The stage is now set to examine the calendar conditions in Jerusalem when Jesus preached and healed.

The adjustment of the lunar calendar to the solar year was becoming more accurate. The beginning of the year in Nisan could be adjusted to fall closer to the Vernal Equinox, which marked the beginning of spring. This was done by the intercalation, or addition, of a thirteenth month, or a second Adar. By the first century CE it was well known that the solar year was 365¼ days, as is evident in the Roman use of the Julian calendar. By this time the Jews were also able to calculate the date of the Vernal Equinox,1 and the addition of a second Adar could be known in advance. Philo (De Spec. Legg. 2:8; De Opificio Mundi 39) and Josephus (Ant. III 10:5) reported Nisan occurring when the sun was in Aries, so the month could be predicted by observation of the stars. Eusebius (Ecclesiastical History VII 32) quotes Aristobulos of Paneas2 as, "In celebrating Passover it is necessary not only for the sun but also the moon to have passed through the equinoctial segment." This means the full moon of Passover must occur after the Vernal Equinox, as in the present calculation of Easter. However, the month might still begin up to two weeks earlier.

The Jews did begin Nisan before the Vernal Equinox. This is confirmed by the later Rabbinical instruction of the early fourth century that, "When you see that the tequphah of Tebeth3 will extend to the sixteenth day of Nisan, declare that year embolismic without hesitation." (BT, Rosh Hashanah 21a) According to this instruction Nisan 1 could occur up to fifteen days before the Vernal Equinox, or in the first century CE as early as March 10. Thus, Nisan would begin with the first new moon following March 10, or not later than April 8. The second Adar was intercalated every two or three years as required. There was some variation possible for the late ripening of the barley or the approach of a Sabbatical year. Unlike the Babylonian calendar that only began Nisanu after the Vernal Equinox, the Second Temple calendar began Nisan with the new moon closest to the equinox, or up to two weeks before or after the opening of spring.

Josephus uses a sunrise day in describing the Passover in, "leaving nothing of what we sacrifice till the day following." (Ant. III 10:5, also Deut. 16:4) The lamb is sacrificed in the afternoon; if the next day began at sunset before the Passover meal this would present a contradiction. The use of sunrise reckoning is also found in the Jewish oral traditions from the end of the first century CE recorded in the Mishna, which states, "for all [offerings] that must be consumed `the same day', the duty last until the rise of dawn." (Berakoth 1:1) Since the references are to sacrifices conducted at the Temple, then the priests must be following the sunrise Second Temple calendar.

The use of the sunrise day is also found in the New Testament. John described resurrection morning with, "Now on the first day of the week Mary Magdalene came early to the tomb." (John 20:1) Later that same day, after dark, "When therefore it was evening, on that day the first day of the week, and when the doors were shut. . . Jesus came and stood in their midst." (John 20:19) John describes the same first day of the week, Sunday, as occurring both before and after sunset. This can only be with a sunrise day, unless he was using Roman reckoning from Midnight. John did not use evening reckoning for his days. There are many other references that seem to support sunrise reckoning, but none are conclusive. The Second Temple calendar appears to have also been used by John's Jewish community in Jerusalem.

Josephus consistently dated his months from Nisan, and the equivalence of the months of the Second Temple calendar and the Syro-Macedonian calendar is repeatedly confirmed.4 The use of the spring new year is also found in the dating of the coins of the revolt against Rome from 66 to 70 CE. Coins of the revolt are found dated year 5, which is correct when reckoned from Nisan.5 With Tishri reckoning there were only four years. The spring new year of the Sadducees was the official reckoning.

The New Testament makes frequent references to Jewish festivals, but the relative month is not given. The only reference to a numbered month is to the Annunciation in the "sixth month," (Luke 1:26) a probable reference to Luke's Syro-Macedonian calendar.6

During this same period it is also known that many early Jews used the Diaspora calendar. This calendar was similar to the Syro-Macedonian calendar, with a fall new year and evening reckoning for days. The Pharisees and many common people observed this calendar of the dispersion. The use of this unofficial reckoning was widespread, including its observance by many in Jerusalem. It became the basis of today's Jewish civil calendar.

In the later Diaspora there developed the tradition of observing some festivals two days in a row. These included Rosh Hashanah, the Tishri new year, and festivals such as the Feast of Unleavened Bread.7 The double observance compensated for the difference in calendars and locations. This ensured that Passover was celebrated on the correct day throughout the Diaspora. The second day was a bonus. This subject is discussed in detail in the chapter, "The Two Passovers."

With the destruction of the Second Temple in 70 CE and the disappearance of the Temple rites, the ecclesiastical Second Temple calendar was soon lost to most Jews. They did not adopt the calendar of their Roman conquerors, but soon shifted to the use of the luni-solar Diaspora calendar. This included the new year in the fall month of Tishri and the day reckoned from sunset. An important characteristic of this calendar was that it was at this time still based on observation of the moon.

The use of the spring new year continued into the second century CE.8 During the years 133-135 CE Bar Kokhba (Son of a Star)9 led the last Jewish revolt against Rome. That rebellion again ended in disaster for the Jewish people and split them from their Christian brothers. The war ended with the fall of Bethar, followed by the slaughter of Jewish men, women and children exceeding 580,000 (Dio Cassius, Roman History LXIX 14:1) and the banishment of the survivors from Jerusalem. Near the end many refugees, especially from En-gedi, hid in caves above the Dead Sea. The Romans tracked down these rebels, laid siege to the caves, and slew all. These caves have been excavated, and have yielded many documents that give some flavor of the dating of the period.10 To understand the dating of these documents requires a look first to the dating of the rebellion.

The third-century Christian historian, Eusebius, gave specific dating of the rebellion. He wrote that the end of the war was "during the eighteenth year of the reign of Hadrian in Beththera." (Ecclesiastical History IV 6) Eusebius also recorded in his Chronicles that the revolt ended in Hadrian's eighteenth year and began in his sixteenth year. Trajan died on August 8, 117 CE, and Hadrian succeeded as emperor later that month. Hadrian's eighteenth year was from August 134 to August 135 CE.11 The last Jewish stronghold, Bethar, fell to the Romans on Ab 9 (Mishna, Ta'anit 4:6). In 134 CE Ab 9 fell on July 19, too early to have been in Hadrian's eighteenth year. Ab 9 in 135 CE fell on August 7, still a few days within the eighteenth year. On August 7, 135 CE Bar Kokhba was killed and the rebellion crushed.

The beginning of the rebellion was in Hadrian's sixteenth year, from August 132 to August 133 CE. The Talmud says that "Bar Koziba reigned two and a half years." (Sanhedrin 93b, 97b) Two and a half years before August of 135 would be in early 133, which falls in the sixteenth year of Hadrian.12 This may indicate the Jew's capture of Jerusalem, with the rebellion having begun somewhat earlier,13 probably as a peasant revolt.14 After a period of a year or more of freedom fighters assaulting the Romans from hideaways, the Jews likely captured Jerusalem by early 133 CE. The Romans retook Jerusalem in early 134, since Hadrian returned to Rome by May 5 of that year.15 The siege and capture of Bethar then marked the end of the revolt in 135 CE.

Later Midrash says that the revolt lasted over three years, saying, "for three and a half years the Emperor Hadrian surrounded Bethar." (Lamentations Rabbah II 2:4; see also PT, Ta'anit 69b and Seder Olam Rabbah 30) However, Bethar was not under siege three years. Such dating is possibly an attempt to justify a claim for Bar Kokhba as messiah using Daniel's "time, times and half a time."16 Even the two and a half years have been explained as a different rabbinic understanding of Daniel 12:7.17 However, the explanation may simply be that the three and a half years were the length of the war and the time Hadrian surrounded Judea, as symbolized by Bethar; the two and a half years were the time from the capture of Jerusalem and the time that Bar Kokhba led the Jews as general.

The coins of Bar Kokhba can support the above dating. His coins are mostly struck over Roman coins dated in 131 and 132 CE.18 The capture of Jerusalem fits in early 133, prior to the issue of any intended new Roman coinage for that year. Coins of Bar Kokhba are known only dated to years 1 and 2 of the rebellion. The coins of year 1 appear to have been struck for the Succot holiday, Pentecost,19 which fell about June 2 in 133 CE. The coins of year 2 would have likely been struck in Nisan of 134 CE. However, Jerusalem's defenses were in ruins, and the retreat to Bethar was probably shortly after the beginning of that new year. There are undated coins that have been attributed to a year 3, but such is only speculation. They might have been struck before the capture of Jerusalem at the beginning of the revolt, or undated copies from years 1 or 2. The rebellion can best be dated to a beginning in early 133, the capture of Jerusalem after Nisan in 133, and the disastrous finish on that fateful day, Ab 9, or August 7, 135 CE.

Documents from the caves of the refugees were found written in Hebrew, Greek and Nabataean. A typical Greek document is dated: "In the ninth year of Imperator Traianus Hadrianus Caesar Augustus, in the consulship of M. Valerius Asiaticus II and Titus Aquilinus (125 CE), IV Id. October (four days before the ides of October, or October 12), and according to the era of the Provincia Arabia Twentieth Year on the twenty-fourth of the month of Hyperberetaius which is called Tishri."20 In addition to the Roman dating, we find the same alignment of the Syro-Macedonian and Jewish months as described by Josephus.

Of more interest is a Hebrew document that is dated from the year of the liberation of Israel by Bar Kokhba that mentions a Sabbatical year. The document reads, in part, "On the twentieth of Shevat, Year 2 of the liberation of Israel by Shimeon ben Kosiba, President of Israel. . . . This land I have leased from you as from today until the end of the eve of the shemittah, which are five complete years, that is harvest-fiscal years."21 The shemittah is the remission, or Sabbatical, year.

Year 2 is akin to a regnal year for Bar Kokhba and matches the dating on the coins. It is, thus, measured from the new year in Nisan of 133 CE. There was the expectation that Israel would remain free, since the writer expected the contract to still be in effect five years later. However, if Year 2 was measured from Tishri, then it would only be necessary to describe the end of the contract as the end of Year 6, since Year 7 would coincide with the Sabbatical year. Instead the contract refers to "harvest-fiscal" years that would begin in Tishri, to distinguish from the year of liberation under Bar Kokhba, as measured from Nisan. Thus, it is possible to identify how the Sabbatical year fell during the revolt.

From this document it is possible to find the years of the Bar Kokhba revolt from the Sabbatical year. This is laid out on Chart VI. N = Nisan; T = Tishri; C = Contract written in the eleventh month of Shebat; Hl, H2, H3, H4, and H5 = Harvest Years; S = Sabbatical year (no harvest). The discussion follows.

Chart VI
The Bar Kokhba Revolt and the Sabbatical Year

Sabbatical years fell in 132/133 and 139/140 CE. This historical sequence is well attested.22 One passage in the Talmud says that the destruction of the Second Temple in 70 CE was, "at the end of the seventh year." (`Arakin 11b; see also 12b.) This was the year 69/70 measured from Tishri, and the Sabbatical year before the Bar Kokhba rebellion then fell in 132/133. Three earlier dates in 1 Maccabees and Josephus confirm this sequence of Sabbatical years as correct.23 Thus, Year 1 of the Bar Kokhba revolt was then in 133/134 CE, measured from Nisan, since from Year 2 this would leave five harvests until the next Sabbatical year.24

The Bar Kokhba revolt is usually measured from 132, or even the autumn of 131 CE. This may be arrived at by several methods. First, there are the late references to the revolt having lasted three and a half years, and this is backdated from its conclusion in 135 CE. However, this time may be given without mention of the conflicting two and a half year war according to the Babylonian Talmud. Also, Eusebius' beginning in Hadrian's year 16 must be discounted or emended. The references in the Talmud and Eusebius suggest the revolt, at least measured from the capture of Jerusalem, did not begin before early 133 CE.

A second method of beginning the revolt early is by counting the years of Bar Kokhba from Tishri. Then the following five years from Year 2 of the above contract would become Years 3 to 7 of Bar Kokhba. The Sabbatical year would coincide with Year 8, with Year 1 being 132/133. This year has also been arrived at by using a different translation of the destruction of the Second Temple: "in a year following a Sabbatical Year." (Ta'anith 29a)25 This would place the Sabbatical year in 68/69,26 and again in 131/132. However, here the five harvests are correctly noted in Years 2 to 6, and 131/132 was the year before Year 1. Thus, 132/133 is again incorrectly said to be Year 1 of the revolt. If the destruction of the Second Temple took place at the end of a Sabbatical year, then Year 1 of the Bar Kokhba revolt is to be measured from Nisan of 133 CE.

A possible difficulty in beginning the revolt in 133 CE is found in a single document dated, "21 Tishri, year 4 of the Redemption of Israel."27 This has been suggested as a scribal error. However, for those who date the revolt from Nisan of 132 this must be dated after the fall of Bethar in 135, although the war was essentially over. This would be the only document to survive during the last six months of the revolt. If the scribe counted the years from Tishri, then it was correctly Year 4 from the suggested beginning in Nisan of 133 (backdated to Tishri of 132). This was still after the fall of Bethar in 135. The document has also been dated to 134, based on incorrectly dating the Sabbatical year in 131/132 and Tishri reckoning. The interpretation of this document is less than certain. It may be dated from the earlier stages of the rebellion, dated in a provincial town after the rebellion, or represent the continuing change from Nisan to Tishri reckoning.

The Bar Kokhba revolt is to be measured from the capture of Jerusalem in about April/May of 133 to August 7, 135 CE. The underground rebellion began somewhat earlier. The year of coins and reign was reckoned from the Nisan new year, still in accord with the Second Temple calendar.28

The use of the Diaspora calendar with the equivalent Syro-Macedonian months allowed the Jews to communicate with common dates. This became the generally used Jewish calendar, with the day measured from sunset. The Nisan new year was still acknowledged for festivals. However, it had long been known that the time of visibility of the new moon at different locations sometimes caused the festivals to fall on different days in different cities. In the Seleucid year 670, 358/359 CE, Hillel introduced the present calendar based on calculation, and the observation of the new moon was no longer necessary.29 This present calendar is no longer a true luni-solar calendar, but an artificial solar one. The new year in Tishri is calculated in a set pattern and the months have a set number of days. The day begins at sunset. And that is the way it is today.30

The stages of the evolution of the Jewish calendar are laid out in Chart VII. For the new year, A = Abib, or the later Nisan that began before or after the Vernal Equinox; B = Babylonian reckoning with Nisanu, or Nisan, beginning only after the Vernal Equinox; T = Tishri based on observation; TC = Tishri calculated according to the present civil calendar; D = Dios, equivalent to the eighth month of Heshvan. The intercalary month is counted from the appropriate new year.

It would appear that a Jewish calendar that begins the new year in the spring and the day at sunrise best fits as the calendar used in Scripture. Religious observances at the Temple were according to this calendar. A calendar using a fall new year and a day beginning at sunset was widely used from early times. This eventually became the base for the present Jewish calendar. These calendars were also influenced by the Babylonian and Syro-Macedonian calendars. From this perspective the question, "when?" is better understood in researching the early history of the Middle East.

1. J. B. Segal, "Intercalation and the Hebrew Calendar," VT 7 (1957), 300.

2. A Jewish philosopher of Alexandria, variously dated from the second century BCE to the second century CE.

3. The period between the winter solstice and the spring equinox.

4. For example, Ant. I 3:3, III 10:2, III 10:5, XI 4:7, XII 5:4, and XII 7:6.  There appears to be an inconsistency in XI 5:4 where the ninth month is called Tebeth instead of Kislev, equivalent to Apellaios; the original statement in Ezra 10:9 and I Esdras 9:5 only notes the ninth month without a secondary identification.

5. B. Kanael, "Notes on the Dates Used During the Bar Kokhba Revolt," IEJ 21 (1971).

6. The reference is to the sixth calendar month, which happens to coincide with the sixth month of Elizabeth's pregnancy.  Luke was a Greek from Antioch writing to Theophilus, a Greek in Antioch.  We can expect that the reference was to the sixth month according to Antioch's Syro-Macedonian calendar, that is Xanthikos in March.  This would accord with the traditional date for the Annunciation on March 25.  See the chapter "The Sixth Month" for further discussion.

7. Encyclopedia Judeaica (Jerusalem: Mc Millian [Keter], 1971), section on Passover; S. Zeitlin, Studies in the Early History of Judaism Vol. 1 (New York: KTAV, 1973), 223-233; A. P. Bloch, The Biblical and Historical Background of the Jewish Holy Days (New York: KTAV, 1978).

8. Kanael, "Bar Kokhba Revolt," 41.

9. PT, Ta'anith 68d presents Bar Kokhba as the fulfillment of Num. 24:17, "A star shall come forth from Jacob."

10. Y. Yadin, Bar-Kokhba (New York: Random House, 1971).

11. Eusebius used inclusive reckoning (Ecclesiastical History I 9-10), and the typical Roman dynastic reckoning.  Dynastic reckoning places Hadrian's year 1 from August 117 to August 118. (See chapter on "The Fifteenth Year of Tiberius.")  Some modern historians have counted his year 1 only until January 1 of 118, then his eighteenth year would be from January 1 to December 31, 134 CE.

12. Kanael, "Bar Kokhba Revolt," 40, n.7, supports a scribal transposition, with the beginning of the revolt in the fifteenth year of Hadrian to support a three and a half year revolt.

13. M. Gichon, "New Insights into the Bar Kokhba War and a Reappraisal of Dio Cassius 69.12-13," JQR 77 (1986).

14. S. Applebaum, "The Second Jewish Revolt (A.D. 131-135)," PEQ 116 (1984), 37.

15. S. Perowne, Hadrian (New York: Norton, 1961), 165.

16. B. Z. Wacholder, "Chronomessianism: The Timing of Messianic Movements and the Calendar of Sabbatical Cycles," HUCA 46 (1975), 217-218 supports the dating of the revolt from the spring of 132 to coincide with the Sabbatical year, as part of the claim of Bar Kokhba as Messiah.  He additionally places a Jubilee year in 132/133.

17. A. Reinhartz, "Rabbinic Perceptions of Simeon Bar Kosiba," JSJ 20, 2 (1989), 188, n. 58.

18. Yadin, Bar-Kokhba, 183.

19. Kanael, "Bar Kokhba Revolt," 41-42 places these coins at Succot in 132 CE, with the beginning year based on a three and a half year revolt.

20. Yadin, Bar-Kokhba, 223-233.

21. Yadin, Bar-Kokhba, 182-183.  For a discussion of this document and the complete text in English and Hebrew see B. Z. Wacholder, "The Calendar of Sabbatical Cycles During the Second Temple and the Early Rabbinic Period," HUCA 44 (1973), 176-179.

22. Wacholder, "Calendar of Sabbatical Cycles," reviews the evidence and provides specific dating from 519 BCE to 441 CE.

23. Sabbatical years are recorded in 163/162 BCE (1 Macc. 6:49; Ant. XII, ch. 9), 135/134 BCE (Ant. XIII 8:1), and 37/36 BCE (Ant. XIV 16:2).  The Sabbatical year sequence has also been interpreted as occurring one year earlier than the dates given here.  For further details see page 77, note 15 and pages 123-124.

24. L. L. Grabbe, "Maccabean Chronology: 167-164 or 168-165 BCE," JBL 110, 1 (1991), 62 notes that grain might be planted as late as Shebat, and Year 2 can be counted as the first of the five years of crops.  Also, a crop may have been planted before the contract was signed.  This fits Chart VI, although Grabbe places the sabbatical year in 138/139 instead of 139/140 CE.

25. This can be due to a difference in the perspective of the writer.  When the year is measured from Nisan the identification of the Sabbatical year is secondary as a "harvest-fiscal" year.  The Sabbatical year is here identified with the Nisan year in which it began.  The second half of the "harvest-fiscal" year is in the following Nisan year, or "in a year following the sabbatical year."

26. D. Blosser, "The Sabbath Year Cycle in Josephus," HUCA 52 (1981) supports this dating primarily by reinterpreting Josephus.  See the reply against Blosser's dating by B. Z. Wacholder, "The Calendar of Sabbath Years During the Second Temple Era: A Response," HUCA 54 (1983).

27. Kanael, "Bar Kokhba," 45.

28. A. Kloner, "Lead Weights of Bar Kokhba's Administration," IEJ 40, 1 (1990), 67 concludes that it is probable the system of weights used at the Second Temple was also used by Bar Kokhba.  By analogy the Second Temple system of time measurement was continued into this period.

29. Encyclopedia Judeaica, section on "Calendar."

30. S. Rogoff, "Israel's Calendar Confusions," JCR 22, 1 (1952), 33, notes the present modification of the calendar to use hours, minutes and seconds, and to drop the use of the day being divided into 1080 parts and 76 moments.  The original "civil" usage is being lost to the widespread use of the Gregorian calendar for such purposes.