By early afternoon the following day, Khepri had discovered a serious flaw in his latest patient carrier. The suspended bed softened a downward jolt. It also swung sideways whenever the outer frame turned.
Ibara rocked the test rig with one hand. Behind her, the bells of the medicine wagons sounded as the afternoon supply round began. The inner platform swayed towards its supports, narrowly avoiding the edge.
Khepri caught the frame.
“That,” he said, “was not in the design.”
“It was in the design,” Ibara replied. “You hung the bed without lateral stops.”
He tightened the suspension cords and tried again. The sideways movement improved, but the platform became rigid enough to transmit every jolt from the outer frame.
Another adjustment made it barely move at all.
“No sideways swing,” he said, testing the rigid bed with one hand.
“Now every jolt goes straight to the patient.”
He put down his tools.
“How do I allow it to move up and down without letting it swing wherever it pleases?”
“There are several ways.”
“Which would work for a patient?”
Ibara returned to shaping a bronze instrument housing. Khepri waited.
“Are you going to?”
“When I finish the commission paying for your bronze.”
He looked across her workbench. Three half-finished housings lay beside a measuring gauge, each marked with precise dimensions. He had been occupying her best bench for most of the morning and had hardly noticed the work waiting around it.
“Can I help?”
She glanced at him.
“Would you follow my measurements, or would you invent a better instrument halfway through?”
“That depends on whether the measurements are good.”
She held his gaze.
“I’ll follow the drawing,” he said.
Ibara selected prepared pieces of bronze and placed a drawing between them.
“Two fittings. Exact shapes. No experiments.”
Chi moved through Khepri’s fingers. The metal separated into bright fragments and lifted above his palm. He gathered the fragments into the shape of the first fitting, checking the drawing before letting the piece settle.
The second took longer. A small curve needed to match the instrument housing with very little clearance, and the effort left an uncomfortable tremor in his hand.
He completed the form and set it beside the first.
“There.”
Ibara examined both pieces with a gauge.
“One meets the drawing. The other needs finishing.”
“It looks right.”
“It nearly is.”
“Where does the curve miss?”
“A physician needs a measured fit, not your guess.”
She handed him a file.
Khepri eyed it. “I have a Spark.”
“And I have a file.”
“The edge still needs filing.”
“File it to the line. Then check it again.”
He took the tool. Working the uneven edge by hand was slower and less impressive than Assembly. It also let his tired Chi circulation recover. By the time the fitting passed inspection, he had begun to appreciate the satisfaction of making a small part right without relying on power.
Ibara collected the pieces.
“Thank you. Now, about your suspension.”
Khepri straightened.
“Where have you seen restraints like this?”
He blinked. “Their instrument supports?”
“Ask the researchers how they stop the samples from striking the frames.”
He stared at his suspended platform. The research wagons might have the answer.
* * *
At seven, Khepri had concluded that Elune’s research wagons were the dullest places in the caravan.
There had been nothing to race, nothing to build and nothing that anyone permitted him to dismantle. Instead, he had spent an afternoon looking at jars with labels whose handwriting was more interesting than their contents.
Three years later, he entered the same row of wagons with a different opinion forming.
The outer walls were painted ivory, their shutters lined with indigo. Within, each wagon held a particular kind of work: anatomical collections, medicinal studies, preserved samples and instruments that needed careful protection from the movement of travel.
A sound came from a nearby bench—wood scraping gently against wood. Two researchers were rocking a broad supporting frame. Inside it, a glass vessel stood almost still. Khepri stopped. They rocked the outer frame again. The vessel shifted a little, then settled.
“How are you doing that?”
One researcher glanced up. Sẹ́ni, standing at the far table with a record book, answered first.
“Ibara sent you?”
“Yes.”
“Then keep your hands in your pockets until you’re invited.”
Khepri tucked them behind his back instead. The demonstration frame had two layers. The outer structure absorbed the movement of the wagon; an inner support carried the vessel. Cords and adjustable restraints joined the two, but the inner support did not hang freely.
He crouched beside it.
“The restraints prevent the sideways swing.”
“Some of it,” the researcher answered.
“Why not all of it?”
“Because the wagon moves in several directions. Hold the vessel too rigidly, and every jolt reaches it. Let it move too freely, and it strikes the frame.”
That was exactly the problem awaiting him in Ibara’s workshop.
A smaller model was brought forward. Khepri tested it cautiously. Certain connections stretched with the platform’s movement; others engaged only when it travelled too far to one side.
“So you control the limit,” he said, “rather than preventing movement.”
“That’s part of it.”
He turned the model and watched the restraints engage from a new angle.
“Can I copy it?”
“You can study it,” Sẹ́ni said. “Copying it without understanding the load would put someone at risk.”
“People aren’t glass vessels,” Khepri agreed. “Different weight, different movements. I’d need to adapt the design.”
His aunt seemed pleased with that answer.
Khepri moved to the next row of shelves.
The glass vessel was only one part of the problem. Some shelves were insulated. Others held narrow cases within larger compartments, so that the wagon walls did not pass every movement straight into the contents. A cabinet at the back had sealed channels and a small Chi-powered regulator.
“What does that do?”
“Keeps some samples under controlled conditions,” a researcher explained.
“Why combine insulation with Chi regulation? Why not use Chi for everything?”
“Insulation does most of the work. The regulator uses Chi only to maintain the conditions it can’t hold by itself.”
Khepri studied the arrangement. Passive protection first; Chi used where it provided an advantage that simpler materials could not.
Khepri read the labels on the containers: dates, distances, settlements he had never visited. Some of the samples had travelled farther than he ever had.
“And if the regulator fails?”
“We follow the handling procedure for that material. Some samples can be moved safely. Others won’t remain useful if their conditions change.”
Beside them were records written in several hands. One handwriting style caught his attention. He recognised the slant of the letters before he could make out a single word, and stopped turning the pages.
Father.
Khepri moved closer.
“Those are Dàre’s notes.”
“Some of them,” Sẹ́ni said.
Dàre had spent years studying resistance to illness and reliable treatment. Khepri had heard about the research whenever his father returned from a journey.
He had not realised how much of it travelled on these shelves.
A record described observations from a settlement beyond the northern hills. Another held samples collected far to the west. Their margins held corrections from different physicians and transport notes in unfamiliar handwriting.
“Because the people we serve don’t all live beside a permanent hospital,” Sẹ́ni answered. “And because the same illness may behave differently in different places. We learn more by travelling.”
Khepri examined the quiet machinery around him.
The transport schedule on the shelf recorded adjustments made by the drivers after difficult journeys.
One cabinet hinge had been mended with braided wire and a thin strip of horn. Khepri bent to examine the join.
“Who designed the suspension?” he asked at last.
“The instrument-makers drew the first version,” said the researcher. “Our transport team helped improve it after testing on the road.”
“So neither could have built the current one alone.”
“Probably not this version.”
Khepri sketched a hinge on the back of a scrap of paper, then scratched it out. It would let the carrier move freely until the first sharp turn knocked a patient against the frame.
He found a table with technical drawings.
“May I copy the measurements?”
Sẹ́ni glanced at the researcher, who nodded.
“Under supervision,” she said.
“Of course.”
“And you can’t take the model.”
“I hadn’t asked.”
“You were about to.”
Khepri accepted a sheet of paper. It took considerable self-control not to try the apparatus at once. Instead, he measured the connections, noted their lengths and compared what happened when the frame moved in different directions. The records included results from journeys over uneven roads. Some earlier designs had failed; others performed well for one kind of cargo and poorly for another.
He copied those failures too. One set of drawings had been crossed out so thoroughly that he could barely see the original design. He asked what had gone wrong.
“The inner platform stayed level on the test bench,” the researcher told him. “On the road, it struck the retaining frame whenever we turned downhill.”
“So it passed the wrong test.”
“It passed one test. That wasn’t enough.”
She let him use a retired demonstration model, provided he changed nothing without permission. Khepri placed a small wooden weight inside and tipped the outer frame downward; the weight barely shifted. He then twisted it sideways, imitating a cart rounding a bend, until a restraint tightened and drew the drifting inner platform back.
He tried with a heavier weight. The motion changed. The return was slower, and one corner dipped lower.
“Same restraints,” he muttered. “Different load.”
“That is why the transport team tested more than one arrangement.”
He noted the result, then asked whether the model could be adjusted. Under the researcher’s supervision, he shortened one restraining cord. The heavier weight stayed closer to the centre during the sideways test, but when he moved the outer frame sharply downward, the platform jolted more than before.
Khepri sat back on his heels. He set the heavier weight on the platform again. The shorter cord held it steady through the turn but transmitted the next downward jolt. One edge of the model was scarred where a weight had struck it during testing. No careful movement of the bench could recreate a real road.
“Could I speak to the transport team?” he asked.
“You may,” Sẹ́ni said, appearing behind him. “After you have written up what you found here.”
He checked his notes. Several cord lengths were crossed out, with the failed test written beside each.
“And if the next design fails?” he asked.
“Then you test the next one carefully,” she replied. “And you write down what failed.”
He imagined showing Ibara three new problems, each marked with measurements instead of guesses.
At the bottom of his page, he wrote a question: *What movement can the patient tolerate?*
He underlined the word patient twice. He would need to ask Simi and Tàlé.
He left the research wagon carrying the notebook close against his chest.
Behind him, Sẹ́ni called, “Khepri.”
He turned.
“Yes?”
“Did you return the model?”
He held up his empty hands. “It’s on the shelf.”
“And nothing is missing?”
“I copied the measurements. That’s all.”