Nothing happened at midnight.
At five to twelve I opened the procedure again. The first item after midnight is to start the machines in order.
Where there is an item that says start, there is a stopping before it.
There is no item that says stop. I went through the seventeen three times and there is none.
At four to twelve I switched off the operational machine and switched off the query machine.
The test machine is not covered by the procedure. There is no item to stop it and none to start it. Head office said at one to leave it.
I left it.
At three to twelve there was one screen left in the room.
There is a clock on the wall and that is the company clock. It has a second hand.
I looked at a screen to see midnight. There are no seconds on a screen. No machine in this room puts out seconds. To know the date one must query, and a query is looked at after it is made.
The changing cannot be seen. Only the after of the change can be seen.
What I did at midnight was watch the second hand of the wall clock. The second hand went up to twelve and the minute hand followed it up. There was no sound.
What the papers wrote did not happen. What was in the articles the company pasted on did not happen.
The electricity did not go. The telephone did not ring. The sound of the water was as it was. There was no sound of people on the stairs and twice there was the sound of a car in the street.
For eight months what happened in this room always had an explanation beside it. They said it was residue from a test, they said it was because the form system had changed, they said it was the limit of two-digit notation. The explanation was there and each time one place was left uncovered.
At midnight tonight there is nothing to explain. If there is nothing for an explanation to cover, there is no uncovered place either.
There is a box for writing that nothing happened. It is the box for no abnormality. In twelve months there are many days I have written none in that box. Today, for the first time, that box is entirely correct.
At two minutes past twelve I looked at the test machine. I queried that machine's date.
It came out as the first of January, year double zero.
After three months that machine became right. A machine that had put out a wrong date since October came, at midnight, to put out the right one. It was right because it was not rolled back.
There is one machine in this room that watched midnight, and it is the machine that is not in the procedure.
At seven minutes past twelve I opened the separate volume.
The previous page is the thirty-first of December, year double zero. On today's page I wrote the first of January, year double zero.
The first few pages of the separate volume are January, year double zero. So within this book January comes first, October and December next, and January again after that. Within one year January came twice.
The separate volume has no box for which volume of a year. If the same year comes twice there must be a mark to tell the earlier from the later, and there is only one year box.
Unable to tell them apart, this book is writing a year of thirteen months.
From ten past twelve until twenty to one I did the six after midnight in order.
I started the operational machine. I queried the date. It came out as the first of January, two thousand. Four digits.
I started the query machine. The same.
I queried the dates of the six programs in turn. Five came out in four digits.
The sixth came out in two.
What is entered as applied comes out in two digits.
For eight months I have watched the corrected spread wrongly again. Today the corrected has not spread, and the corrected place stands uncorrected.
That one is the index query.
At one I compared the copy. It is the twenty-second item after midnight.
The figure box of page two hundred and thirty-two was filled. It was five hundred and twenty-five.
The morning search was five hundred and twenty-two. Three have been added. It is a number that has grown by three a day, and today it is still one o'clock.
At five past one the guard came down. He said all three were in order. He said the man from funds had made coffee earlier, and went up.
At ten past one I queried the index. It is the twenty-third item after midnight.
The result count agreed. Seven thousand three hundred and sixty-two.
I looked at the order.
The first line was not twenty-three.
The first line was double zero.
Beneath double zero are twenty-three and twenty-four and thirty-four and thirty-seven. Beneath those are ninety-seven and ninety-eight and ninety-nine.
Standing in two digits, double zero is foremost. I knew that in October and then I saw it on the test machine. Today I see it in this room's index.
There is one thing under double zero.
I traced the material number. It is not something that has been in this room these eleven months and it did not come from the third floor.
It is what I registered on the fifth of November.
It is the fifth bundle.
I left the year box empty and registered it. I left it empty because there was nothing to write.
An empty box is stored as two-digit double zero. Double zero is read as nineteen hundred.
A box left empty has become the oldest year in this list.
In July I wrote that empty boxes are of two kinds. Then it was of the box left empty for want of time and the box whose writer had not yet gone. There is a third. It is empty because there is nothing to write.
That third has become the first line of the list.
At one thirty-five I copied out the top eight lines of that list.
Double zero is the first line, and the four beneath it are the four in the box, and the three beneath those are the three records of this room. All eight lines are things in the box under my desk or in the drawers of this room. Eight of seven thousand three hundred and sixty-two are so.
That the eight are gathered at the top is because of the years. Double zero and twenty-three and twenty-four and thirty-four and thirty-seven stand in front, and ninety-seven and ninety-eight and ninety-nine stand behind. The three that stand behind are this room's three records, and the last of those three is my log.
Having copied them out I looked at the eight lines once more. There is no box for a name in this list. There is only the material number, the year and the title.
Whether there are eight people in the eight lines standing by number and year, the list does not say.
At one forty I put in the report to head office. It is the twenty-fourth item after midnight. One line of no abnormality and one line for the one program in two-digit notation.
I did not write the index order. There is no box for writing an order.
At two the duty ended.
I opened the log. On which page do I write the work up to two.
There is no box in the form for the unit of a duty. There is a box for the date.
Written by date it is two pages.
I wrote on two pages. Up to midnight on page two hundred and twenty-four and after midnight on page two hundred and twenty-five. One duty spans two pages. For twelve months a day has been a page, and today one night is two.
Filling in the date box of page two hundred and twenty-five, my hand stopped.
The year box of the form is two digits. For twelve months I have written ninety-nine in that box.
I wrote double zero.
It is the first line in this room in which double zero is written by my hand.