Typically, ancient papyrus codices were made by cutting off sheets from a long roll of blank papyrus, stacking the sheets, and folding them into quires. There is, however, a small group of somewhat odd papyrus codices that were made in a different way. A number of papyrus codices, some of which can be associated with the Egyptian city of Panopolis, were made by reusing the blank sides of documentary rolls. Two papyrus rolls containing old documents were glued to one another by pasting together the written sides (with horizontal fibers) to create a double-thick roll of papyrus with two blank sides (with vertical fibers). This double-thick roll was then cut into sheets which were folded into a quire to form a kind of codex.
Because it was so well preserved, probably the best known example of this type is the tax codex in the Chester Beatty Library, which is a part of the so-called “archive of the descendants of Alopex.” This single quire composed of double-thick papyrus sheets contained a personal collection of tax receipts from the years 339-345 CE. The codex was disassembled in the 1950s, and the “recto” sides of the two pasted rolls were found to contain official correspondence from the years 298 and 300 CE. In 1964, Theodore Skeat published the text of the “recto” (the older official correspondence on the rolls, now known as “P.Panop.Beatty” copied along the fibers), while the “verso” texts (the more recent accounts copied in the codex against the fibers) were published a few years later and now known as part of the “P.Panop.” series. The Beatty tax codex gives us a good sense of what one of these books looked like in finished form:
Chester Beatty Tax Codex Before it was Dismantled; image source: Skeat, Papyri from Panopolis in the Chester Beatty Library (Dublin: Hodges Figgis & Co., 1964), plate I.
And the images of the sheets after the pasted rolls were peeled apart gives us an idea of the original reused documents:
A column from P.Panop.Beatty 2; image source: Skeat, Papyri from Panopolis in the Chester Beatty Library (Dublin: Hodges Figgis & Co., 1964), plate III; [Update: 9 April 2019: See color images of the dismounted “leaves” here.]
In the case of the Chester Beatty tax codex, the reuse of the codex was of a documentary nature. There are, however, some examples of Christian literary works that were preserved in this type of codex composed of reused documentary materials. I will discuss them in future posts.
Sources:
Gascou, Jean. “Les codices documentaires Égyptiens.” Pages 71-101 in Alain Blanchard (ed.), Les débuts du codex. Turnhout: Brepols, 1989.
Skeat, Theodore C. Papyri from Panopolis in the Chester Beatty Library. Dublin: Hodges Figgis & Co., 1964.
Youtie, L.C., D. Hagedorn and H.C. Youtie. “Documents from Panopolis II.” Zeitschrift für Papyrologie und Epigraphik 8 (1971), 207-234.
Variant Readings 
Pingback: P.Ryl. 1.1: A Datable Papyrus Codex of Deuteronomy in Greek | Variant Readings
Brent, I have just come across this page (in particular the picture of the codex before disassembly), and it has made wonder how close this is to how P46 originally looked. Do you have any information on that, in particular how thick the whole codex was in assembled form? Also, it is clear that P46 was written after the outer edges of the pages had been trimmed, and because of this I have always assumed that it was written on in assembled form. However, is there any evidence to suggest either that it was never actually bound, or that it was trimmed, then written, and only then bound?
Well, it’s usually estimated that P46 had 104 leaves. So, when thinking about the thickness of a single-quire codex with that number of leaves, you would probably be looking at something thicker even than Nag Hammadi Codex III (78 leaves), which is pretty thick (I made a scale model of it, but I can’t recall the exact thickness). There do appear to be remnants of binding holes along the edges of the leaves of P46, and they form a pattern not incompatible with the method used to bind most of the Nag Hammadi codices (two tackets through the center fold).
Thank you. The thing I’m having difficulty with is the thickness of 104 leaves. I’ve found very little on how thick each leaf would be, but in any case the picture on your blog shows that each leaf is unlikely to be perfectly flat (although I know that in that case the codex is made of two sheets stuck together), so that a codex of 104 leaves is likely to be more than 104 times the thickness of one leaf. Also, I think the info on the binding holes confirms what I see from the differences in average line length between verso and recto, which is that it was written on after having been bound.