For Meric Casaubon (1599–1671), the study of spirits, including the soul, devils and angels, was properly a part of physics, a discipline treating natural entities. This formed part of the arts curriculum, whose purpose was the inculcation of the general learning that Casaubon, an early critic of the Royal Society of London, applauded. To conduct his own investigations into the activity of spirits, Casaubon drew on what he called ‘converse with books’. This exchange took place in part within his own library, but to a greater extent through reading carried out in those of others, notably that of Sir Thomas Cotton in London, where Casaubon gained access to the manuscripts of the Elizabethan magus John Dee.

During the century or so after 1640, the use of collections of books and manuscripts was unquestionably part of the work of scholars such as Casaubon, active in the disciplines of natural philosophy. But those disciplines embraced a great deal that would currently surprise practitioners of ‘science and medicine’. Debates about the content and form of natural philosophy often pitted knowledge that could be obtained from experiment and practice against wisdom derived from books.

Francis Bacon (1561–1626), the philosopher on whose ideas many of the proponents of intellectual reform based their programmes, did not make explicit mention of libraries in his scientific utopia, The new Atlantis (1626), although the descriptions of several of the categories of employees with whom Bacon peopled ‘Salomon’s House’ suggested some of the uses to which he believed libraries could be put. ‘The Merchants of Light’ thus ‘bring us the books, and abstracts, and patterns of experiment of all other parts’; the ‘Depredators’ also ‘collect the experiments which are in all books’; finally, the ‘Compilers’ were supposed to ‘draw the experiments … into titles and tables’.