Every one who has had any thing to do with publication, is well aware of the great difference between the space occupied by the written and that filled by the printed letters. It is well known that the volumes of ancient libraries consisted of rolls, which generally were written only on one side. Thus the written surface of one of these volumes would correspond to but half the written surface of one of our books, of which every page is covered with letters. A library, then, composed of 100,000 rolls, would contain no more matter than one of our libraries composed of 50,000 manuscripts. It is well known, also, that a work was divided into as many rolls as the books which it contained. Thus the Natural History of Pliny, which in the _Princeps_ edition of Venice forms but one folio volume, would, since it is divided into thirty-seven books, have formed thirty-seven rolls or volumes. If it were possible to compare elements of so different a nature, we should say that these rolls might be compared to the sheets of our newspapers, or to the numbers of our weekly serials. What would become of the great library of Paris were we to suppose its 824,000 volumes in folio, quarto, &c., to be but so many numbers of five or six sheets each? Yet this is the rule by which we ought to estimate the literary wealth of the great libraries of ancient times; and "hence," says M. Balbi, "notwithstanding the imposing array of authorities which can be brought against us, we must persist in believing that no library of antiquity, or of the middle ages, can be considered as equivalent to a modern one of 100,000 or 110,000 volumes."