mits of being demonstrated; by the help of the date which Josephus assigns to the death of Vitellius-a certain day in the month Apellæus, or rather, as we shall see by and by, Audenæus. The same proof also will make it appear that the Jewish months, beginning with Nisan, are much more probably to be reckoned in every year, (an unintercalated one, as well as an intercalated,) at twenty-nine and thirty days each alternately, than vice versa, at thirty days and twenty-nine.

From the great minuteness with which Josephus has specified the lengths of the several reigns between the demise of Nero and the death of Vitellius, it is clear that he intended to be very exact in each of these instances; and to express their lengths not merely by months, but also by days. The integrity of his text, however, has suffered greatly throughout from corruptions as to numbers: and no where more so than here.

The rule which he follows, in stating the lengths of the reigns in question, is to reckon by calendar Julian months, and by those only: as may thus be demonstrated.

с

I. The reign of Galba is stated at seven months and as many days. The last day of the reign of Galba was XVIII. kal. Feb. (Jan.15.) This computation, then, supposes it to have begun on the ninth of some month. Reckon backwards seven Julian months, and its beginning will coincide with June 9. The correctness of this conclusion is proved by what is asserted concerning the length of the reign of Nero.

Nero is said to have reigned thirteen years and eight days; in which assertion, as the text now

a Bell. iv. xi. 4. Epitome, in eodem.

b Bell. iv. ix. 2. Cf. Aurelius Victor, in Galba. Also the c Tacitus, Historiæ, i. 27. d Bell. iv. ix. 2.

stands, there is undoubtedly a great omission. The reign of Nero began October 13. U. C. 807. and he reigned full thirteen years, and almost eight months of a fourteenth. Reckoned by Julian months, the eighth month of this fourteenth year would begin May 13. U. C. 821. From that day inclusive, to the 9th of June inclusive also, are twenty-eight days exactly. I have no doubt that this is what Josephus meant; viz. that the reign of Nero expired, as the reign of Galba began, June 9, U. C. 821. His text then must have stood originally, τρισκαίδεκα ἔτη, μῆνας ἑπτὰ, καὶ ἡμέρας ὀκτὼ καὶ εἴκοσι—η καὶ κ' ; not simply ή : a mode of notation which might easily be corrupted, especially if expressed by nk', or even by kŋ'.

Again, the reign of Otho is computed at three months and two days. According to Dio, Otho died on April 17, eleven days before his birthday, April 28. To this day inclusive, from January 15. preceding it, reckoned as before, there are just three months and two days. It is clear, then, that Josephus reckons the day of his death as April 17. Hence it constitutes no objection that Suetonius supposes Otho to have died xcv imperii diek; which, dated from January 15 inclusive, would make his death to have happened April 19. That statement also might once have stood in Suetonius XCIV imperii die; which, reckoned from January 15, as before, inclusive, would be only one day wide of the truth. Josephus is further confirmed by Dio, loco citato, who computes the reign of Otho in general terms at ninety days *.

It is another example of Josephus' mode of reckoning by calendar or Julian months, that Ant. Jud. xx. viii. 1. the reign

of Claudius is stated at thirteen years, eight months, and twenty days. This is exactly the interval of time between

k Otho, II.

e Tacitus, Annales, xii. 69. f Dio, lxiii. 29. Suetonius, Nero, 40, 1. g Bell. h Dio, lxiv. 15. i Suetonius, Otho, 2.

iv. ix. 9.

On this principle, when Vitellius is said to have reigned eight months and five days', we must understand his eighth month to have expired December 16, U. C. 822, and his death to have happened December 21, on the fifth day afterwards. Accordingly Dio also states him to have reigned a year, save ten daysTM; which the necessity of the case requires should be understood to imply that he died on December 21. He asserts also that he survived his birthday eighty-nine days m. His birthday was either September 7, or September 24". From the former inclusive, the eightyninth day would fall on December 4-but from the latter, on December 21: and we have the further assurance of Tacitus, that Vitellius was alive considerably after December 4. All these circumstances must concur to fix the day of his death to December 21.*

January 24, U.C. 794, inclusive, and October 13, U. C. 807, inclusive, so computed.

* Dio, lxvi. 17, from the death of Nero, U. C. 821, to July 1, U. C. 822, the first of Vespasian, it is reckoned one year, twenty-two days: which proves that Dio considered June 9 to be the date of Nero's death.

Hippolytus, Opera, 58. Chronicon, section xix. Nero's reign is stated at 13 years, 8 months, 28 days. So also by Clemens Alex. i. 406, 1. 16: Stromatum i. 21. Theophilus ad Autolycum, iii. 27. p. 387, puts it at 13 years, 6 months, 28 days: Epiphanius, ii. 168 C: De Mensuris et Ponderibus, xii. at 13 years, 7 months, 27 days. But Cassiodorus, in Chronico, computes it at 13 years, 7

months, 28 days, exactly. The same authority reckons the reign of Galba at 7 months: that of Otho at 3 months 5 days: that of Vitellius at 8 months 1 day. The intermediate date agrees with that of Suetonius: and as reckoned from April 19 exclusive, Vitellius' reign would thus expire on December 20.

Tacitus, Historiæ, i. 18: Piso was adopted by Galba, January 10, U.C. 822. Ibid. 19. 27, he was killed four days after (which must be reckoned exclusively,) January 15, xviii Kal. February: for, ibid. 29, (cf. Plutarch, Galba, 24,) it is said, in reference to the date of the adoption itself, on the very day of the death, Sextus dies agitur, &c.

The arrival of Icelus, Plutarch, Galba, 7, in Spain, with the news

1 Bell. iv. xi. 4. m lxv. 22. n Suetonius, Vitellius, 3. o Historiæ, iii. 67.

Now this day Josephus makes to coincide with the third of Apellæus. Apellæus or Chasleu was the ninth month in the Jewish year, reckoned from Xanthicus or Nisan: and Tisri or Hyperberetæus, was the seventh. If the 15th of Nisan coincided with March 26, or (what is the same thing) the 15th of Tisri coincided with September 19; the 24th of September coincided with the 20th of Tisri, and the 3d of Apellæus with the 5th of November. But if Apellæus be reckoned at twenty-nine days, and the 3d of Apellæus coincided with the 5th of November: then the 29th of Apellæus coincided with the 1st of December: and the 3d of Audenæus with the 4th of December: and the 20th of Audenæus with the 21st of December. Now either

of Nero's death, seven days after that event, would be June 16; what Plutarch calls θέρος ἤδη.

Plutarch, Otho, 18, also dates his reign at three months. Tacitus, Historiæ, ii. 55. the Ludi Cereales were going on at Rome when news arrived of the event of his death. The old Roman calendar dates these games April 10 or 12 and they were celebrated for six or seven days, as late as April 19. Cf. Ovid, Fasti, iv. 389, 393, 619-621, 681. Cicero, Ad Atticum, ii. 12.

If Otho died at Brixellum on the 17th of April, his death might easily be known at Rome on the nineteenth: and this is the most probable account of the mistake of Suetonius; that he has confounded the day on which the news of the death was received, with the day of the death itself.

Tacitus, Historiæ, iii. 67.begins to relate the circumstances of the death of Vitellius, from xv. Kal. Jan. Dec. 18; supplying the fol

lowing notes of time, after (cap. 69) the night of that day.

Cap. 70, Luce prima, December 19: 78, the Saturnalia, which began December 17, and lasted to December 21 at least: 79, the night of December 19, and the morning of December 20: 82, ad serum usque diem; which is the end of December 20: 85, 86, diem latebra... præcipiti in occasum die: which seems to be spoken still of December 20, and certainly is to be understood either of that day or of the next. One of these two, then, according to Tacitus, December 20 or 21, was the day of Vitellius' death.

Eutropius also, lib. vii. 18, supposes Vitellius to have reigned eight months and one day. If he reckons these from the death of Otho, ninety-five days as he supposes after that of Galba, the first month began April 20, and the last expired December 19, and Vitellius died on the 20th.

of these dates per se might express the day of the death of Vitellius; but the latter only can be the true, or that which agrees both with Josephus' own statement respecting the length of his reign, and also with the testimony of Dio and Tacitus. What remains, then, except to suppose that instead of Tρity unvòs 'ATεaíoυ, Josephus either wrote, or intended to write, εἰκάδι μηνὸς Αὐδηναίου ? κ' μηνός Αὐδηναίου, not γ' μηνὸς Απελλαίου ? How easily κ' might be mistaken for γ is too obvious to require proof *.

But this supposes that the month Apellæus consisted of twenty-nine days; for had it consisted of thirty, the 3d of Audenæus, the next month, must have fallen on December 5, and the 20th of Audenæus on December 22; and this would place the death of Vitellius one day too late. If then we would avoid such an error at last, or in the final result of a computation, which Josephus, as it is clear, intended should be precise and exact; we have no alternative except to suppose that Apellæus consisted of twenty-nine days, not of thirty; and on the same principle Tisri also.

I am aware that the contrary is the commonly received opinion; and therefore that it may appear presumptuous in me to venture to dissent from it. Nor should I have thought of dissenting from the

*Such a mistake as the above, with reference to the name of a month, would not be unexampled in Josephus; for, Ant. Jud. xi. v. 4: Apellæus is confounded with Tebeth, though both the Latin version in loco, and Ant. Jud. xii. v. 4, prove it to have been in reality the same with Chasleu. Josephus might confound Apellæus with Aude

næus, in this instance, intending to express by either a certain date in the Jewish month Chasleu, because παρὰ Μακεδόσιν, according to Suidas, in voce, Apellæus was reckoned the same with December, and so was Chasleu among the Jews. Audenæus, on the other hand, according to Suidas, in voce, was the Macedonian January.