susceptibility to being read in private) reinforces its thematics of solitude. Yet the result is a catch-22: when such analyses of the material conditions of production and consumption corroborate some formal or thematic analysis of the text itself, they become redundant; and in those rarer cases where they contradict conclusions reached through textual explication, they become irrelevant. Heads you win, tails I lose. It may prove more productive to turn our attention to moments where these two strands of evidence pull apart, when (for example) an anticolonial manifesto is printed on paper imported from themetropole, or when an oath of revenge is sworn upon the same bible whose text preaches forgiveness. Book historians are hardly alone in facing this challenge: the evidence that other schools of literary historicism invoke is equally incommensurate with the language in which they frame their claims. William St Clair is right to charge our entire profession with basing historical claims on literary evidence—more specifically, with supporting arguments about early nineteenth-century culture with texts that were produced in that era but found an audience only later (St Clair). By extension, historicist literary critics could be accused of wanting to have their cake and eat it—to make cultural claims about a different time and place while restricting ourselves to evidence that gives a particular set of professional readers aesthetic pleasure in the here and now. I plead guilty to a different bait and switch: I draw my evidence for bibliographical claims from sources chosen at least in part on the basis of their linguistic content. If I had taken my own premises more seriously, the table of contents would have looked rather different. Unlike Dickens, neither Trollope nor Charlotte Brontë employed particularly innovative publishing strategies. More fundamentally, if publishing strategies were my primary focus, authors would not provide the marquee names for my table of contents in the first place. Much less Victorian authors, given that the books most used by Victorians were not books written by Victorians. In the case of books whose life span exceeded that of human beings (notably the Bible) they were not even books printed by Victorians. Worse, this study of those uses of the book that exceed or even replace reading is based primarily on the evidence of my own . . . reading. My occasional appeal to material evidence (the traces of wear and tear, of handling or ignoring) is dwarfed by my more regular use of textual proof. Certainly my interpretations are conditioned by the list price and current condition of the books that I study, but they appeal more heavily (to state the obvious once again) to the evidence of words. I confess not only to the absence of a volume of sources or a systematicity of sampling that would satisfy historians—as for most literary critics’, my sources are printed (or reprinted or digitized), not manuscript; skewed toward the literary canon; too few in number to support any bibliometric claim—but also to the presence of literary-critical tastes and priorities. My subject is Victorian representations and perceptions of, and fantasies and illusions about, the circulation of books, not the circulation of books itself. And even within representations, my corpus skews heavily toward the literary and journalistic: a search across a few digital repositories for the character string “locked bookcase” yields one kind of result; a search of locksmiths’ records to find out how often bookcases tampered with by servants needed to be repaired would have yielded quite another. To make book-historical claims on the basis of textual evidence is not the same as making cultural-historical claims on the basis of literary evidence: the book’s material form can give aesthetic pleasure as easily as its textual content can withhold it. It’s a fallacy to assume that analyses of noncanonical or nonliterary