In many, or in most of these particulars ancient and modern usages are very dissimilar. But something more should first be said indicative of the purpose with a view to which these facts are brought forward. It need scarcely be said that the antiquity and integrity of a book can be open to no question, if in any case the existence of any one copy of it can be traced back, with certainty, to the time of its first publication. If, for example, a manuscript of a work in the author’s handwriting were still extant, and if the fact of its being such could by any means be proved, our argument would be concluded, and any other evidence must be deemed superfluous. There are however few such unquestionable _autographs_ to be found, even of modern works, and none, of any ancient one. Yet the circumstances attending the preservation and transmission of manuscripts are, in some instances, as we shall see, such as to prove the antiquity and genuineness of a work with little less certainty than as if the very first copy of it were in existence. But before we enter into the particulars of this proof it should be mentioned--especially as we intend to follow the order of time _retrogressively_, that the history of _manuscripts_ need not be traced through any later period than that of the early part of the fifteenth century, when most of the classic authors passed through the press. For the invention of printing has served, as well to ascertain, beyond doubt, the existence of books at certain dates, as to secure the text from extensive interpolation and corruption. A printed book is not susceptible of subsequent interpolation or alteration by the _pen_: it bears also a date, and the issuing of different editions of the same work from distant places, would render any falsification of date in one of them, or any material corruption of the text by an editor, a nugatory attempt. For example, there are now extant, printed copies of the history of the Peloponnesian war, dated “Venice, 1502;” other copies of an edition of the same work are dated “Florence, 1506;” others are dated “Basil, 1540;” and others, printed within a few years of the same time at Paris and Vienna. On being compared with each other, these editions are found to agree _in the main_; and yet to disagree in many small variations of orthography, syntax, or expression; so as to prove that they were derived independently from different manuscripts; and not successively from each other. These printed editions, therefore, sufficiently prove the existence of the work in the fifteenth century; and also that the text of the modern editions has not been materially impaired or corrupted during the last four hundred years.