MERGE TO GYRENE. 555 do not see what other sense could be given to the words in question, with so much probability of being that which the writer intended ; and it is plain, that as the sentence now stands it is incomplete A\e could not succeed in finding any Greek dates of antiquity, although the Greek names are very numerous ; but a person accustomed to the many negligent modes of writing the character, with plenty of time and light at his disposal, might probably succeed in finding Greek inscriptions of more interest than we were able to discover in the mass of waiting here alluded to ; a great portion of which, as might naturally be expected, consists of rude scrawls and hasty scratches — mere apologies in fact for letters almost of any kind. That the fountain continued to be an object of curiosity, and probably of religious veneration, after the cession of the country to the Eomans, may, however, be inferred from what we have stated; and a minimum may at least be established with respect to the date of the excavated channel, if w'e cannot ascertain the precise time of its formation, or wdiether it was cut at one or at several periods.

We have already mentioned that several hours had elapsed, from the time of our entering the channel to that of our re-appearance at its mouth ; and we really believe that the Arabs of the place, w'ho had * We imagine the words to have been, em ieeos- rou fj.syiarov AtsoKKoivos, but the os is wanting after AttoXXoiv, and the ix. in fAiyisrov ; in which latter word also the s and y look more like an a and a t. The rest of the inscription is clear ; and were we only to give it as a fragment, mi iseos rov...i(srou AKoXkcov. . , there would, probably, be no doubt raised with respect to the manner of reading it.

4 B 2