On some special difficulties in Pindar. 215 wood has them, &c." To this method of dealing with the pas- sage before us I have many objections. In the first place, it does not seem very natural that the active naTeKkaae, predicated of the charioteer, should be so directly opposed to the passive Kpefiarai, predicated of the chariot. Then, there is no statement of the place where the chariot was hung up, until we get into a totally different sentence, beginning with to acpe. Then again, I do not think that, as the re/troW oaioaXa clearly refer to the 6os dlcppos of the antistrophe, suspension in the temple would be the proper or usual mode of dealing with such an offering. Lastly, I do not agree with Bockh that, " onoo-a dictum est ad omissum et cogitatione supplendum iravra ; " or with Dissen, that the empha- tic roa-cra " nescio quomodo ineptum est et non Pindaricum." On the contrary, it appears to me, that as the relative clause, which precedes, contains a description or statement of the objects con- secrated, the antecedent ought to be expressed in the following or correlative clause, which tells us in what part of the temple at Delphi the chariot was deposited. If the verb Kpeparai had not made its appearance in the text, no one I conceive would have objected for a moment to the construction 67760-0, ayav apei^ev, rdW' t^et H&aOpop. For these reasons, I think that the old reading toW is better than the to 0-$', which has been reimported from the MSS., and that the genuine text is to be sought in some restitution which will make Carrhotus the subject of all the verbs, and pre- serve the unbroken tenor of the passage. The primary corrup- tion is the verb Kpeparai, which is inappropriate in itself, and wants a local predicate to help out its meaning. If we would discover the diplomatic or palaeographic origin of this corruption, we must apply a principle of great importance in verbal criticism, especially in the case of ancient vitiations of the text. This principle, which I have applied to the correction of one of the most extensive corruptions in the text of Sophocles (Antig. 607, p. 186), is thus defined : resemblances between the terminations of successive lines produce interpolations or absorptions of sylla- bles. Thus we have here r[e/ii]/os Kp[e/x]arai Xjepuipav aycov. As the terminations -at and -c<ov are often confused, I see in this
