For Rome, the year which we call 46 B.C. is called by Macrobius the last year of the muddled reckoning, annus confusionis ultimus, and it was 445 days long, so much had the nominal dates got behind the real ones; with the next year began the Julian reckoning, albeit with sundry boggles on the part of the Roman officials, who did not quite understand it, and long delays before the whole Western world adopted it.
“The Pre-Caesarian Calendar: Facts and Reasonable Guesses”
H. J. Rose
The Classical Journal, Vol. 40, No. 2 (Nov., 1944), pp. 65-76
If this 1944 article were the first and last word on the subject, it would be closed. The last year of confusion, 46 B.C., would be fixed at 445 days long, based on the addition of a regular intercalation of 23 days, plus Caesar’s extra 67 days.
A retired professor wrote me to ask whether the number of days in the last year of confusion should be 445 or 455 because I had both numbers on my site and he had found both numbers elsewhere, including in print. His OCD didn’t inspire him with confidence because where it cited Censorinus, (which should be De Die Natali 20, 8) the entry listed it wrong. He thought the 355 number might be from people not realizing the basic year to which the extra days were added was not 365, but only 355 days long.
I told him that my 455 was a typing error since elsewhere on my site I had 445 and I am embarrassingly inclined to make such errors. He then pointed out some other references, including Mommsen whose observations might lead one to the conclusion that there were 422 or so days in 46 (before proofing, the 46 was 466, so wrong-number-typing is a serious problem I have to face). The reason for this is that Mommsen says only 67 extra days were added and 355+67=422:
Of a kindred nature was the reform of the calendar. The republican calendar, which strangely enough was still the old decemviral calendar–an imperfect adoption of the -octaeteris- that preceded Meton (117)–had by a combination of wretched mathematics and wretched administration come to anticipate the true time by 67 whole days, so that e. g. the festival of Flora was celebrated on the 11th July instead of the 28th April.
Mommsen seems either to have forgotten to mention the intercalation that normally occurred at regular intervals (albeit, missed recently) or to have followed ancient writers who either failed to pay attention to the February intercalation or considered it unremarkable. In all, the 23-day February intercalation may have been assumed, may not have been assumed, or may have been thought not to have happened. How Mommsen gets a 67 day gap between the floralia’s proper and scheduled dates I don’t understand, although a helpful forum poster said it’s only by wretched mathematics, in this case, Mommsen’s (although given his customary care, that seems improbable), because he’s the one using the figures.
Further evidence for the loss of intercalation can be gleaned from Caesar’s
correction of the calendar in 46. Dio says that there was a special supplement of sixty-seven days, while according to Censorinus the year 46 had a total of 445 days. I This implies that there had already been a normal intercalary month of twenty-three days. By his further addition of sixty-seven days Caesar was supplying three Republican intercalary months. Now in the period March 65-March 45 B.C. there had probably been only six intercalations-in 63, twice between 62 and 58, in 54, 52, and 46. At first sight this indicates the loss of four intercalary months, but intercalation every alternate year would have produced a surplus of twenty days after twenty years, and so Caesar’s calculations would have only required the addition of three intercalary months to bring the calendar into order by March 45 B.C. It is also possible that the pontifices in the fifties were allowing for this and one of the intercalary months was omitted for the sake of the correctness of the calendar, not through negligence or for political purposes.
“Nundinae and the Chronology of the Late Roman Republic”
A. W. Lintott
The Classical Quarterly, New Series, Vol. 18, No. 1 (May, 1968), pp. 189-194
There is also some doubt as to whether the year 45 began on January 1 or 2, so the length of the preceding year of confusion might have gained a day at the very end.
My correspondence with the retired professor continued. Nothing was resolved. The professor thinks the reason for the confusion has to do with the fact that fully credentialed scholars are unwilling to put their reputations on the line by siding with Censorinus or Dio.
What are your thoughts on the length of the final year of confusion?
Another question: Based on the following passage from Rose’s article, what does Rose think people should call the year?
2 Macrobius, Saturnalia I, 14, 3; no one, except moderns who should know better, ever calls it the annus confusionis simply. He does not give the date, but that is supplied by Censorinus, De Die Natali 20, 8, who says it was in Caesar’s third consulate, and we know when that was from a number of other authors; cf. W. Fynes Clinton, Fasti Hellenici (Oxford, 1827-41), under the year 46 B.C.