.
The academy had built a network of permanent rune systems.
Arin raised his hand.
Instructor Valeor looked at him.
“Yes?” “Can a spell model be converted into a rune?” The instructor’s eyes narrowed.
“Not at your level.” “Is it possible?” “In theory.” “What is the difference?” “A spell model expects a conscious spiritual force to guide the energy.
A rune contains more of its own instructions.” “So the rune is more independent.” “Sometimes.” “Then it is more complex.” “Usually.” Arin wrote that down mentally.
The difference between a spell model and a rune was not simply shape.
It was where the instructions lived.
A spell model left part of the execution to the caster.
A rune encoded more of the process into the structure itself.
That made runes less flexible but more repeatable.
Framework Zero gained another note.
Spell model: caster-dependent.
Rune: structure-dependent.
The distinction might become useful later.
At the end of class, Arin approached Instructor Valeor.
“I found a damaged practice plate at the market.” “Many people do.” “It appears to contain a partial spell structure.” “Then it is a damaged practice plate.” “May I ask what it was designed to do?” “No.” “Why?” “Because you are not authorized to study unknown models.” “I am not asking to activate it.” “You are asking questions that lead toward activation.” “That depends on the answer.” Valeor held out his hand.
“Give it to me.” Arin hesitated.
The instructor’s gaze hardened.
“Now.” Arin removed the wrapped plate and placed it in Valeor’s hand.
Valeor examined the damage.
His expression changed for less than a second.
Arin noticed.
“You recognize it.” “I recognize poor workmanship.” “Is the channel deliberately severed?” “Perhaps.” “Why would someone do that?” “To prevent use.” “Then it was once functional.” Valeor returned the plate.
“That does not follow.” “If it had never worked, there would be no reason to disable it.” “People damage useless things all the time.” “That is true.” The instructor stared at him.
“Do not become attached to broken objects.” “I am not attached.” “You spent a copper on it.” “That is not attachment.
It is investment.” “Your investment is dead.” Arin looked down at the plate.
The phrase stayed with him.
Dead plate.
It was simple and accurate.
The model contained no active response.
Its channels were empty.
Its outer structure had been deliberately cut.
He returned to the dormitory and placed the plate on the desk.
Then he began comparing it with the academy’s Cleanse model.
The central mark was a basic input point.
The first two channels guided energy outward.
The third channel did not release energy.
It redirected it back toward the outer boundary.
That design was not meant to produce an external effect.
It was meant to maintain a stable loop.
A practice model.
Perhaps the plate had been designed to let apprentices feel energy circulating through a model without casting a real spell.
If so, the missing channel explained why it was inert.
Arin examined the break again.
The damage was narrow, but not clean enough to be a deliberate safety lock.
The metal around it was discolored.
Heat damage.
Someone had overloaded the model.
The channel had melted rather than been cut.
Arin wrote: Likely failure cause: excessive energy.
Not intentional disabling.
The plate had died during use.
That made it more valuable.
A failed process preserved evidence of what had gone wrong.
Arin placed the plate beside his notes.
He could not repair the spell structure.
But he could reconstruct the missing channel on paper.
Using the existing lines as a guide, he estimated the direction and width of the damaged path.
The first reconstruction was too thick.
The second curved too sharply.
The third followed the geometry of the remaining channels.
He drew a small mark where the energy would likely split.
Then he stopped.
A reconstruction was not a repair.
If the original channel had been damaged by overload, copying it exactly might repeat the same failure.
He needed to understand the intended function before attempting anything.
The next afternoon, Arin visited Joren’s repair stall.
Joren was repairing a metal clasp with a fine tool.
“You returned the wire.” Arin placed it on the table.
“I need another favor.” “That usually means you have no money.” “I have notes.” Joren looked up.
“What kind?” “Market prices for common herbs and writing supplies.” Joren considered.
“You have been recording prices?” “Yes.” “Why?” “Prices vary by stall.” Joren leaned back.
“How much information?” “Enough to tell you which vendors overcharge.” “Now you have my attention.” Arin handed him the page.
Joren read it.
“You marked the vendors who sell damaged tools without disclosure.” “They use different methods.
Some disclose damage.
Some call it ‘old.'” “That sounds familiar.” “Would you accept a repair consultation in exchange for information?” Joren smiled.
“You negotiate like an old merchant.” “I am trying to survive like a poor apprentice.” “What do you need?” “A magnifying lens.” “For how long?” “An hour.” “Five copper.” “I have one quarter.” “Then you have one quarter of a lens.” Arin pointed to a box of damaged metal plates.
“You have three items here with similar channel damage.” Joren glanced at them.
“Can you repair them?” “No.” “Then what can you do?” “Compare them.
Identify which damage is structural and which is surface-level.” Joren was silent.
“An hour,” he said.
“No leaving the stall.” Arin worked.
The three plates were old, but their damage differed.
One had a shallow scratch.
Another had a broken channel.
The third had a warped outer ring caused by heat.
Arin separated them and explained the difference.
Joren listened carefully.
When the hour ended, he handed Arin a small lens.
“Return it tomorrow.” Arin took the lens back to the dormitory.
Under magnification, the practice plate revealed more information.
The metal around the damaged channel contained tiny bubbles.
Heat damage.
The model had overheated from within.
Arin traced the surviving route.
The energy had entered the central mark, divided into two channels, then attempted to pass through the damaged section.
The obstruction had caused pressure to build.
The model had not failed because the spell was poorly designed.
It had failed because the caster had supplied more mana than the structure could process.
The same pattern appeared in Cleanse.
Excess input.
Insufficient control.
System failure.
Arin sat back.
The broken plate had taught him something important.
A spell model had limits independent of the caster’s spiritual force.
Even a talented Magus could damage a model by exceeding its designed capacity.
Power did not remove constraints.
It moved them.
That evening, he took the practice plate to Instructor Valeor again.
The instructor examined his sketches.
“You mapped the failure.” “I think so.” “You used a magnifying lens.” “Yes.” “Where did you get it?” “I borrowed it.” “From whom?” “Joren.” Valeor sighed.
“You are spending more effort studying a dead model than learning the living one.” “The dead model shows where it failed.” “Dead things do not teach.” “This one does.” The instructor looked at him for a long moment.
Then he pointed to Arin’s reconstruction.
“Your estimated channel is too narrow.” Arin stared at the drawing.
“Why?” “Because the original model used a wider distribution line.” “How can you tell?” “The curve near the second junction.” Arin examined it.
The remaining metal did support Valeor’s interpretation.
“Then the failure began at the junction?” “Likely.” “Could the model be repaired?” “Not safely.” “Because the damaged metal has changed.” “And because you do not understand the full structure.” Arin nodded.
That was a fair limitation.
Instructor Valeor handed back the plate.
“You are permitted to study dead models in the public hall.
Do not activate them.
Do not repair them.
Do not use them in class.” “Understood.” “And stop asking questions about converting spell models into runes.” “Why?” “Because you are a Grade One.” “That does not answer the question.” “It is the answer you are receiving.” Arin returned to the dormitory.
He placed the practice plate beside Framework Zero.
New observation: A spell model has capacity limits.
Failure may be caused by excess input, not incorrect intent.
Damaged models preserve evidence of past execution.
Study can improve control without activation.
He looked at the dead plate.
It had produced no spell.
It had not helped him gather mana.
It had cost him one copper, several hours, and a favor.
Still, it had given him a better understanding of failure.
Arin smiled faintly.
The plate was dead.
Its information was not.