Pliny's encyclopaedia demands a very particular approach from its readers. The key paradox of the genre is that, having gone to the immense effort of gathering together information covering all the branches of human knowledge, it expects its readers not to read it all. The tension between the unifying discourse of complete knowledge and its practical segmentation into digestible chunks requires a certain amount of conceptual shimmying on the part of the reader. How relevant is the surrounding information to the single fact retrieved? In the case of Pliny, the possibility of understanding disembodied sections and facts in isolation is a promise held out to the reader by the book itself: Pliny explicitly gives his blessing in the preface and provides a summarium to aid in the process. This summarium takes up all of Book One of the Natural History, listing the contents of other 36 books, together with the sources Pliny consulted. But although the paratext promotes one model of how to read the Natural History, the text itself is ambivalent; the insistency of linear narrative never quite surrenders to the allure of complete segmentation. There is a logic to the Natural History that only reveals itself to the reader who follows the stream of information from fact to fact, section to section, book to book, subject to subject.