Chapter 63: The Duplicate That Burned

The portable saddle could not be examined until three people agreed to make it movable again.

Tertia held the stone cradle. Mara held the seal across the release band. The Examiner held the retaining pin and its numbered witness tab. Beneath their separate restraints, the saddle gave back every vibration as a small, imprisoned complaint.

Rain ticked against the high windows. Wax, wet wool, and old machine oil thickened the inquiry room’s air.

A seam opened in Tertia’s eastern wall. Her writing rose from the slate within it.

Comparison authorized: fourth-socket fitting, indexed witness strip, established RST-12-B defect record, returned wafer, East Three admission, cabinet return, and winding log. No E4 access. No unrelated enclosure.

The Examiner read it aloud.

“Does your consent include a non-winding interlock test after removal?”

If the test cannot produce a code, advance an evidentiary index, or damage a recorded surface.

“It cannot.”

Mara broke only her saddle seal. The Examiner withdrew his pin. Tertia lowered one side of the cradle, leaving the mechanism bounded on the other three.

Each action entered the custody sheet before the next began.

The documents were simpler once nobody asked one page to prove everything.

The destruction return recorded RST-12-B as destroyed. That return was already established as false.

The identity record established that the authorized duplicate designated RST-12-B had survived that entry. Its authorization, identity, and continued existence were not under examination again.

The Renewed Continuity service order placed that established RST-12-B into controlled service through this portable saddle.

East Three’s admission book recorded the saddle entering the station fourteen minutes before the seven-minute gap.

The cabinet register certified RST-12-B as returned nine minutes before the gap.

Finally, the winding log and seized wafer recorded a second-code pulse three minutes into it.

“Twelve minutes after the certified return,” Mara said.

“Which proves that a second-code pulse occurred,” the Examiner replied. “Not yet that RST-12-B occupied the saddle.”

Tertia marked his objection onto the wall.

Accepted. Continue.

Mara opened the saddle’s service packet. At Tertia’s request, she narrated what her hands and eyes found.

“The return seal is intact except for the previously recorded trimmed registration margin. The cut is at the same angle and location as the margin already under inquiry. I am not attributing an owner.”

“Recorded without attribution,” said the Examiner.

Mara turned to the equipment diagram.

“The fourth position is not a winding socket by itself. It receives removable witness fitting F-4-771. The fitting carries a sequential pressure strip and completes the saddle’s engagement interlock.”

The old inventory had listed only three ordinary fittings. The service packet named the missing fourth component, gave its serial number, and instructed that it be removed before depot return.

It had not been removed.

Tertia felt Mara turn the saddle. Metal rasped softly against its bounded supports. A catch resisted, released, then stopped against the custody cord threaded through it.

“Fourth socket exposed,” Mara reported. “The fitting is present beneath the cover. Serial visible: F-4-771.”

The Examiner checked the packet number and repeated it for the record.

Removing the fitting required both of them. Mara held the release collar without turning it. The Examiner inserted a numbered extraction fork through the collar’s two opposing eyes. Only then did Tertia relax the narrow stone brace preventing rotation.

The fitting came free with a dry metallic rasp.

Cold brass touched the padded evidence tray. Beneath its oil and tarnish lingered the sharper smell of heated waxed fiber.

Mara described an asymmetric polish along the fitting’s outer shoulder, matching the one-sided wear previously found in the fourth socket. The Examiner measured both arcs. Their widths and load direction agreed.

“That establishes which component made the socket wear,” he said. “It does not establish which key entered the component.”

“Open the strip chamber,” Mara said.

“Not until its seal is authenticated.”

The chamber carried a depot wire beneath two layers of darkened wax. Mara read the impressed service number. The Examiner compared it with the service order, then used side-light and a measuring fork while Mara described the result for Tertia: no lifted edge, no replacement knot, no second wire channel. The deliberately trimmed registration margin appeared again on the outer return mark, but the chamber seal itself remained whole.

The Examiner signed the authentication line.

Mara cut the wire between its numbered witness beads. When she lifted the chamber cover, a spring relaxed with a faint click.

The indexed strip emerged reluctantly. Its waxed backing dragged across the guide with a papery rasp, each numbered cell passing under Mara’s gloved fingers.

“Strip serial WS-771-4,” she said. “Continuous from the authenticated head. No splice visible. Sequential cells intact through two hundred seven. Two hundred eight is unstruck.”

The Examiner checked the serial against the service packet.

“Agreed.”

RST-12-B’s settled identity record included more than its code. During its earlier authentication, examiners had recorded a unique defect on the lower contact edge: a flattened shoulder ending in a raised burr. Under pressure, that defect produced a short deep bite followed by a shallower feathered drag. The interval and angle had been measured before the present saddle was seized.

Mara laid the established defect plate beside cell 207.

She did not announce a match.

Instead, she described cell 205: an ordinary centered pressure fan. Cell 206 bore another complete impression without an edge bite. Cell 207 began with a narrow depression near its lower margin. A shallow drag extended from it at an angle.

The room held only rain, the lamp’s hiss, and the Examiner adjusting his depth fork.

He tested the depression at three points. At the first, the fork descended cleanly. At the second, it met the displaced ridge. At the third, it followed the feathered drag until the waxed fiber rose again.

“Distinct resistance at the damaged edge,” he dictated. “Deep bite, displaced ridge, trailing drag.”

He compared the distance from bite to ridge against the established defect plate.