328 Journal of Philology. (2) " This is proved by Luke iii. 1 : In the 15th year of Tiberius" the year beginning with the fixt or moveable 1 Thoth = 29 or 16 Aug. a.d. 28, M. Nab. 776 "the word of the Lord came to John the Baptist, and iii. 21, 23, our Lord, coming to his baptism was about thirty years old. Moreover, he preached for the space of one year, as foretold by Isaiah lxi. 1. (cf. Luke iv. 19) : TTwreKaibcKciTco ovv erei Ttfieplov nal irevrcKaideicaTCd Avyovarov [Cle- ment doubtless wrote te ovv Zttj T., Kal if' err) Auy.] ovto> 7rr)povTai ra. rpiaKovra itt) ea>s ov enadev." (3) " From the Passion (i(j> ov 8c cnaOcv) to the destruction of Jerusalem are 4:2 y. 3m." (4) "And thence to the death of Commodus 128y. 10m. and 3d." (5) " The sum, therefore, from the Nativity to the death of Commodus is 194y. lm. 13d." In (4) for pier)' we must read pica, 121, which is the number of years required, and also expressed in 140 for the interval from the destruction of Jerusalem to the death of Commodus. Also in ijpepcu y we must restore the numeral iota absorbed by the final iota of the noun: jpepauy. So the sum of 30y., 42y. 3m., and 121y. 10m. 13d. becomes 194y. lm. 13d., as expressed in (.">). In (2) and (3) there is this manifest inconsistency, that the Baptism and the Passion are assigned to the same year, = 15 Tiberius, and the " one year" of our Lord's ministry is not in- cluded in the summation an error which, as Mr Clinton justly remarks, F. R. s. a. 29, " is not palliated by the explanation of Pagi adv. Baron, t. i. p. 19." Mr Clinton suggests the correction, err] pa (for pft) prjvts y. For reasons which will appear in the sequel, I do not adopt this alteration. In the nature of the case, it is probable that the form of year employed by Clement in this enumeration is the vague year of the Astronomical Canon or Mm of Nabonassar, that being the instrument commonly used in his age and country in the treat- ment of astronomical questions, and the question relative to the date of our Lord's Passion being such*. And in fact Clement sets out with the canon, 144.
- In this form of year 'the cycle of period for the Julian year, nann-ly,
Apis' (=25 vague years) was a highly 25 x 365 = 9125 days exceeding only by convenient instrument for the calcula- 0*048 d. the sum of 309 mean lunations tion of lunar dates, more so, in fact, (= 9i24 # 95i9d.) than the Metonic cycle and Callippic
