By far the greater proportion of all “various readings”--perhaps nineteen out of twenty, are purely of a _verbal_ kind, and they are such as can claim the attention of none but philologists and grammarians: a few may deserve the notice of every reader of ancient literature; and a few demand the consideration of the student of history. But, taken in a mass, the light in which they should be regarded is that of their furnishing a significant and conclusive proof of the care, fidelity, and exactness with which the business of copying was ordinarily conducted. For it is certain that nothing less than a high degree, as well of technical correctness, as of professional integrity, on the part of those who practised this craft, could have conveyed the text of ancient authors through a period--in some instances--of two thousand years, with alterations so trivial as are those which, for the most part, are found actually to have taken place. When the discrepancies of manuscripts of an author are such as materially to affect the sense of a passage in itself important, so as to demand the exercise of discrimination on the part of the student of history, it becomes necessary to understand, and to bear in mind, what were probably the most common sources of such diversities. The following may be named as the most common causes of the various readings which are met with in comparing several copies of the same ancient author. 1. Nothing can be more probable than that authors who long survived the first publication of their works, should, from time to time, issue revised copies of them; and each of these altered copies would, if the work were in continual request, and were widely diffused, become the parent, as we may say, of _a family_ of copies. Thus it would be that, without any fault on the part of the transcribers, a considerable amount of such diversities would be originated, and perpetuated. A large proportion, perhaps, of those variations which occupy the diligence and acumen of editors and critics, and for the rectification of which so many learned conjectures are often hazarded, have, in fact, arisen from the author’s own hand in revising the copies which, at intervals, he delivered to his amanuenses. The perpetual opportunity afforded for introducing corrections, when a book was continually in request, would not fail to encourage, in fastidious authors, the habit of frequent revision: meantime transcribers, in distant countries, might have no opportunity to collate the earlier with the later exemplars. This source of various readings seems to have been too little adverted to by critics; though it might serve to solve some perplexing questions relative to the genuineness of particular expressions or sentences, which have fallen under suspicion from their non-existence in certain manuscripts.