Fourth revision. Issued during active deviation. Destination not assigned.
A navigation office exists to distinguish movement from error.
This is normally straightforward. A vessel moves. The chart remains. A bearing changes. The compass remains. A datum is fixed and the object is measured against it.
The Meridian assembly has made this arrangement inconvenient.
Following the drydock intake recorded in Ballast 03, the keel-blocks, the drainage manifold, the Orchard Engine, the timber associated with Graft 73, and the hull under observation were treated as one structural system. The system was expected to settle.
It did not settle.
It rotated.
Not visibly. The drydock remained where it was. The walls did not turn. The slipway did not swing around its foundations. No survey marker moved sufficiently to produce an ordinary explanation.
The strain minimum moved.
Every twelve minutes it occupied a slightly different azimuth until, at 04:17, the changes stopped. The instruments then agreed on a direction that was not north, not magnetic north, not the original slipway bearing, and not any bearing present in the charts available to the watch.
The direction was reproducible.
That is the problem.
The first instinct was to call the 623 bearing a fault in the reference frame. This is reasonable. Reference frames are old things. They accumulate errors. Datums are revised. Instruments drift. Operators misread scales. A good navigation office assumes error before assuming the world has changed.
The second instinct was to call it contamination from the previous documents.
This was less reasonable, but easier to test.
The test failed.
We removed every textual reference we could identify to 623. The bearing returned. We changed the labels. It returned. We reversed the order of the three predecessor documents. It returned. We removed the Orchard Engine from the manifest. The strain minimum moved away from the expected position and then returned to 623 after 41 minutes.
We did not remove Graft 73.
This was not because Graft 73 was necessary.
It was because it was under the hull.
Ballast 01 taught the system to retain a disagreement rather than resolve it. Ballast 02 taught it to retain a contaminated identity. Ballast 03 taught it to bear load with both.
Ballast 04 inherits the word deviation because it is the closest word we have for the condition in which a thing is measurably different from what the inherited frame predicts without thereby being wrong.
This is not a semantic refinement.
It is an operational necessity.
If 623 is an error, the correction is to force the system back toward zero.
If 623 is a bearing, the correction is to update the frame.
We currently have insufficient information to determine which of those operations would constitute the larger error.
The term was proposed verbally at 04:23.
It was entered here at 04:31.
No meeting occurred.
This is not an objection to the term. It is a record of how terms enter systems when systems are under load.
At 04:46 the navigation watch requested the original Meridian survey. The archive returned three versions.
Version A placed the drydock longitudinal axis at 117°. Version B placed it at 116°42′. Version C did not contain the drydock.
Version C contained the orchard.
The orchard was not present on Version A.
The drainage interceptor was present on all three.
The vessel was present on none.
This is not unusual in itself. Maps are selective. Infrastructure is built after surveys are completed. Orchards disappear. Drydocks move. Vessels arrive later.
What is unusual is that the angular difference between the orchard boundary in Version C and the drydock axis in Version B is 623 nm.
We checked the arithmetic twice.
Then we checked the units.
Then we stopped checking the arithmetic.
The number had already become a bearing.
A note must be made here because the term "anchor" has appeared in three separate reports this week and is being used in three different senses.
In navigation, an anchor fixes a vessel against movement.
In structural engineering, an anchor transfers load into something assumed to be fixed.
In ordinary speech, an anchor is the thing you return to when the rest of the system moves.
The fused assembly satisfies all three definitions until you measure it for long enough.
Then the fourth definition appears.
The assembly does not prevent movement.
It makes movement measurable.
This distinction should have been obvious.
It was not obvious because we were using "anchor" as a synonym for "stationary."
They are not synonyms.
An anchor can move.
What makes it an anchor is that other things acquire their position relative to it.
The reference is not the thing that does not move. The reference is the thing against which movement becomes legible.
Callis was asked whether the controls could be restored.
They said no.
This was not said dramatically. The answer was given while checking the pressure on the west cradle.
The garden cannot be separated from the timber under the keel without lifting the vessel.
The timber cannot be removed without releasing the load.
The load cannot be released without closing the Scour Gate.
Closing the Scour Gate raises the pressure in Interceptor 7.
Raising the pressure changes the strain minimum.
Changing the strain minimum removes the only presently reproducible measurement of 623.
Therefore the measurement cannot currently be repeated without destroying the condition that produced it.
This is a familiar problem in science.
It is less familiar when the apparatus has started behaving like the reference frame.
At 05:02 the watch was instructed to determine the destination associated with Bearing Zero.
No destination could be identified.
The bearing crosses the original drainage basin.
It crosses the former orchard.
It crosses the drydock.
It crosses open water.
It continues beyond the range of the available Meridian survey.
At 05:11 the watch extended the line manually using the old deep-range chart.
The line ends at the edge of the chart.
At that location the chart has a notation.
The notation is not a place name.
It is a small black anchor.
The chart predates the drydock by twenty-two years.
The notation predates the orchard register by sixty-one years.
The notation is not present on the replacement chart.
The original notation has been photographed and sealed.
The photograph has not been attached to this document.
There is no technical reason for that omission.
The reason is that the image contains something at the end of the bearing that the watch has not yet agreed how to name.
We were told the point was an anchor because it did not move.
I don't think that's right anymore.
I think it is an anchor because everything else has started moving relative to it.
There is a difference.
I have marked the difference.
It is 623.
Whether 623 is a bearing, a residual deviation, a transferred property of the fused ballast, or an artifact of the reference frame remains open.
Whether the Meridian datum was ever as fixed as subsequent documents claim remains open.
Whether the Orchard Engine was named before or after the first appearance of the anchor notation remains open.
Whether Graft 73 is the source of the observed structural compliance or merely the first material capable of accepting it remains open.
Whether the occupant of the vacant custodian post is the same person who left the wax remains open.
Whether there is a destination at the end of Bearing Zero remains open.
Whether a destination is required remains open.
The navigation office's function is not to resolve these matters.
Its function is to preserve the bearing while the matters remain unresolved.
Accordingly, no correction shall be applied to 623.
No destination shall be assigned to it.
No map shall be redrawn around it.
For the present, the line is simply there.
Do not correct a bearing merely because the inherited chart disagrees. Do not correct the chart merely because the bearing persists. Do not remove the instrument that produces an inconvenient reference. Do not declare the destination merely because a line has been found. If the reference moves, record the movement. If the movement repeats, record the pattern. If the pattern acquires a direction, record the direction. If the direction acquires a name, record the name beside the direction. Do not confuse the name for the destination.
This standing order is provisional.
It has not been approved.
It is nevertheless being followed.