On some special difficulties in Pindar. 221 by an eminent scholar, who was then engaged in examining for the Classical Tripos at Cambridge, as to the amount of discredit which ought to attach to a candidate for high honours who had construed naL nep with a finite verb in his Greek composition. I expressed my opinion that it was a blunder of the gravest character, and ought to be visited accordingly. In this opinion he concurred. But the subject was discussed a day or two afterwards in the company of other scholars, the passage from Plato was adduced, and the doubt was raised, whether there might not be many exceptions to a rule which appeared to us to be imperatively required by the genius of the Greek language. As the assertion of a negative is proverbially dangerous both in law and logic, I thought it best to make no remarks on the passage in Pindar until I had ascertained that there were no other examples of the construction in question. And having now sought in vain, for many years, to find any fourth violation of this idiom, I can have no hesitation in pronouncing that Kai nep with a finite verb is utterly inadmissible. I am no lover of Procrustean canons, but the general analogy of a language, fortified by a thousand examples, must override three exceptional cases, in which there is so plain and simple a road to the neces- sary correction. V. Although I did not, in editing Pindar, venture to remove from the text the faulty construction of Kai irep with a finite verb to which I have just now directed attention, I did not hesitate to correct the converse error, namely, the appearance of a parti- ciple with a conditional particle in 01. 11. 56, where for tl de piv ex>v tis oldev to /xeXXov, I read ev <' p,iv ex cov Tls ^ V TO p*)&ov, taking ev with otda. I said in the preface (p. xi.) that the emphasis on cv justifies its position, and I have since then fallen in with the following passage of Plato (Eesp. vi. p. 492 e.), where we have a similar prominence of the same adverb : eS yap x pv dbimu. With regard to the general question whether a participle can be used for the finite verb in conditional sentences, the alleged examples are so few and the necessity for such a construction is so inca- pable of proof, that I should not hesitate to adopt the simplest mode of getting rid of the difficulty. In addition to those, which
