In addition to the image layer from the scanned old book, PDFs can have an "invisible" layer over that that is the optical character recognition (OCR) text. You can drag your mouse over the page and select and copy the invisible text. There are typically many errors in the OCR of old books. As far as I know the only time you can have this dual layer of the original image and the OCR'd text is with a PDF. The EPUB you downloaded was the OCR'd text from the PDF. You can also find many of these dual layer PDFs on archive.org.
|