INTRODUCTION
sequence of thought; the chapters in A are marked by rubricated capitals, but the resultant divisions are frequently incoherent. Yet it is in the actual text that the difference is most remarkable. If we read P, we meet many small errors, such as are common in all manuscripts, but the general impression left is of a text with many idiosyncrasies but an intelligible text; if, on the other hand, we take up A at any point, not only do we find continual omissions of lines, parts of lines, even of longer passages of some two or three lines, but the amount of corruption of individual words is such that it is possible to make only an approach to the meaning of the author, sometimes not even that. Besides this, especially in the later Books, we meet forms of words which are corrupted according to no known rules of manuscript interpretation. The problem of the origin of all these difficulties is intensified by the fact that the hand of the scribe is quite a good one, although late, that he has often patched up a lapsus calami, and has occasionally written a correction of a form in the space above the line, without erasing his first attempt. He appears to have tried to be intelligent.
Clearly the principal problem of an editor is to determine what weight is to be attached to the evidence of P and A respectively in a case of difference between them, and what are the grounds for his decision. In regard to P there is one question to be answered first, viz. how widely does the printed text of Gesner, which Xylander edited, differ from the lost original P cod.? There are many obvious misprints in P, which Xylander remedied in the second or Basel edition; there are many other mistakes in the form of words, which Xylander did not correct, some of which certainly appear to be misprints, though others may have been errors of the scribe. Schenkl's estimate is that Xylander corrected thirty-six mistakes but overlooked forty-four, generally graver, errors, which he had silently amended in his original Latin version. It is certain that
