May 9. A tide has been added.
At half past nine I took out sub-number three.
The lighthouse keeper's table of tides. Of thirty-four, a hundred and ninety-eight pages.
One page is two days. Each line has a date and a tide name and the weather.
The tide names are numbers. From the first tide to the thirteenth. The day after the thirteenth returns to the first.
I copied it by hand over fourteen days.
At twenty to ten I ran off the query record.
Over fourteen days the set's row has been opened a hundred and thirty-two times.
In the account field, outside a hundred and thirteen and the eight Latin characters nineteen.
The eight characters are one. How many outside is I do not know.
From ten to eleven I counted by name.
The original has thirteen names.
The copy has fourteen.
The fourteenth name is the fourteenth tide.
It stands twenty-six times in the copy.
It is not in the original. I went through a hundred and ninety-eight pages twice. Not once.
From half past eleven to twelve I looked at the order.
A tide is an order. Insert one and everything after shifts.
Nothing shifted.
In the copy the day after the thirteenth is the first. The cycle stands.
The fourteenth tide is anywhere. It comes after the third and it comes after the ninth. Its twenty-six positions have nothing to do with the order.
The names are fourteen and the order is thirteen.
At ten past twelve I looked at the writing in those twenty-six places.
It is my hand. The same weight, the same slope, the same pen as the lines before and after.
I chose one and wrote the fourteenth tide beside it now.
The same.
There is nothing to tell it by in the writing. I wrote it.
When I wrote it I do not know. Over fourteen days I wrote a hundred and ninety-eight pages, and of those twenty-six lines I have no memory of writing any. I have no memory of four pages' worth of the rest either.
At half past twelve I found those twenty-six days in the original.
All twenty-six have an empty tide box.
The lighthouse keeper did not write a tide on those days.
From one to two I looked at the weather boxes for those days.
All twenty-six carry the same character.
Fog.
At ten past two I looked for why thirteen is thirteen.
It is on the first page. He divided the days from least water to most into thirteen. Thirteen because at that shore the water takes thirteen days to come round once.
Thirteen is that shore's number. It says elsewhere is different.
For a fourteenth to exist it would have to take fourteen days to come round.
In the copy it still takes thirteen days.
Only the names are fourteen. There is one name more, and no day for that name to point at.
From half past two to three I looked for how a tide is decided.
It is on the lighthouse keeper's first page.
It says it is decided by looking at the water. In the morning he goes down to the shore, sees how far the water has come, and gives it a name.
In fog he cannot see.
It says a day not seen is left empty. It says he does not write by guess.
In the place where he did not write by guess, the copy wrote a name.
That name is nowhere in the original.
From three to four I counted all the original's empty boxes.
Thirty-one days.
Twenty-six are fog.
Three are the last three days. Three lines running from the back of page a hundred and ninety-seven onto page a hundred and ninety-eight, where the weather box is empty too. The table ends there.
Two read clear in the weather box with only the tide left empty. Why they are empty is not written.
I looked at the copy.
The last three days are empty in the copy too.
The two clear days are empty in the copy.
What is filled is the twenty-six fog days and nothing else.
Not every empty box was filled.
Something chose.
Whether the choosing was my hand there is nothing to tell by.
If I chose, then over fourteen days I chose twenty-six times by the same rule. And never once knew the rule.
If I did not choose, then there is a choosing side other than me.
At a quarter past four I made one test.
I chose one of the twenty-six days. I copied that line of the original onto fresh paper again.
I left the tide box empty. The original being empty, I left it empty.
I looked at it after writing. It is empty.
An hour later I looked again. It is empty.
Fourteen days cannot be counted by one hour. This test tells nothing apart.
I did it all the same. Not doing it would have left the not doing.
At twenty past four I opened the right-hand material.
Three hundred and forty-seven pages. The date of the added page is 7 June 98.
My log is on its three hundred and forty-seventh page today.
At twenty to five I opened the box.
Seven. Six impressions. The seventy-fifth day.
At five I looked at the sixth box.
Eight hundred and twelve.
At ten past five I wrote the ten-line summary for sub-number three.
I wrote that there are thirteen tide names.
By the copy it is fourteen.
The procedure says to draw the table of contents and the summary from the transcript. From the copy, that is.
Writing thirteen, I went against the procedure.
I wrote why in the margin. It is the first margin entry this month.
At twenty to six I counted the dates of the twenty-six fourteenth-tide days.
Nine are December.
One of them is the sixteenth of December, thirty-four.
That is the fourth time this room has seen that date.
Four different years, and four different people's records.
Next is the telegraph operator's.
The librarian's gained a row. The surveyor's grew in the values inside the rows. The lighthouse keeper's had a name put into an empty place.
All three on the copy's side. In all three the place that increased is different.
The place that increases moves each time I copy.
The telegraph operator's has dates.
What dates become when they increase I do not know today.