17), but there is much weight in Wordsworth’s remark, that “such suppositions as these have a tendency to destroy the credit of the ancient MSS.; and if such surmises were true, those MSS. would hardly be worth the pains of collating them” (_note on_ 1 Cor. xv. 49). VII. “Apparent probabilities of erroneous transcription, permutation of letters, itacism and so forth,” have been designated by Bp. Ellicott “_paradiplomatic_ evidence” (Preface to the Galatians; p. xvii, first edition), as distinguished from the “_diplomatic_” testimony of codices, versions, &c. This species of evidence, which can hardly be deemed internal, must have considerable influence in numerous cases, and will be used the most skilfully by such as have considerable practical acquaintance with the rough materials of criticism. We have anticipated what can be laid before inexperienced readers on this topic in the first chapter of our first volume, when discussing the sources of various readings(255): in fact, so far as canons of internal or of paradiplomatic evidence are at all trustworthy, they instruct us in the reverse process to that aimed at in Vol. I. Chap. I; the latter showing by what means the pure text of the inspired writings was brought into its present state of _partial_ corruption, the former promising us some guidance while we seek to retrace its once downward course back to the fountain-head of primeval truth(256). To what has been previously stated in regard to paradiplomatic testimony it may possibly be worth while to add Griesbach’s caution “lectiones RHYTHMI fallaciâ facillimè explicandae nullius sunt pretii” (N. T., Proleg. p. lxvi), a fact whereof 2 Cor. iii. 3 affords a memorable example. Here what once seemed the wholly unnatural reading ἐν πλαξὶ _καρδίαις_ σαρκίναις, being disparaged by dint of the rhyming termination, is received by Lachmann in the place of καρδίας, on the authority of Codd. AB (_sic_) CDEGLP, perhaps a majority of cursive copies (seven out of Scrivener’s twelve, and Wake 12 or Paul. 277); to which add Cod. א unknown to Lachmann, and that abject slave of manuscripts, the Harkleian Syriac. Codd. FK have καρδίας, with all the other versions. If we attempt to interpret καρδίαις, we must either render with Alford, in spite of the order of the Greek, “on fleshy tables, [your] hearts:” or with the Revisers of 1881 “in tables _that are_ hearts of flesh;” yet surely σαρκίναις as well as λιθίναις must agree with πλαξί. Dr. Hort in mere despair would almost reject the second πλαξί (Introd., Notes, p. 119).