The last chapter of the book is called "Toward the Middle Ages", so it really deals with what is prehistory to us, but I'd still recommend it if you are interested for example in library classification/organisation or in how books were acquired, i.e. copied and spread, in the ancient world. The book has a reasonable list of literature and sources, but it is not an academic treatise and I found it easy to read and suitable for e.g. commuting literature that one reads in smaller bits (I started on the plane and had it as bedtime literature after that, reading a few pages at a time).
A small snippet of information to tickle your imagination:
In the ruins of the Royal Palace of Ebla, a Sumerian palace near Aleppo in Syria, burned down by invaders in 2300-2250 B.C, a room was found that had to be a working library of the palace scribes. With mounts of administrative records on clay tablets, there was a small collection of tablets with among others word lists, many bilingual in Sumerian and Eblaite. This collection would have fitted on one shelf and the tablets could have easily been browsed through when checking on a word. This was over 3000 years ago!
However, what really got me thinking is the transition of book formats that happens during the period the book describes. After the small excursion into the clay tables, comes the time of long rolls of papyrus leaves, two layers set crossways and pounded to stick together and rolled up - the (papyrus) roll. Codex, our book, doesn't appear until around 100 A.D.
The Greek and Roman libraries - including the famous one in Alexandria - stored their book rolls piled up in niches. At least in the Roman libraries each niche for books would need to be fully lined with wood, i.e. have a book case, to isolate the rolls from the cold and damp wall and thus to stop them from going moldy; a precaution a lot less needed in the drier Egyptian climate. Each roll had a tag at the top end that had the basic information: whose book, which part of that book and in libraries often also a locator mark (bookcase number, for example). Book rolls were fragile, so library assistants would go and get them and bring them to the reader in a bucket, tags up, instead of letting visitors browse the shelves freely.
(BTW, most libraries in the non-Greek parts of the Roman empire seem to have had two sections, often in opposite ends of a building complex: the Greek library and the Latin library. In the Greek area one would not bother with a Latin section or it would be really small - Latin may have been the language of engineering, politics and bureaucracy, but it never won over Greek in the more cultured setting!)
The advantage of the papyrus roll is that material was pretty easy to get and apparently doesn't need too much preparation, either. But the roll format has quite many disadvantages: you need two hands to read a book roll (the classic era roll is read or "scrolled" sideways!) and the text can only be on one side of the paper. Because of the latter, the famous (and less famous) classic works did not fit on one roll, you needed a whole pile. And if you wanted to make notes, you had to weigh the roll down on a table to keep the correct spot visible while you write. Or then you'd do like Pliny the Elder and employ an educated slave to read the books aloud for you while you make notes!
The codex, developed in Rome from a bound pile of wax tables, was a welcome novelty. It was the "pocket edition" of the day, a way of taking a whole book along on a trip. Leafing through is easier than in a roll and one can even stick something into the book as a bookmark and it'll stay there. It still took about 300 years for the codex really to become the norm, despite it's advantages. Codices were also made out of papyrus, so writing material alone doesn't explain the development - Casson thinks it was just a matter of habit, both that of readers and that of scribes; they were used to rolls.
However, there's an exception and this is of importance to us. Casson writes (page 129 in my edition):
The finds from Egypt demonstrate unequivocally that from the very beginning Christians used only the codex for their copies of the Bible and strongly favored it for their other religious writings.Casson's theory is that this is partly due to the fact that early Christianity had a stronghold in the city of Rome, the home of codex format. He also points out that the codex format is free from associations that the roll of papyrus (pagan) or parchment (Jewish), would have in the Roman context. Whatever the reasons, the adoption of codex as the format for Christian writings had far reaching consequences:
..., as Christianity expanded, so did its use of the codex, and when, from Constantine's reign on, Chrstianity was more or less the established religion, codices were to be seen in every church and school of the empire. This surely speeded up the overall displacement of the roll. (p. 130)...and so was laid ground to the earliest medieval (monastic) libraries with shelves of stacked codices, the library we know.
Lionel Casson: Libraries in the Ancient World. Yale Nota Bene, Yale University Press/New Haven and London. First published as a Yale Nota Bene book in 2002. Hardcover edition first published by Yale University Press in 2001.
ISBN of the paperback version: 0-300-09721-2
Price on back cover: USD 12.95 / Sterling 8.99; I think I paid the list price on it, so about 7-8 euros. Amazon UK has it for 6.80 pounds, but for people outside of the British Isles Book Depository's EUR 8.10 is a better deal as it already includes shipping anywhere in the world; I've found both shops to be reliable and quite fast (and this book should ship as a letter, it is so small and lightweight).