Suppose a philosopher had a great deal of money to spend—though this is not in accordance with experience, it is not inherently impossible—and suppose he thought, as any philosopher does think, that the British public ought to read much more and better books than they do, and that founding public libraries was the way to induce them to do so, what sort of public libraries would he found? That, I submit, is a suitable topic for a disinterested speculator.

He would, I suppose, being a philosopher, begin by asking himself what a library essentially was, and he would probably come to the eccentric conclusion that it was essentially a collection of books. He would, in his unworldliness, entirely overlook the fact that it might be a job for a municipally influential builder, a costly but conspicuous monument to opulent generosity, a news-room, an employment bureau, or a meeting-place for the glowing young; he would never think for a moment of a library as a thing one might build, it would present itself to him with astonishing simplicity as a thing one would collect. Bricks ceased to be literature after Babylon.

His first proceeding would be, I suppose, to make a list of that collection. What books, he would say, have all my libraries to possess anyhow? And he would begin to jot down—with the assistance of a few friends, perhaps—this essential list.

He would, being a philosopher, insist on good editions, and he would also take great pains with the selection. It would not be a limited or an exclusive list—when in doubt he would include. He would disregard modern fiction very largely, because any book that has any success can always be bought for sixpence, and modern poetry, because, with an exception or so, it does not signify at all. He would set almost all the Greek and Roman literature in well-printed translations and with luminous introductions—and if there were no good translations he would give some good man #500 or so to make one—translations of all that is good in modern European literatures, and, last but largest portion of his list, editions of all that is worthy of our own. He would make a very careful list of thoroughly modern encyclopaedias, atlases, and volumes of information, and a particularly complete catalogue of all literature that is still copyright; and then—with perhaps a secretary or so—he would revise all his lists and mark against every book whether he would have two, five or ten or twenty copies, or whatever number of copies of it he thought proper in each library.

Then next, being a philosopher, he would decide that if he was going to buy a great number of libraries in this way, he was going to make an absolutely new sort of demand for these books, and that he was entitled to a special sort of supply.

He would not expect the machinery of retail book-selling to meet the needs of wholesale buying. So he would go either to wholesale booksellers, or directly to the various publishers of the books and editions he had chosen, and ask for reasonable special prices for the two thousand or seven thousand or fifty thousand of each book he required. And the publishers would, of course, give him very special prices, more especially in the case of the out-of-copyright books. He would probably find it best to buy whole editions in sheets and bind them himself in strong bindings. And he would emerge from these negotiations in possession of a number of complete libraries each of—how many books? Less than twenty thousand ought to do it, I think, though that is a matter for separate discussion, and that should cost him, buying in this wholesale way, under rather than over #2,000 a library.

And next he would bethink himself of the readers of these books. "These people," he would say, "do not know very much about books, which, indeed, is why I am giving them this library."

Accordingly, he would get a number of able and learned people to write him guides to his twenty thousand books, and, in fact, to the whole world of reading, a guide, for example, to the books on history in general, a special guide to books on English history, or French or German history, a guide to the books on geology, a guide to poetry and poetical criticisms, and so forth.

Some such books our philosopher would find already done—the "Bibliography of American History," of the American Libraries' Association, for example, and Mr. Nield's "Guide to Historical Fiction"—and what are not done he would commission good men to do for him. Suppose he had to commission forty such guides altogether and that they cost him on the average #500 each, for he would take care not to sweat their makers, then that would add another #20,000 to his expenditure. But if he was going to found 400 libraries, let us say, that would only be #50 a library—a very trivial addition to his expenditure.

The rarer books mentioned in these various guides would remind him, however, of the many even his ample limit of twenty thousand forced him to exclude, and he would, perhaps, consider the need of having two or three libraries each for the storage of a hundred thousand books or so not kept at the local libraries, but which could be sent to them at a day's notice at the request of any reader. And then, and only then, would he give his attention to the housing and staffing that this reality of books would demand.

Being a philosopher and no fool, he would draw a very clear, hard distinction between the reckless endowment of the building trade and the dissemination of books. He would distinguish, too, between a library and a news-room, and would find no great attraction in the prospect of supplying the national youth with free but thumby copies of the sixpenny magazines. He would consider that all that was needed for his library was, first, easily accessible fireproof shelving for his collection, with ample space for his additions, an efficient distributing office, a cloak-room, and so forth, and eight or nine not too large, well lit, well carpeted, well warmed and well ventilated rooms radiating from that office, in which the guides and so forth could be consulted, and where those who had no convenient, quiet room at home could read.

He would find that, by avoiding architectural vulgarities, a simple, well proportioned building satisfying all these requirements and containing housing for the librarian, assistant, custodian and staff could be built for between #4,000 and #5,000, excluding the cost of site, and his sites, which he would not choose for their conspicuousness, might average something under another #1,000.

He would try to make a bargain with the local people for their co-operation in his enterprise, though he would, as a philosopher, understand that where a public library is least wanted it is generally most needed. But in most cases he would succeed in stipulating for a certain standard of maintenance by the local authority. Since moderately prosperous illiterate men undervalue education and most town councillors are moderately illiterate men, he would do his best to keep the salary and appointment of the librarian out of such hands. He would stipulate for a salary of at least #400, in addition to housing, light and heat, and he would probably find it advisable to appoint a little committee of visitors who would have the power to examine qualifications, endorse the appointment, and recommend the dismissal of all his four hundred librarians. He would probably try to make the assistantship at #100 a year or thereabout a sort of local scholarship to be won by competition, and only the cleaner and caretaker's place would be left to the local politician. And, of course, our philosopher would stipulate that, apart from all other expenditure, a sum of at least #200 a year should be set aside for buying new books.

So our rich philosopher would secure at the minimum cost a number of efficiently equipped libraries throughout the country. Eight thousand pounds down and #900 a year is about as cheap as a public library can be. Below that level, it would be cheaper to have no public library. Above that level, a public library that is not efficient is either dishonestly or incapably organised or managed, or it is serving too large a district and needs duplication, or it is trying to do too much.