John Morgan
Using the recent determination of the pattern of leap years in the Roman civil calendar between 41 BCE and 8 BCE and the epigraphically attested dates of 27 triumphs from 45 BCE to 19 BCE on the Fasti Triumphales Capitolini and the Fasti Triumphales Barberiniani, I shall demonstrate that the Romans avoided scheduling triumphal processions through the streets of Rome on the nundinae, the market-day of their 8-day “week”.
A combination of epigraphic and literary evidence shows that during the late Republican and early imperial periods, Rome and several other Italian cities held their markets every ninth day (by inclusive reckoning) on days called nundinae, which followed an 8-day cycle similar to the Judaeo-Christian 7-day weekly cycle. In the late Republic, when the Roman pontiffs exercised considerable freedom in deciding whether a year should be an ordinary year of 355 days or an intercalary year of 377 or 378 days, there was a general tendency to avoid a coincidence of the Kalends of January with the nundinae, which was considered a bad omen (Cassius Dio 40.47.1; Macrobius, Saturnalia 1.16.34). So strong was this tendency that after Julius Caesar’s reform of the Roman calendar in 46 BCE, the pontiffs extraordinarily made the year 41 BCE a leap year of 366 days so that the Kalends of January of 40 BCE would not fall on the nundinae (Cassius Dio 48.33.4). The equation of Peritios 14 of the Macedonian lunar calendar with January 23 in a leap year in the letter by the proconsul Paullus Fabius Maximus to the Greeks of Asia (Inschriften von Priene (1906), no. 105 = R.K. Sherk, Roman Documents from the Greek East (1969), no. 24) shows that not 9 BCE (as had generally been supposed by previous scholars) but rather 8 BCE was a leap year (Chris Bennett, ZPE 147 (2004), 167). This recent determination that not 9 BCE but 8 BCE was a leap year indicates that when for 36 years the pontiffs erroneously inserted a leap day every fourth year (by inclusive reckoning, based on their misunderstanding of quarto quoque anno), they did so in the 12 years
41, 38, 35, 32, 29, 26, 23, 20, 17, 14, 11, 8 BCE,
until in 8 BCE their error was discovered and corrected by Augustus, who ordered that no leap day be inserted in the following 12 years (Julius Solinus, Collectanea Rerum Memorabilium 1.40-47; Macrobius, Saturnalia 1.14.13-15). Since the number of days in 3 consecutive years, including the leap day in each 3rd year, is 365+365+366=1096, which is an integral multiple of 8, the pontiffs’ erroneous pattern of intercalating leap days ensured that in the 36 years from 41 to 5 BCE the Kalends of January fell on only 3 days of the 8-day Roman “week” and never coincided with the nundinae.
Between 45 and 19 BCE the precise dates of 27 triumphs are attested by the Fasti Triumphales Capitolini and the Fasti Triumphales Barberiniani. Knowledge of the pattern of inserting leap days during this period allows one to determine that these 27 triumphs fell on every day of the 8-day Roman “week” except one: the nundinae. The probability that this avoidance would be merely coincidental is (7/8)27 ~ 0.0272, less than a 3% chance. This suggests that the Romans deliberately scheduled their triumphs to avoid coincidences with the nundinae, for the obvious reason that it would have been distracting or even disruptive if a triumphal procession through the crowded streets of Rome and the Forum had coincided with the “weekly” market on the nundinae. This discovery that triumphs were not celebrated on the nundinae, together with the many attested dates of earlier triumphs on the Fasti Triumphales Capitolini and in Livy’s Ab Urbe Condita and other literary sources, should facilitate further progress in reconstructing the operation of the Roman calendar in the last centuries of the Republican period.