Chapter 363: What Arrived Late

The sealed system — a watch post.

The keeper on duty compared the reporting sheet with the open ledger. Two successive checks had produced late returns above the posted range. The first had been observed on one measuring line, the next on the other. Beside each entry, the reader had confirmed that the instrument’s settings matched the procedure card.

The keeper requested higher review, attaching the report and a checked copy of the relevant ledger entries. The original ledger remained at the post.

Keeper Tamar Ost arrived with the request. She had authority to direct the review and amend the watch’s procedures. While the duty keeper attended the instruments, she sat at the desk with the reader and recorder and opened the ledger at the last normal entry before the deviation. ℴ

"These two," the reader said, indicating the later times. "The first return was read on the first line. We used the second line for the next check. They were separate returns."

Tamar followed the entries across the page.

"Both late. Both above range."

"Yes."

"Show me what you checked afterward."

The reader brought the procedure card. The recorder laid out the interval cards used for those observations, with their expected windows and limits. Tamar compared them with the report before going to the first instrument.

The scheduled check was part of the maintenance procedure: a prescribed working sent through a designated section of the containment structure, with a return the post could measure. The keepers maintained the containment continuously. These instruments measured its response to that particular working; they did not report an opening whenever a return arrived late.

"We’ll check the measuring equipment first," Tamar said. "One line at a time. Keep the other attended."

They traced each measuring line to its connection and checked the instrument against a reference input. Both readings fell within the reference limits. The timing references agreed with the post’s master clock, which they also checked against a separate time standard brought for the review.

The recorder entered the results beside the settings used. After each reference test, the measuring line was restored to its ordinary connection and the reader checked the setting again. The test times remained in the record, separate from the times of workings sent through the containment structure. Tamar examined the completed checks before authorizing a third instrument.

It was installed beside the first, on a separately checked measuring path, with its own timing reference. The second instrument remained across the post. The reader compared the third instrument’s response to the reference input while the recorder added a column for it.

For the next scheduled check, all three would observe the same return. Each line had its own expected arrival window and range, recorded on the card placed beside its instrument.

"We haven’t found a fault in these checks," Tamar said. "That doesn’t exclude a fault shared by the measuring paths. Keep the results with the return readings."

"Will you send the checking working?" the duty keeper asked.

"The assigned keeper will send it as prescribed. I want the same input and schedule."

The recorder prepared the entry with the scheduled time. When the sending record confirmed the working had gone out, he entered its actual time beside it and checked the section and input against the procedure.

The reader watched the first instrument, Tamar the third. Across the post, the duty keeper took the second. The recorder kept the ledger open where he could hear all three.

The first instrument’s expected window closed without a mark appearing. The reader called the time and continued watching. At the other instruments, their windows also ended before a return registered.

When the first mark appeared, it stood above the posted limit.

The reader gave its arrival time and value. The recorder entered both, then repeated them for her to confirm. Across the post, the duty keeper called the second line’s reading. Tamar supplied the third.

All three observations were late and above range against their respective limits. The values differed slightly between the measuring paths, but each recorded the same kind of deviation as the earlier checks.

Tamar waited while the entries were checked against the instruments. The recorder grouped all three observations beneath the same sending entry. The earlier report retained its two separate checks; this test added one further check observed through three paths. Then Tamar returned to the desk and placed the new results beside the original report.

"The deviation has recurred under the prescribed checking conditions," she said. "Late and above range. The cause remains unestablished."

The recorder wrote the conclusion beneath the comparison. She read it before giving him the revised watch instructions.

The third instrument would remain connected. Observations would be taken at shorter intervals, including between the expected return windows, and further deviations would be reported directly to Tamar as well as to the duty keeper. The checking schedule and prescribed inputs would remain unchanged during that observation period.

The reader drew the interval card closer and followed the revised spacing with her finger.

"We’ll need another reader to keep all three attended at these intervals."

"I’ll arrange the cover. Until it arrives, the keeper and I will take the other stations."

The recorder added the staffing arrangement to the order. Tamar checked that line, then asked him to leave space beneath it for a further instruction.

Compare every observed response with the keepers’ sending record, including earlier workings whose returns remain outstanding. Record any response that cannot be matched as unmatched and refer it for immediate review.

"Include the possible matches with the report," Tamar said. "If a return could belong to an earlier working, keep that uncertainty in the entry."

The recorder looked up. "Even if it falls inside the next check’s window?"

"Yes. We now have late returns. The nearest sending time may not be the right one."

He read the instruction back, then placed the revised orders with the original report, the instrument checks and the new observations. The review record referenced the corresponding ledger entries. Tamar signed the orders and left the working copy at the desk.

The reader returned to her station and reached for the stool. Tamar drew it forward and sat at the first instrument instead.

"Take the third for now."

The reader set its interval card within reach, then stood beside her with the first card while they checked the next expected window. Across the room, the duty keeper remained at the second line. The recorder turned to a fresh row in the ledger.

The next scheduled working was sent at the prescribed time. He entered the confirmed sending details, and the observers watched their instruments.

A mark appeared on Tamar’s scale within its window. She read the time and value aloud. Both fell within the posted limits.

The other lines registered the return within their own windows. The reader called the third instrument’s value, checked its card once more, and looked toward Tamar.

"Normal on all three."

Tamar compared the first reading with the recorder’s entry. The duty keeper and reader confirmed theirs. He completed the row with the result and kept the ledger open for the closer observations ordered between checks.

The reader replaced the interval card beside Tamar with the revised one. Tamar checked the next observation time and remained on the stool, facing the instrument.