A book scanner photographs each opening and counts the pages as it goes, and the count has to match the book, because a missing page is a rejected job. One volume in our collection scans to four hundred and twelve pages every time. It has three hundred and ninety. The extra twenty two are in the files, in the book's own hand, and they are not between its covers.
I run the digitisation suite for a county library service, and the page counts are how my work is checked.
How a book is scanned.
It goes on a cradle under two cameras. Each opening is captured as two images, the operator turns the page, and the software numbers the images in sequence. Before we start, the book is collated: somebody counts its pages and writes the number on the sheet, and at the end the image count has to match, and the images are checked against the book by a second person.
That is the whole quality system. Count, scan, count, check.
One volume, a bound manuscript ledger from the eighteen hundreds that came to us from a parish, scans to four hundred and twelve images.
It has three hundred and ninety pages. I have counted them. Three people have counted them.
The twenty two extra images are pages.
Not blanks, not duplicates, not the cradle. Pages, in the ledger's ruled format, in the same hand as the rest of the book, with entries in the same style, numbered in the top corner in the same ink, with numbers that fit into the sequence, between pages that are consecutive in the physical book.
The physical book goes from page two hundred and six to two hundred and seven with nothing between. The scan has pages two hundred and six A through two hundred and six V between them, twenty two of them, lettered in a hand that looks like the clerk's.
What has been checked, because a scanner producing pages is a scanner that would be sent back.
The scanner. It has been recalibrated and its software reinstalled, and the volume has been scanned on the other scanner in the suite, which is a different make. Four hundred and twelve.
The operator. Three of us have scanned it, including me, and the extra images appear on the turn from two hundred and six to two hundred and seven, and I have watched my own hands turn a single leaf while the software captured, counted, and moved on, and the screen showed a page I was not looking at.
The book. Collated four times, with a ruler in the gutter to find stubs where leaves might have been cut out. No stubs. The sewing is original and continuous.
What is on the pages.
Entries. The ledger is a parish account book, receipts and payments, and the twenty two pages are more of the same, in the same hand, dated inside the same year, for sums and names that do not appear anywhere else in the book.
I have had an archivist look at them, from the images, without telling her they were not in the book. She read them as a genuine continuation of the accounts and remarked that the clerk had been unusually busy that quarter.
The names on those pages are not in the parish registers. I have checked all of them, and there is a name that recurs on every one of the twenty two pages, receiving small sums, that appears nowhere else in any record I can find.
What is done about it.
The digital copy is the digital copy. Our policy is that images are not deleted from a job, and I have not deleted them, and the volume's online record shows four hundred and twelve pages and a note that says page count differs from collation, which is a category we have because bindings are unreliable.
It is not the binding.
What I do.
I scan the collection and I keep the counts.
The one change is that I do not scan that volume, and it is not on the list to be rescanned at the higher resolution when the equipment is upgraded, and I have made sure it is not by putting a note on its record.
The last time was 2023, for the third check, at the end of a day with the suite empty. At the turn from two hundred and six to two hundred and seven I stopped and did not turn the page, and I looked at the screen, and the software captured anyway, and the image it showed was two hundred and six A, and I looked down at the book on the cradle, and my hand was on page two hundred and six, and the leaf under it was warm.