Venus and Eclipse Observations and Chronology

The earliest surviving Babylonian observational records that refer to appearances and disappearances of Venus are from the reign of the king Ammisaduqa (or Ammizaduga), king of Babylon 146 years after the accession of his ancestor, Hammurabi. Alternative possible dates for Ammizaduqa's reign have been determined from the Venus data13 in combination with historical records, in particular the contemporaneity of Hammurabi and Shamshi-Adad, 39th King of Assyria. It has been generally accepted that these data permit only four alternative interpretations for the beginning date of the reign of Ammizaduqa: b.c. 1702 (the "Long Chronology" favored by Huber and Weir), 1646 (favored by Rowton in the Cambridge Ancient History, the most influential source in English), 1638, and 1582 (the "Short Chronology," vigorously defended by van der Waerden 1974).14 Rowton argued that the average length of reign of

13 The text referring to these observations, which are tied into the lunar months, is known in a large number of fairly late copies and has been published by Langdon and Fotheringham with Schoch (1928) and more recently by Weir (1972). The text has been further updated by Reiner and Pingree (1975), who emphasize that the first of the two sets of apparent observations seems much more reliable, whereas the second set might have been completed from some other set of observations entirely. A discussion by Weir (1983) has much new data and interpretation. Weir (1995) is probably the most understandable account of the nature of the data and the problems of interpretation.

14 Mitchell (1990) has recently proposed an "Ultra-Short Chronology" that meets the astronomical criteria very well but requires massive rejection of the information in the Assyrian king list and reinterpretation of much else.

contemporary Hittite kings was too high under the Long Chronology and too low under the Short Chronology. This does not seem to us a strong historical argument. Average reign lengths vary greatly depending on many different cultural and biological factors, particularly mechanisms of inheritance. van der Waerden's astronomical arguments seem to us considerably stronger. He demonstrated clearly that the rising dates of Venus were consistently too early with the 1646 solution and consistently too late with the 1638 solution. Neither the 1702 solution nor the 1582 solution showed much consistency but rather varied in both directions from the norm, as would be expected with real phenomena. Therefore, the two middle chronologies were the least likely, astronomically. In 1974, Huber discovered that the arcus visionis (see ยง3.1.5) values determined from late Babylonian sources clearly indicated that the position in the text that had been uniformly identified as the "last visibility" of Venus, was instead the "first invisibility." This discovery improved the correspondence between the tablet and astronomical backcalculations for all solutions. Unfortunately, all solutions still contained some apparent impossibilities. Weir's analysis for the first time differentiated between western observations, uninterupted across the desert to the horizon, and eastern observations looking toward the mountains.

The chronicles assign 21 years to Ammizaduqa and 31 to his successor, Samsuditana. Of the five solar-lunar eclipse pairs that have been suggested as possibly relevant to the end of the reign of Samsuditana in their relationship to postulated Venus dates of Ammizaduqa, the pair of 1713 b.c. would fall 45 years after 1758 b.c. (when the appropriate Venus phenomena are found), 56 years still earlier than the "Long Chronology." The pair of 1659 b.c. are 43 years after the 1702 b.c. date of the "Long Chronology." There is no eclipse pair to match either of the two middle chronologies. The eclipse pair of 1532 b.c. matches the Venus data of 1582 b.c., 50 years earlier. Huber (1982, pp. 40-41) argues that the physical conditions of the eclipse pair of 1713 b.c. match the description better than do either of the other two. Mitchell found an eclipse pair in 1362 b.c., some 57 years after Venus phenomena that fit 1419 b.c. These dates constitute his preferred solution, which we think is incompatible with the Assyrian king list, with the Anatolian tree-ring data, with the Egyptian historical data, and with our astronomical date for Ramesses II.

Mitchell found another eclipse pair in 1178 b.c., some 30 years after appropriate Venus phenomena of 1208 b.c. The discrepancy between the 30-year interval and the 52-year interval of the chronicles would be enough to eliminate the last possibility. Clearly, the 50 years between 1532 b.c. and 1582 b.c. come closer to matching the 52 years of the chronicles than do any of the others. Both Huber and Mitchell have emphasized the bad fit of the lunar data of Ammizaduqa's reign with an accession in 1582 b.c.; in particular, known inscriptions have an excessive number of 30-day lunations when contrasted with backcalculations for that period. In light of what we wrote earlier about the 30-day months of the 360-day year, we ask whether it is possible that some of these 30-day months were not lunations. Only a careful study of the originals by a competent scholar of Assyro-Babylonian can answer this question. On nearly all other grounds known to us, the Short Chronology seems to be the most acceptable.

Some problems remain with respect to the appropriate dating of the earlier lunar eclipses. The death of Irra-imitti of Isin is not included in Mitchell's account because we have no astronomical details. If we accept the Short Chronology, with the accession of Hammurabi in 1728 b.c., the death of Irra-imitti should fall on historical grounds between 1803 b.c. and 1797 b.c. The total lunar eclipse of 1801 b.c. seems a probable candidate, but this selection is based on the historical evidence rather than on astronomy. This would put the accession of Ishbi-erra of Isin at about 1964 b.c. He was a contemporary of Ibbi-Sin, the last king of the Third Dynasty of Ur. According to Huber, the only probable lunar eclipse fitting the conditions in this time period is the partial eclipse of 26 April 1932 b.c., with the next earlier eclipse at 2004 b.c. Although Mitchell substantially increased the number of possibilities for this eclipse, he added no others for the relevant time period. Because the Ur III Dynasty lasted 109 years and began shortly after the defeat of the Gutians, the latter eclipse should date within a few years before 2040 b.c. There was, in fact, an eclipse matching the specifiable conditions on 26 August 2041 b.c., but the total eclipse of 26 July 2049 b.c. is probably more likely, assuming that Utu-hegal's defeat of the Gutians was near the beginning of his short reign (less than eight years).

Surprisingly, there are apparently no lunar eclipses at Ur between 2056 b.c. and 1946 b.c. that agree with the descriptive data and, hence, no possible matches for a king of Ur murdered by his son during the entire III Dynasty in the Short Chronology.

There seem to be four possible explanations for this discrepancy:

(1) The Short Chronology is wrong.

(2) There are errors in the description of the eclipse.

(3) The astronomical calculations somehow missed an appropriate eclipse.

(4) The King of Ur (who is not named) was not one of the Kings of Dynasty III, but one of the Kings of Dynasty II.

The data on the eclipse during the Dynasty of Agade seem to us insufficient to identify a lunar eclipse that may be associated with any of seven rulers in a realistic range of about 200 years for any of the proposed chronological schemes. The solar eclipse during the reign of Sargon may ultimately be more promising.

Mitchell (1990, p. 24) maintains that the only solution that fits all of the eclipse data and the Venus data is the Ultrashort Chronology, which also provides a very good fit for the lunar data. We agree with him that solutions should be made with respect to sets. If it is true that the best solution astronomically is contrary to the best nonastronomical evidence, we think that this presents an important methodological problem. We are also bothered by the evidence, particularly on lunations (assembled by Huber) and on Venus (assembled by Weir), which supports the Long Chronology. Overall, in spite of doubts and problems, we continue to favor the Short Chronology.

0 0

Post a comment