From The Outline of History by H.G. Wells

Volume I, p 367

In those days, it must be remembered, books were not in pages, but rolled like the music-rolls of the modern piano-player, and in order to refer to any particular passage, a reader had to roll back or roll forward very tediously, a process which wore out books and readers together.

One thinks at once of a simple and obvious little machine by which such a roll could have been wound to and fro for reference, but nothing of the sort seems to have been used. Every time a roll was read it was handled by two perspiring hands...

It is curious to note how slowly the mechanism of intellectual life improves. Contrast the ordinary library facilities of a middle-class English home, such as the present writer is now working in, with the inconveniences and deficiencies of the
equipment of an Alexandrian writer, and one realizes the enormous waste of time, physical exertion, and attention that went on through all the centuries during which that library fluorished.

Before the present writer lie half a dozen books, and there are good indices to three of them. He can pick up any one of these six books, refer quickly to a statement, verify a quotation, and go on writing. Contrast with that the tedious unfolding of a rolled manuscript. Close at hand are two encyclopaedias, a dictionary, an atlas of the world, a biographical dictionary, and other books of reference. They have no marginal indices, it is true; but that perhaps is asking for too much at present. There were no such resources in the world in 300 B.C. Alexandria had still to produce the first grammar and the first dictionary.

This present book is being written in manuscript; it is then taken by a typist and typewritten very accurately. It can then, with the utmost convenience, be read over, corrected amply, rearranged freely, retyped, and recorrected. The Alexandrian author had to dictate or recopy every word he wrote. Before he could turn back to what he had written previously, he had to dry his last words by waving them in the air or pouring sand over them; he had not even blotting paper.

Whatever an author wrote had to be recopied again and again before it could reach any considerable series of readers, and every copyist introduced some new error. Whenever a need for maps or diagrams arose, there were fresh difficulties. Such a science as anatomy, for example, depending as it does upon accurate drawing, must have been enormously hampered by the natural limitations of the copyist. The transmission of geographical fact again must have been almost incredibly tedious.

No doubt a day will come when a private library and writing-desk of the year A.D. 1919 will seem quaintly clumsy and difficult; but, measured by the standards of Alexandria, they are astonishingly quick, efficient, and economical of nervous and metal energy.