The decision made in the previous days was not revisited.

  Training resumed the very next morning, not with excitement, but with routine. There was no ceremony to mark the start—only repetition, correction, and exhaustion. Fire was taken first, it was powerful in experienced hands. If it was going to be used later, it needed to be understood early.

  The first twelve days were spent on a single spell.

  A basic fireball.

  Not shaping flames, not increasing temperature—just constructing it correctly, again and again, until the structure no longer collapsed mid-casting. Most attempts failed due to wasted mana or unstable form. When it finally worked, it was small, unremarkable, and costly for what it did. That was acceptable. Function mattered more than appearance.

  After fire, the training shifted without pause.

  Light followed for fifteen days. The spell itself was simple—illumination, nothing more. Mana was distributed evenly, spread thin rather than condensed, producing steady visibility instead of intensity. It served no combat purpose and was treated as such. Darkness was not opposed, corrected, or erased—only displaced where vision was required.

  Earth came last.

  Thirteen days were spent learning its limits.

  There was no creation involved. No raising walls or forming weapons. Only manipulation of what was already present—adjusting ground elevation, forming shallow bumps for footing, stabilizing loose terrain. Every attempt depended on surroundings. Without material, there was nothing to shape. It was restrictive, but reliable when conditions allowed.

  Magic training concluded without announcement.

  Parallel to it, the sword never left his side.

  Over the same forty days, quick draw was repeated until the movement no longer required thought. Grip, pull, strike—clean and efficient. No wasted motion, no hesitation. The blade came out when needed and stayed sheathed when it wasn’t. By the end, speed was no longer forced. It simply happened.

  Unauthorized duplication: this narrative has been taken without consent. Report sightings.

  The learning phase was over.

  Now came application.

  After that, time blurred.

  The following two months were different.

  What had been learned was no longer practiced in isolation. Sparring became regular, sometimes daily. Spells were cast while moving, while reacting, while thinking ahead. Fire was used to pressure, not overwhelm. Light maintained visibility under strain. Earth adjusted footing mid-exchange when terrain allowed.

  None of it was enough.

  Against either of them, the result never changed.

  He did not land a single hit.

  Distance was closed before spells finished forming. Openings disappeared the moment they were identified. Even when timing felt correct, it wasn’t. Every exchange ended the same way—with him disarmed, outmaneuvered, or simply stopped.

  The gap was clear.

  Progress had been real.

  But it was not sufficient.

  The problem was no longer knowledge.

  It was combat experience.

  ___

  The conclusion was unavoidable.

  Sparring showed me what I lacked, but it couldn’t give it to me. Controlled exchanges ended the moment they reached their limit. Past that, they only repeated the same result.

  Combat experience couldn’t be learned in a courtyard.

  That left only one option.

  Dungeon.

  Mana flowed through the earth the same way water flowed through soil—unevenly, under pressure, following paths that weren’t always stable. Most of the time, those flows corrected themselves.

  Sometimes, they didn’t.

  When a flow became disrupted long enough, mana accumulated. Density increased. Structure followed. Eventually, the pressure forced an eruption—not upward, but outward.

  That was how dungeons formed.

  Layers developed where mana concentration differed. Environments stabilized around those layers, and creatures formed where excess mana condensed into physical patterns. They weren’t summoned or designed. They were produced.

  The same process created resources.

  Herbs grew warped by saturation—some toxic, some valuable, some capable of producing effects that didn’t exist outside a dungeon environment. Their rarity wasn’t intentional. It was a byproduct.

  Dungeons weren’t malicious.

  They were inefficient.

  And until they exhausted themselves or were cleared often enough to destabilize, they remained.

  The nearest dungeon lay less than a day’s travel from our territory.

  It wasn’t considered dangerous. The layout was mapped, the monster population stable, and clearances happened regularly. It was used by parties that needed exposure to risk without inviting death.

  Suitable for survival.

  Not for comfort.

  Lyra didn’t object when I brought it up.

  She reviewed what I could do, what I couldn’t, and what that meant in an uncontrolled environment. Equipment was adjusted. Limits were set—depth, duration, retreat conditions.

  “If anything deviates,” she said, “we leave.”

  I agreed.

  I didn’t expect the deviation to come from the dungeon itself.