The Book on the Bookshelf

by Henry Petroski

Reviewed by John Wright
Printable ViewReturn to Contents


This is a wonderful collection of essays on the form and storage of books, covering a range of details much wider than the title suggests.

The first chapter discusses bookshelves. After a few cursory remarks about the unnoticability of bookcases, the author describes various means of removing books from them. Pulling one out by the sides of the spine is recommended. Pushing its neighbors forward and then grasping the sides of the spine is suggested as an emergency alternative. Pulling a book by the headband (the loop of fabric at the top of the spine) is verboten.

The author treats also of the difficulty of returning books once removed, and mentions the evident tendency of a book to expand during its time away from the shelf (it is legended that the width of a book can increase by as much as a quarter-inch, provided it is watered and well-fed). The realignment of the spines of books once one of them has been returned is touched on, but without finally resolving the issue. And there is mention of various contrivances and furniture for keeping books upright in their places.

The second chapter discusses the scrolls. They were commonly unrolled from left to right, and tied up with strings or straps, though more valuable scrolls might be fitted in cylinders to the same purpose as a modern slipcase. There were various means of storing them—lidded cylindrical boxes with tags for the scrolls to distinguish their subjects, titles or authors, and cases with pigeonholes. A scroll was usually held flat and in place by use of a stone or other weight, but there were also slotted lecterns (with the portion being read rising from one slot and disappearing into another).

The remaining nine chapters are called Chests, Cloisters, and Carrels; Chained to the Desk; The Press of Books; Studying Studies; Up Against the Wall; Books and Bookshops; Bookstack Engineering; Shelves That Move; The Care of Books. And since the concerns of these chapters overlap, they will be described summarily.

From the use of the stylus and various kinds of impressionable tablets, arose the codex, or bound book, and the use of scrolls declined from the beginning of the fourth century. The binding of codices was commonly of wood, ¼ to ½ inch thick, with beveled, cambered or fluted fore-edges. The covers were often fastened with clasps. Codices were sometimes stored on shelves that slanted like a lectern.

Chests were commonly used to transport or store small collections and were not infrequently locked, since books, being handwritten and laborious to produce, were valuable. As more books accumulated, armaria1, armoires for books, were used in their place.

Places for secluded reading were valuable as well. In monstaries, novice monks had no privileges in this respect; those of higher grades could use a cloister; those of highest seniority enjoyed the luxury of a carrel.

In time, the increasing number of books with the continued necessity of securing them led to the chaining of books to desks or lecterns. And this led to the design of libraries in such a manner as to provide adequate light from the windows for reading books that, by their nature, could not be removed to better light. The practice of chaining books began to be disused in the 18th century.

It is worth mentioning here that Petroski’s book is supplied with many historical illustrations such as those depicting various lectern seats. One picture shows the dark, intricately carved and pew-like seats with their straight backs, and books laid out side by side on the slanted reading surface, the chains hanging down like garlands between them.

The reading lectern endured a slow transformation into the modern bookcase, according to the multiplication of books and the need for more economical storage and ease of access. The lectern was consequently provided with an abundance of shelves subdivided by uprights and very similar in appearance to the modern bookcase except that there hung, nearly on a level with the bottom of the lowermost shelf, a desk on which to rest the book being consulted. The books were chained each as before from the fore-edge and shelved with their spines inward; at the bottom of each shelf ran a rod to which the other end of the chain was attached.

This arrangement, called a press, was without the enclosing doors of the armarium (although book-cabinets with doors are also sometimes referred to as presses; in Wilde’s novel, Dorian Gray secretes evidence of his crime inside a press). The desk was at a height to be used while standing, although stools were sometimes provided. At the end of the press was posted a list, framed and sometimes enclosed by a door, of all the books it contained. The whole formed an innovation of the late 16th century. It was also an innovation of the 16th c. to print the title and author’s name on the spine, although it was not at first practiced universally; and exceptions to the rule had identifying slips of paper placed at the fore-edge, so that some books were shelved with their spines inward and others with their spines out.

The next chapter is devoted largely to renaissance studies and the various ways books were arranged in scholars’ apartments. It is accompanied by numerous woodcuts and engravings illustrative of developments during that period, although the subjects of the illustrations are commonly more ancient; e.g. Durer’s engraving of St. Jerome.

The author calls attention to several points of special interest. Mention is made of Agostino Ramelli2, a 16th c. engineer who wrote Diverse and Ingenious Machines (1588) and whose inventions include a “reading wheel” which rotated books into position before the reader’s desk.

There is a discussion of fore-edge painting and design. I find it written in an article elsewhere that fore-edge painting was developed in the 17th century but little practiced until the 18th. Nevertheless, Petroski presents examples of painting going back to the 15th, such as a German book of that century decorated with simple outlines of flowers on the fore-edges as well as the top and bottom edges. Other books’ fore-edges bore merely their titles. And others artfully depicted scenes appropriate to their contents. But the first purpose of these embellishments was identification.

(It is worth mentioning that the spine of a book has traditionally been considered best concealed because plainer and more functional than the cover and back which could be highly ornamented. It was only with the printed book that it was found expedient to exploit the blank area of the spine as a place for identification, which in turn dictated the shelving of spines outward, and which in turn encouraged printers to make the best of a bad situation by gilding them.)

As with so many other modernities, the tooled leather cover and gilt-lettered spine date to the 16th c. To the same period dates the first dos á dos books: those printed back to back, in nearly the same manner as the Ace paperbacks in which Dean Koontz’ early novels appeared, except that the early dos á dos books, printed in Germany, had a third board dividing the two works and forming a shared back cover.

The author contasts the English and continental modes of library arrangement. In the former, the cases are arranged in rows and the windows are narrow and closely spaced around the perimeter in order to distribute light to the stacks. But in the latter, the arrangement is very different. The cases here are placed along the walls, the interior of their circle being left open, and the windows are few and large. In order to make the most of space by this method, it is helpful to provide a second tier of cases and this is often done. A notable example can be seen in pictures of the library at Salamanca University.

In a discussion of booksellers, it is worthwhile to note that bookshops existed as early as the 15th century; one is depicted in a 1499 Danse Macabre, which depicts both printer and dealer as getting the worst of it from a pack of skeletons. It is also noted that, at least in the latter part of the 17th c., books were not sold bound at all but as loose sheets with the pages marked for appropriate folding; it was the business of the book-buyer to have the sheets bound as he liked. The traditional names for sizes of books derive their names from the folding that is still a part of printing: folio (each sheet folded to make two leaves, and forming a large book as a consequence), quarto (four leaves), octavo (eight), duodecimo (twelve).

The use of these terms today is perhaps less practical and absolute than affectionate and indicative; different sizes of sheets are used, producing similarly sized volumes with differing numbers of folds. But the usual hardbook back is octavo; a coffee-table book, a quarto; a very large book (possibly the Book of Birds), a folio; and I have a gold-colored volume of Poe called Weird Tales which is about the size of my hand and should probably be called duodecimo.

Toward the close of the book, various experiments and innovations are described—libraries with glass floors to collect and transmit light, glass ceilings for similar reasons, the domed and radially disposed reading room of the British Museum, slotted metal shelves to improve air flow (at the Library of Congress), the increasing introduction of electric light as less damaging to books than sunlight. There is a chapter on bookcases that roll, swing and revolve; a commentary on the problems of dust and dirt; and finally an appendix, which discusses the following rationales for ordering books in a library: author, title, subject, size, color, binding, publisher, familiarity, date (of publication or acquisition), number of pages, Dewey decimal system, library of congress system, ISBN, price, provenance, etc.

The whole is provided, as every good book about books should be, with endnotes, bibliography, illustration notes and index.


1 “One of the oldest representations of an armarium or bookcase, from the tomb of Galla Placidia in Ravenna.”: http://ccat.sas.upenn.edu/jod/texts/class.pictures.html

2 http://www.sil.si.edu/ondisplay/ramelli/