Short URL for this page:

[image ALT: Much of my site will be useless to you if you've got the images turned off!]
Bill Thayer

[image ALT: Cliccare qui per una pagina di aiuto in Italiano.]

[image ALT: Haga clic aquí para una página en Español.]

[Link to a series of help pages]
[Link to the next level up]
[Link to my homepage]

 p226  Calendarium

The Roman section only (pp226‑233)
of an article by Thomas Hewitt Key, M.A., Professor of Comparative Grammar in University College, London
on pp222‑233 of

William Smith, D.C.L., LL.D.:
A Dictionary of Greek and Roman Antiquities, John Murray, London, 1875.

CALENDA′RIUM, or rather KALENDA′RIUM, is the account-book, in which creditors entered the names of their debtors and the sums which they owed. As the interest on borrowed money was due on the Calendae of each month, the name of Calendarium was given to such a book (Senec. De Benef. I.2, VII.10). The word was subsequently used to indicate a register of the days, weeks, and months, thus corresponding to a modern almanac or calendar.

[. . .]

2. Roman Calendar — The Year of Romulus. — The name of Romulus is commonly attached to the year which is said to have prevailed in the earliest times of Rome; but tradition is not consistent with regard to the form of it. The historians Licinius Macer and Fenestella maintained that the oldest year consisted of twelve months, and that it was already in those days an annus vertens, that is, a year which coincided with the period of the sun's course. Censorinus, however, in whose work this statement occurs (De Die Natali, c20; compare also the beginning of c19), goes on to say that more credit is due to Graccanus, Fulvius (Nobilior), Varro, and others, according to whom the Romans in the earliest times, like the people of Alba from whom they sprang, allotted to the year but ten months. This opinion is supported by Ovid in several passages of his Fasti (I.27, I. 43, III.99, 119, 151); by Gellius (Noct. Att. III.16),º Macrobius (Saturn. I.12), Solinus (Polyh. I), and Servius (ad Georg. I.43). Lastly, an old Latin year of ten months is implied in the fact, that at Laurentum (Macrob. I.15) a sacrifice was offered to Juno Kalendaris on the first of every month except February and January. These ten months were called Martius, Aprilis, Maius, Junius, Quinctilis, Sextilis, September, October, November, December. That March was the first month in the year is implied in the last six names; and even Plutarch, who ascribes twelve months to the Romulian year (Numa, c18), places Januarius and Februarius at the end. The fact is also confirmed by the ceremony of rekindling the sacred fire in the temple of Vesta on the first day of March, by the practice of pla­cing fresh laurels in the public buildings on that day, and by many other customs recorded by Macrobius (I.12). With regard to the length of the months, Censorinus, Macrobius, and Solinus agree in ascribing thirty-one days to four of them, called pleni menses; thirty to the rest called cavi menses. The four longer months were Martius, Maius, Quinctilis, and October; and these, as Macrobius observes, were distinguished in the latest form of the Roman calendar by having their nones two days later than any of the other months. The symmetry of this arrangement will appear by pla­cing the numbers in succession:— 31, 30; 31, 30; 31, 30, 30; 31, 30, 30. Ovid, indeed, appears to speak of the months as coinciding with the lunar period:—

"Annus erat decimum cum luna repleverat annum:"

but the language of a poet must not be pressed too closely. On the other hand, Plutarch, in the passage already referred to, while he assigns to the old year twelve months and 365 days, speaks of the months as varying without system between the limits of twenty and thirty-five days. Such an irregularity is not incredible, as we find that even when Censorinus wrote (A.D. 238), the Alban calendar gave 36 days to March, 22 to May, 18 to Sextilis, and 16 to September; while at Tusculum Quinctilis had 36 days, October 32; and again at Aricia the same month, October, had no less than 39 (Censorinus, c22). The Romulian year, if we follow the majority of authors, contained but 304 days; a period differing so widely from the real length of the sun's course, that the months would rapidly revolve through all the seasons of the year. This inconvenience was remedied, says Macrobius (I.13), by the addition of the proper number of days required to complete the year; but these days, he goes on to say, did not receive any name as a month. Servius speaks of the intercalated period as consisting of two months, which at first had no name, but were eventually called after Janus and Februus. That some system of intercalation was employed in the Romulian year, was also the opinion of Licinius Macer (Macrob. I.13). This appears to be all that is handed down with regard to the earliest year of the Romans.

As a year of ten months and 304 days, at once falls greatly short of the solar year, and contains no exact number of lunations, some have gone so far as to dispute the truth of the tradition in whole or in part, while others have taxed their ingenuity to account for the adoption of so anomalous a year. Puteanus (De Nundinis, in Graevius' Thesaurus, vol. VIII), calling to mind that the old Roman or Etruscan week contained eight days,​1 every eighth  p227 day being specially devoted to religious and other public purposes, under the name of nonae or nundinae, was the first to point out that number 304 is a precise multiple of eight. To this observation, in itself of little moment, Niebuhr has given some weight, by further noti­cing that the 38 nundines in a year of 304 days tally exactly with the number of dies fasti afterwards retained in the Julian calendar. Another writer, Pontedera, observed that 304 bore to 365 nearly the ratio of 5 to 6, six of the Romulian years containing 1824, five of the longer periods 1825 days; and Niebuhr (Rom. Hist. vol. 1 p271), who is a warm advocate of the ten-month year, has made much use of this consideration. He thus explains the origin of the well-known quinquennial period called the lustrum, which Censorinus (c18) expressly calls an annus magnus, that is, in the modern language of chronology, a cycle. Moreover, the year of ten months, says the same writer (p279), was the term for mourning, for paying portions left by will, for credit on the sale of yearly profits; most probably for all loans; and it was the measure for the most ancient rate of interest. [Fenus.] Lastly, he finds in the existence of this short year the solution of certain historical difficulties. A peace, or rather truce, with Veii was concluded in the year 280 of Rome, for 40 years. In 316 Fidenae revolted and joined Veii, which implies that Veii was already at war with Rome; yet the Veientines are not accused of having broken their oaths (Liv. IV.17). Again, a twenty-years' truce, made in 329, is said, by Livy, to have expired in 347 (IV.58). These facts are explained by supposing the years in question to have been those of ten months, for 40 of these are equal to 33⅓ ordinary years, 20 to 16⅔; so that the former truce terminated in 314, the latter in 346. Similarly, the truce of eight years concluded with the Volscians in 323, extended in fact to no more than 6⅔ full years; and hence the Volscians resumed the war in 331, without exposing themselves to the charge of perjury.

These ingenious and perhaps satisfactory speculations of the German critic, of course imply that the decimestrial year still survived long after the regal government had ceased; and in fact he believes that this year, and the lunar year, as determined by Scaliger's proposed cycle of 22 years, co-existed from the earliest times down to a late period. The views of Niebuhr do not require that the months should have consisted of 31 or 30 days; indeed it would be more natural to suppose that each month, as well as the year, contained a precise number of eight-day weeks; eight of the months, for instance, having four such weeks, the two others but three. Even in the so‑called calendar of Numa we find the Etruscan week affecting the division of the month, there being eight days between the nones and ides, from which circumstances the nones received their name; and again two such weeks from the ides to the end of the month; and this, whether the whole month contained 31 or 29 days.

The Year of Numa.— Having described the Romulian year, Censorinus (c20) proceeds thus:— "Afterwards, either by Numa, as Fulvius has it, or according to Junius by Tarquin, there was instituted a year of twelve months and 355 days, although the moon in twelve lunations appears to complete but 354 days. The excess of a day was owing either to error, or what I consider more probable, to that superstitious feeling, according to which an odd number was accounted full (plenus) and more fortunate. Be this as it may, to the year which had previously been in use (that of Romulus) one-and‑fifty days were now added; but as these were not sufficient to constitute two months, a day was taken from each of the before-mentioned hollow months, which added thereto, made up 57 days, out of which two months were formed, Januarius with 29, and Februarius with 28 days. Thus all the months henceforth were full, and contained an odd number of days, save Februarius, which alone was hollow, and hence deemed more unlucky than the rest." In this passage it is fitting to observe that the terms pleni and cavi menses are applied in a sense precisely opposite to the practice of the Greek language in the phrases μῆνες πληρεῖς and κοῖλοι. The mysterious power ascribed to an odd number is familiar from the Numero deus impare gaudet of Virgil. Pliny also (Plin. H. N. XXVIII.5) observes, — Impares numeros ad omnia vehementiores credimus. It was of course impossible to give an odd number of days at the same time to the year on the one hand, and to each of the twelve months on the other; and yet the object was in some measure effected by a division of February itself into 23 days, and a supernumerary period of five days. (See the mode of intercalation below.) The year of Numa then, according to Censorinus, contained 355 days. Plutarch tells us that Numa estimated the anomaly of the sun and moon, by which he means the difference between twelve lunations and the sun's annual course at eleven days, i.e. the difference between 365 and 354 days. Macrobius, too, says that the year of Numa had at first 354, afterwards 355 days. Compare herewith Liv. I.19; Ovid. Fasti, I.43, III.151; Aurel. Vict. c3; Florus, I.2; Solinus, c1.

Twelve lunations amount to 354 days, 8h. 48′ 36″, so that so‑called year of Numa was a tolerably correct lunar year; though the months would have coincided more accurately with the single lunations, if they had been limited to 30 and 29 days, instead of 31, 29, and 28 days. That it was in fact adapted to the moon's course is the concurrent assertion of ancient writers, more particularly of Livy, who says: (Numa) omnium primum ad cursum lunae in duodecim mensis discribit annum. Unfortunately however, many of the same writers ascribe to the same period the introduction of such a system of intercalation as must at once have dislocated the coincidence between the civil month and the lunar period. At the end of two years the year of Numa would have been about 22 days in arrear of the solar period, and accordingly it is said an intercalary month of that duration, or else of 23 days, was inserted at or near the end of February, to bring the civil year into agreement with the regular return of the seasons. Of this system of intercalation a more accurate account shall presently be given. But there is strong reason for believing that this particular mode of intercalation was not contemporary in origin with the year of Numa.

In antiquarian subjects it will generally be found that the assistance of etymology is essential; because the original names that belong to an institution often continue to exist, even after such changes have been introduced, that they are no longer adapted to the new order of things; thus they survive as useful memorials of the past. In this  p228 way we are enabled by the original meaning of words, aided by a few fragments of a traditional character, to state that the Romans in early times possessed a year which altogether depended upon the phases of the moon. The Latin word mensis (Varro, De Ling. Lat. VI., or in the old editions, V.54), like the Greek μήν or μείς, and the English month, or German monath, is evidently connected with the word moon. Again, while in the Greek language the name νουμηνία (new-moon), or ἕνη καὶ νέα, given to the first day of a month, betrays its lunar origin, the same result is deduced from the explanation of the word kalendae, as found in Macrobius (I.15). "In ancient times," says that writer, "before Cn. Flavius the scribe, against the pleasure of the patricians, made the fasti known to the whole people (the end of the 4th century B.C.), it was the duty of one of the pontifices minores to look out for the first appearance of the new moon; and as soon as he descried it, to carry word to the rex sacrificulus. Then a sacrifice was offered by these priests, after which the same pontifex having summoned the plebs (calata plebe) to a place in the capitol, near the Curia Calabra, which adjoins the Casa Romuli, there announced the number of days which still remained to the nones, whether five or seven, by so often repeating the word καλῶ." There was no necessity to write this last word in Greek characters, as it belonged to the old Latin. In fact, in this very passage, it occurs in both calata and calabra; and again, it remained to the latest times in the word nomenclator. In regard to the passage here quoted from Macrobius, it must be recollected that while the moon is in the immediate vicinity of the sun, it is impossible to see it with the naked eye, so that the day on which it is first seen is not of necessity the day of the actual conjunction. We learn elsewhere that as soon as the pontifex discovered the thin disc, a hymn was sung, beginning Jana novella, the word Jana (Macrob. Sat. I.9; Varro, De Re Rust. I.37) being only a dialectic variety of Diana, just as Diespiter or Diupiter corresponds to Jupiter; and other examples might readily be given, for the change occurs in almost every word which has the syllables de or di before a vowel. Again, the consecration of the kalends to Juno (Ovid. Fasti, I.55, VI.39; Macrob. Sat. I.9.15) is referred by the latter writer to the fact that the months originally began with the moon, and that Juno and Luna are the same goddess; and the poet likewise points at the same connection in his explanation of Juno's epithet Lucina. Moreover, at Laurentum Juno was worshipped as Juno Kalendaris. Even so late as 448 B.C. strictly lunar months were still in use; for Dionysius (Antiq. X.59) says that Appius, in that year, received the consular authority on the ides of May, being the day of full moon, for at that time, he adds, the Romans regulated their months by the moon. In fact, so completely was the day of the month, which they called the ides, associated with the idea of the full moon, that some derived the word ἀπὸ τοῦ εἴδους, quod eo die plenam speciem luna demonstret (Macrob. ibid.). Quietly to insert the idea of plenam, when the Greek word signified merely speciem, is in accordance with those loose notions which prevailed in all ancient attempts at etymology. But though the derivation is of course groundless, it is of historical value, as showing the notion connected with the term ides.

For the same reason probably the ides of March were selected for the sacrifice to the goddess Anna Perenna, in whose name we have nothing more than the feminine form of the word annus, which, whether written with one n or two, whether in its simple form annus, or diminutive annulus, still always signifies a circle. Hence, as the masculine form was easily adopted to denote the period of the sun's course, so the feminine in like manner might well be employed to signify, first the moon's revolution, and then the moon herself. The tendency among the Romans to have the same word repeated, first as a male and then as a female deity, has been noticed by Niebuhr; and there occurs a complete parallel in the name Dianus, afterwards Janus, for the god of dies, or light, the sun; Diana, afterwards Jana, for the goddess of light, the moon; to say nothing of the words Jupiter and Juno. That the month of March should have been selected arose from its being the first of the year, and a sacrifice to the moon might well take place on the day when her power is fully displayed to man. The epithet Perenna itself means no more than ever-circling. Nay, Macrobius himself (c12) connects the two words with annus, when he states the object of the sacrifice to be — ut annare perennareque commode liceat.

Another argument in favour of the lunar origin of the Roman month, is deducible from the practice of counting the days backward from the Kalends, Nones, and Ides; for the phrases will then amount to saying — "It wants so many days to the new moon, to the first quarter, to full moon." It would be difficult, on any other hypothesis, to account for the adoption of a mode of calculation, which, to our notions at least, is so inconvenient; and indeed it is expressly recorded that this practice was derived from Greece, under which term the Athenians probably are meant; and by these we know that a strictly lunar year was employed down to a late period (Macrob. I.16).

But perhaps the most decisive proof of all lies in the simple statement of Livy (I.19), that Numa so regulated his lunar year of twelve months by the insertion of intercalary months, that at end of every nineteenth year it again coincided with the same point in the sun's course from which it started. His words are — Quem (annum) intercalaribus mensibus interponendis ita dispensavit ut vicensimo anno ad metam eandem solis unde orsi sunt, plenis annorum omnium spatiis, dies congruerent. We quote the text; because editors, in support of a theory, have taken the liberty of altering it by the insertion of the word quarto, forgetting too that the words quarto et vicensimo anno signify, not every twenty-fourth year, which their theory requires, but every twenty-third, according to that peculiar view of the Romans which led them to count both the extremes in defining the interval from one point to another; and which still survives in the medical phrases tertian and quartan ague, as well as in the French expressions huit jours for a week, and quinze jours for a fortnight. Accordingly, it is not doing violence to words, but giving the strict and necessary meaning to them, when, in our own translation of the passage in Livy, we express vicensimo anno by every nineteenth year.

Now 19 years, it is well known, constitute a most convenient cycle for the conjunction of a lunar and solar year. A mean lunation, or synodic month, according  p229 to modern astronomy, is 29d. 12h. 44′3″, and a mean tropical year 365d. 5h. 48′48″. Hence it will be found, that 235 lunations amount to 6939d. 16h. 31′45″, while 19 tropical years give 6939d. 14h. 27′12″, so that the difference is only 2h. 4′33″. Although it was only in the second century B.C. that Hipparchus gave to astronomical observations a nicety which could pretend to deal with seconds*; yet even in the regal period of Rome, the Greek towns in the south of Italy must already have possessed astronomers, from whom the inhabitants of Latium could have borrowed such a rough practical knowledge of both the moon and sun's period, as was sufficient to show that at the end of 19 solar years the moon's age would be nearly what it was at the commencement; and it will be recollected that the name of Numa is often connected by tradition with the learning of Magna Graecia. At any rate a cycle of 19 years was introduced by Meton at Athens, in the year 432 B.C.; and the knowledge of it among the learned may probably have preceded by a long period its introduction into popular use, the more so as religious festivals are generally connected with the various divisions of time, and superstition therefore would be most certainly opposed to innovations of the almanack. How the Romans may have intercalated in their 19 lunar years the seven additional months which are requisite to make up the whole number of 235 (= 12 × 19 + 7) lunations, is a subject upon which it would be useless to speculate. From a union of these various considerations, it must be deemed highly probable that the Romans at one period possessed a division of time dependent upon the moon's course.

Year of the Decemviri (so called by Ideler).— The motives which induced the Romans to abandon the lunar year are no where recorded; nor indeed the date of the change. We have seen, however, that even in the year 448 B.C., the year was still regulated by the moon's course. To this must be added that, according to Tuditanus and Cassius Hemina, a bill on the subject of intercalation was brought before the people by those decemviri, who added the two new tables to the preceding Ten (Macrob. I.13), that is in the year 450 B.C. That the attention of these decemviri was called to the calendar is also proved by the contents of the Eleventh Table, wherein it is decreed that "the festivals shall be set down in the calendars." We have the authority of Varro indeed, that a system of intercalation already existed at an earlier date; for he says that there was a very ancient law engraved on a bronze pillar by L. Pinarius and Furius in their consulate cui mentio intercalaris ascribitur. We add the last words in Latin from the text of Macrobius (c. 13), because their import is doubtful. If we are right in interpreting them thus — "the date upon which is expressed by a month called intercalary", all that is meant may be one of the intercalary lunations, which must have existed even in the old lunar year. At the period of the decemviral legislation there was probably instituted that form of the year of 354 days, which was corrected by the short intercalary month, called Mercedonius, or Mercidinus; but so corrected as to deprive the year and months of all connection with the moon's course. The length of the several ordinary months was probably that which Censorinus has erroneously allotted to the months of Numa's lunar year, viz.:—

Martius 31 days
Aprilis 29 days
Maius 31 days
Junius 29 days
Quinctilis 31 days
Sextilis 29 days
September 29 days
October 31 days
November 29 days
December 29 days
Januarius 29 days
Februarius 28 days

Such, at any rate, was the number of days in each month immediately prior to the Julian correction; for both Censorinus and Macrobius say that Caesar added two days to Januarius, Sextilis, and December, and one to Aprilis, Junius, September, and November. Hence Niebuhr appears to have made an error when he asserts (vol. II note 1179) that July acquired two more days at the reformation of the calendar, and founds thereon a charge of carelessness again Livy. Moreover that November had but 29 days prior to the correction, in other words, that the XVII Kal. Dec. immediately followed the Idus Nov., appears from a comparison of Cicero's letters to Tiro (Ad Fam. XVI.7.9); for he reaches Corcyra a. d. V. Id. Nov., and on the XV Kal. Dec. complains — Septumum jam diem tenebamur. The seven days in question would be IV. Id., III. Id., Prid. Id., Id. Nov., XVII. Kal. Dec., XVI. Kal. Dec., XV. Kal. Dec. That the place of the nones and ides was in each month the same before the Julian correction as afterwards, is asserted by Macrobius.

The main difficulty is with regard to the mode of intercalation. Plutarch, we have already observed, speaks of an intercalation, by him referred to Numa, of 22 days in alternate years in the month of February. Censorinus, with more precision, says that the number of days in each intercalation was either 22 or 23, and Macrobius agrees with him in substance. Of the point at which the supernumerary month was inserted, the accounts are these:— Varro (De Ling. Lat. VI.55) says, the twelfth month was February; and when intercalations take place, the five last days of this month are removed. Censorinus agrees herewith, when he places the intercalation generally (potissimum) in the month of February, between the Terminalia and the Regifugium, that is immediately after the day called by the Romans a. d. VI. Kal. Mart. or by us the 24th of February. This, again, is confirmed by Macrobius. The setting aside of the last five days agrees with the practice which Herodotus ascribes to the Egyptians of considering the five days over the 360 as scarcely belonging to the year, and not pla­cing them in any month. So completely were these five days considered by the Romans to be something extraneous, that the soldier appears to have received pay only for 360 days. For in the time of Augustus the soldier received deni asses per day, i.e. 1016 of a denarius; but Domitian (Suet. Dom. 7) addidit quartum stipendium aureos ternos. Thus, as 25 denarii made an aureus, the annual pay prior to Domitian was (360 × 10) ÷ 16 denarii = (360 × 10) ÷ (16 × 25) aurei = 9 aurei; and thus the addition of three aurei was precisely a fourth more. Lastly, the festival Terminalia, as its name implies, marked the end of the year, and this by the way again proves that March was originally the first month.

The intercalary month called Μερκίδινος, or Μερκηδόνιος (Plutarch, Numa, 19; Caes. 59).  p230 We give it in Greek characters, because it happens somewhat strangely that no Latin author has mentioned the name, the term mensis interkalaris or interkalarius supplying its place. Thus, in the year of intercalation, the day after the ides of February was called, not as usual a. d. XVI. Kalendas Martias, but a. d. XI. Kalendas interkalares. So also there were the Nonae interkalares, and Idus interkalares, and after this last came either a. d. X. or XVI. Kal. Mart., according as the month had 22 or 23 days, or rather, if we add the five remaining days struck off from February, 27 or 28 days. In either case the Regifugium retained its ordinary designation a. d. VI. Kal. Mart. (See Asconius, Ad Orat. pro Milone, and the Fasti Triumphales, 493, A.U.C.). When Cicero writes to Atticus (VI.1), Accepi tuas litteras a. d. V. Terminalia (i.e. Feb. 19); he uses this strange mode of defining a date, because, being then in Cilicia, he was not aware whether any intercalation had been inserted that year. Indeed, he says, in another part of the same letter, Ea sic observabo, quasi intercalatum non sit.

Besides the intercalary month, mention is occasionally made of an intercalary day. The object of this was solely to prevent the first day of the year, and perhaps also the nones, from coinciding with the nundinae, of which mention has already been made (Macrob. I.13). Hence in Livy (XLV.44), Intercalatum eo anno; postridie Terminalia intercalares fuerunt. This would not have been said had the day of intercalation been invariably the same; and again Livy (XLIII.11), Hoc anno intercalatum est. Tertio die post Terminalia intercalares fuere, i.e. two days after the Terminalia, so that the dies intercalaris was on this occasion inserted, as well as the month so called. Nay, even after the reformation of the calendar, the same superstitious practice remained. Thus, in the year 40 B.C., a day was inserted for this purpose, and afterwards an omission of a day took place, that the calendar might not be disturbed (Dion Cass. XLVIII.33).

The system of intercalating in alternate years 22 or 23 days, that is ninety days in eight years, was borrowed, we are told by Macrobius, from the Greeks; and the assertion is probable enough, first, because from the Greeks the Romans generally derived all scientific assistance; and secondly, because the decemviral legislation was avowedly drawn from that quarter. Moreover, at the very period in question, a cycle of eight years appears to have been in use at Athens, for the Metonic period of 19 years was not adopted before 432 B.C. The Romans, however, seem to have been guilty of some clumsiness in applying the science they derived from Greece. The addition of ninety days in a cycle of eight years to a lunar year of 354 days, would, in substance, have amounted to the addition of 11¼ (= 90÷8) days to each year, so that the Romans would virtually have possessed the Julian calendar. As it was, they added the intercalation to a year of 355 days; and consequently, on an average, every year exceeded its proper length by a day, if we neglect the inaccuracies of the Julian calendar. Accordingly we find that the civil and solar years were greatly at variance in the year 564 A.U.C. On the 11th of Quinctilis, in that year, a remarkable eclipse of the sun occurred (Liv. XXXVII.4). This eclipse, says Ideler, can have been no other than the one which occurred on the 14th of March, 190 B.C. of the Julian calendar, and which at Rome was nearly total. Again, the same historian (Liv. XLIV.37) mentions an eclipse of the moon which occurred in the night between the 3rd and 4th of September, in the year of the city 586. This must have been the total eclipse in the night between the 21st and 22nd of June, 168 B.C.a

That attempts at legislation for the purpose of correcting so serious an error were actually made, appears from Macrobius, who, aware himself of the cause of the error, says that, by way of correction, in every third octoennial period, instead of 90 intercalary days, only 66 were inserted. Again it appears that M'. Acilius Glabrio, in his consul­ship 169 B.C., that is, the very year before that in which the above-mentioned lunar eclipse occurred, introduced some legislative measure upon the subject of intercalation (Macrob. I.13). According to the above statement of Macrobius, a cycle of 24 years was adopted, and it is this very passage which has induced the editors of Livy to insert the word quarto in the text already quoted.

As the festivals of the Romans were for the most part dependent upon the calendar, the regulation of the latter was intrusted to the college of pontifices, who in early times were chosen exclusively from the body of patricians. It was therefore in the power of the college to add to their other means of oppressing the plebeians, by keeping to themselves the knowledge of the days on which justice could be administered, and assemblies of the people could be held. In the year 304 B.C., one Cn. Flavius, a secretary (scriba) of Appius Claudius, is said fraudulently to have made the Fasti public (Liv. XI.46; Cic. Pro Murena, c. 11; Plin. H. N. XXXIII.1; Val. Max. II.5; A. Gellius, VI.9; Macrob. I.15; Pomponius, De Origine Juris in the Digest 1 2; and Cicero, ad Att. VI.1). It appears however from the last passage that Atticus doubted the truth of the story. In either case, the other privilege of regulating the year by the insertion of the intercalary month gave them great political power, which they were not backward to employ. Every thing connected with the matter of intercalation was left, says Censorinus (c. 20), to the unrestrained pleasure of the pontifices; and the majority of these, on personal grounds, added to or took from the year by capricious intercalations, so as to lengthen or shorten the period during which a magistrate remained in office, and seriously to benefit or injure the farmer of the public revenue. Similar to this is the language employed by Macrobius (I.4), Ammianus (XXVI.1), Solinus (c. I), Plutarch (Caes. c. 59), and their assertions are confirmed by the letters of Cicero, written during his proconsulate in Cilicia, the constant burthen of which is a request that the pontifices will not add to his year of government by intercalation.

In consequence of this licence, says Suetonius (Caes. 40), neither the festivals of the harvest coincided with the summer, nor those of the vintage with the autumn. But we cannot desire a better proof of the confusion than a comparison of three short passages in the third book of Caesar's Bell. Civ. (c. 6), Pridie nonas Januarias navis solvit — (c. 9) jamque hiems adpropinquabat — (c. 25) multi jam menses transierant et hiems jam praecipitaverat.

Year of Julius Caesar.— In the year 46 B.C. Caesar, now master of the Roman world, crowned  p231 his other great services to his country by employing his authority, as pontifex maximus, in the correction of this serious evil. For this purpose he availed himself of the services of Sosigenes, the peripatetic, and a scriba named M. Flavius, though he himself too, we are told, was well acquainted with astronomy, and indeed was the author of a work of some merit upon the subject, which was still extant in the time of Pliny. The chief authorities upon the subject of the Julian reformation are Plutarch (Caes. c. 59), Dion Cassius (XLIII.26), Appian (De Bell. Civ. II. ad extr.), Ovid (Fasti, III.155), Suetonius (Caes. c. 40), Pliny (H. N. XVIII.57), Censorinus (c. 20), Macrobius (Sat. I.14), Ammianus Marcellinus (XXVI.1), Solinus (I.45). Of these Censorinus is the most precise:— "The confusion was at last," says he, "carried so far that C. Caesar, the pontifex maximus, in his third consulate, with Lepidus for his colleague, inserted between November and December two intercalary months of 67 days, the month of February having already received an intercalation of 23 days, and thus made the whole year to consist of 445 days. At the same time he provided against a repetition of similar errors by casting aside the intercalary month, and adapting the year to the sun's course. Accordingly, to the 355 days of the previously existing year, he added ten days, which he so distributed between the seven months having 29 days, that January, Sextilis, and December received two each, the others but one; and these additional days he placed at the end of the several months, no doubt with the wish not to remove the various festivals from those positions in the several months which they had so long occupied. Hence in the present calendar, although there are seven months of 31 days, yet the four months, which from the first possessed that number, are still distinguishable by having their nones on the fifth of the month. Lastly, in consideration of the quarter of a day, which he considered as completing the true year, he established the rule that, at the end of every four years, a single day should be intercalated, where the month had been hitherto inserted, that is, immediately after the Terminalia; which day is now called the Bissextum."

This year of 445 days is commonly called by chronologists​b the year of confusion; but by Macrobius, more fitly, the last year of confusion. The kalends of January, of the year 708 A.U.C., fell on the 13th of October, 47 B.C. of the Julian calendar; the kalends of March, 708 A.U.C., on the 1st of January, 46 B.C.; and lastly, the kalends of January, 709 A.U.C., on the 1st of January, 45 B.C. Of the second of the two intercalary months inserted in this year after November, mention is made in Cicero's letters (ad Fam. VI.14).

It was probably the original intention of Caesar to commence the year with the shortest day. The winter solstice at Rome, in the year 46 B.C., occurred on the 24th of December of the Julian calendar. His motive for delaying the commencement for seven days longer, instead of taking the following day, was probably the desire to gratify the superstition of the Romans, by causing the first year of the reformed calendar to fall on the day of the new moon. Accordingly, it is found that the mean new moon occurred at Rome on the 1st of January, 45 B.C., at 6h. 16′ P. M. In this way alone can be explained the phrase used by Macrobius: Annum civilem Caesar, habitis ad lunam dimensionibus constitutum, edicto palam proposito publicavit. This edict is also mentioned by Plutarch where he gives the anecdote of Cicero, who, on being told by some one that the constellation Lyra would rise the next morning, observed, "Yes, no doubt, in obedience to the edict."

The mode of denoting the days of the month will cause no difficulty, if it be recollected, that the kalends always denote the first of the month, that the nones occur on the seventh of the four months March, May, Quinctilis or July, and October, and on the fifth of the other months; that the ides always fall eight days later than the nones; and lastly, that the intermediate days are in all cases reckoned backwards upon the Roman principle already explained of counting both extremes.

For the month of January the notation will be as follows:—

14 a. d. XIX. Kal. Feb.
15 a. d. XVIII. Kal. Feb.
16 a. d. XVII. Kal. Feb.
17 a. d. XVI. Kal. Feb.
18 a. d. XV. Kal. Feb.
19 a. d. XIV. Kal. Feb.
20 a. d. XIII. Kal. Feb.
21 a. d. XII. Kal. Feb.
22 a. d. XI. Kal. Feb.
23 a. d. X. Kal. Feb.
24 a. d. IX. Kal. Feb.
6 a. d. VIII. Id. Jan. 25 a. d. VIII. Kal. Feb.
7 a. d. VII. Id. Jan. 26 a. d. VII. Kal. Feb.
8 a. d. VI. Id. Jan. 27 a. d. VI. Kal. Feb.
9 a. d. V. Id. Jan. 28 a. d. V. Kal. Feb.
2 a. d. IV. Non. Jan. 10 a. d. IV. Id. Jan. 29 a. d. IV. Kal. Feb.
3 a. d. III. Non. Jan. 11 a. d. III. Id. Jan. 30 a. d. III. Kal. Feb.
4 Prid. Non. Jan. 12 Prid. Id. Jan. 31 Prid. Kal. Feb.
1 Kal. Jan. 5 Non. Jan. 13 Id. Jan.

The letters a d are often, through error, written together, and so confounded with the preposition ad, which would have a different meaning, for ad kalendas would signify by, i.e. on or before the kalends. The letters are in fact an abridgement of ante diem, and the full phrase for "on the second of January" would be ante diem quartum nonas Januarias. The word ante in this expression seems really to belong in sense to nonas, and to be the cause why nonas is an accusative. Hence occur such phrases as (Cic. Phil. III.8), in ante diem quartum Kal. Decembris distulit, "he put if off to the fourth day before the kalends of December," (Caes. Bell. Gall. I.6) Is dies erat ante diem V. Kal. Apr., and (Caes. Bell. Civ. I.11) ante quem diem iturus sit, for quo die. The same confusion exists in the phrase post paucos dies, which means "a few days after," and is equivalent to paucis post diebus. Whether the phrase Kalendae Januarii was ever used by the best writers is doubtful. The words are commonly abbreviated; and those passages where Aprilis, Decembris, &c. occur, are of no avail, as they are probably accusatives. The ante may be omitted, in which case the phrase will be die quarto nonarum. In the leap year (to use a modern phrase), the last days of February were called —

Feb. 23 a. d. VII. Kal. Mart.
Feb. 24 a. d. VI. Kal. Mart. posteriorem
Feb. 25 a. d. VI. Kal. Mart. priorem
Feb. 26 a. d. V. Kal. Mart.
Feb. 27 a. d. IV. Kal. Mart.
Feb. 28 a. d. III. Kal. Mart.
Feb. 29 Prid. Kal. Mart.

In which the words prior and posterior are used in  p232 reference to the retrograde direction of the reckoning. Such at least is the opinion of Ideler, who refers to Celsus in the Digest (50 tit. 16 s.98).

From the fact that the intercalated year has two days called ante diem sextum, the name of bissextile has been applied to it. The term annus bissextilis, however, does not occur in any writer prior to Beda, but in place of it the phrase annus bissextus.

It was the intention of Caesar that the bissextum should be inserted peracto quadriennii circuitu, as Censorinus says, or quinto quoque incipiente anno, to use the words of Macrobius. The phrase, however, which Caesar used seems to have been quarto quoque anno, which was interpreted by the priests to mean every third year. The consequence was, that in the year 8 B.C. the Emperor Augustus, finding that three more intercalations had been made than was the intention of the law, gave directions that for the next twelve years there should be no bissextile.​c

The services which Caesar and Augustus had conferred upon their country by the reformation of the year, seem to have been the immediate causes of the compliments paid to them by the insertion of their names in the calendar. Julius was substituted for Quinctilis, the month in which Caesar was born, in the second Julian year, that is, the year of the dictator's death (Censorinus, c. 22); for the first Julian year was the first year of the corrected Julian calendar, that is, 45 B.C. The name Augustus, in place of Sextilis, was introduced by the emperor himself, at the time when he rectified the error in the mode of intercalating (Suet. Aug. c. 31), anno Augustano XX. The first year of the Augustan era was 27 B.C., viz., that in which he first took the name of Augustus, se VII. et M. Vipsanio Agrippa coss. He was born in September; but gave the preference to the preceding month, for reasons stated in the senatus-consultum, preserved by Macrobius (I.12). "Whereas the Emperor Augustus Caesar, in the month of Sextilis, was first admitted to the consulate, and thrice entered the city in triumph, and in the same month the legions, from the Janiculum, placed themselves under his auspices, and in the same month Egypt was brought under the authority of the Roman people, and in the same month an end was put to the civil wars; and whereas for these reasons the said month is, and has been, most fortunate to this empire, it is hereby decreed by the senate that the said month shall be called Augustus." "A plebiscitum, to the same effect, was passed on the motion of Sextus Pacuvius, tribune of the plebs."

The month of September in like manner received the name of Germanicus from the general so called,​d and the appellation appears to have existed even in the time of Macrobius. Domitian, too, conferred his name upon October; but the old word was restored upon the death of the tyrant.

Our days of the Month. March, May,
July, October
have 31 days.
August, December
have 31 days.
April, June,
September, November
have 30 days.
has 28 days,
and in Leap Year 29.
2. VI. ante Nonas. IV. ante Nonas. IV. ante Nonas. IV. ante Nonas.
3. V. III. III. III.
4. IV. Pridie Nonas. Pridie Nonas. Pridie Nonas.
6. Pridie Nonas. VIII. ante Idus. VIII. ante Idus. VIII. ante Idus.
8. VIII. ante Idus. VI. VI. VI.
9. VII. V. V. V.
10. VI. IV. IV. IV.
11. V. III. III. III.
12. IV. Pridie Idus. Pridie Idus. Pridie Idus.
14. Pridie Idus. XIX. ante Kalendas
(of the month following).
XVIII. ante Kalendas
(of the month following).
XVI. ante Kalendas Martias.
16. XVII. ante Kalendas
(of the month following).
18. XV. XV. XIV. XII.
21. XII. XII. XI. IX.
22. XI. XI. X. VIII.
23. X. X. IX. VII.
24. IX. IX. VIII. VI.
26. VII. VII. VI. IV.
27. VI. VI. V. III.
28. V. V. IV. Pridie Kalendas Martias.
29. IV. IV. III.
30. III. III. Pridie Kalendas
(of the month following).
31. Pridie Kalendas
(of the month following).
Pridie Kalendas
(of the month following).

 p233  The Fasti of Caesar have not come down to us in their entire form. Such fragments as exist may be seen in Gruter's Inscriptiones, or more completely in Foggini's work, Fastorum Anni Romani . . . Reliquiae. See also some papers by Ideler in the Berlin Transactions for 1822 and 1823.

The Gregorian Year. —e The Julian calendar supposes the mean tropical year to be 365d. 6h.; but the year, as we have already seen, exceeds the real amount by 11′12″, the accumulation of which, year after year, caused at last considerable inconvenience. Accordingly, in the year 1582, Pope Gregory the XIII, assisted by Aloysius, Lilius, Christoph. Clavius, Petrus Ciaconius, and others, again reformed the calendar. The ten days by which the year had been unduly retarded were struck out by a regulation that the day after the fourth of October in that year should be called the fifteenth; and it was ordered that, whereas hitherto an intercalary day had been inserted every four years, for the future three such intercalations in the course of four hundred years should be omitted, viz., in those years which are divisible without remainder by 100, but not by 400. Thus, according to the Julian calendar, the years, 1600, 1700, 1800, 1900 and 2000 were to have been bissextile; but, by the regulation of Gregory, the years 1700, 1800, and 1900, were to receive no intercalation, while the years 1600 and 2000 were to be bissextile, as before. The bull which effected this change, was issued Feb. 24, 1582. The fullest account of this correction is to be found in the work of Clavius, entitled Romani Calendarii a Gregorio XIII. P. M. restituti Explicatio. As the Gregorian calendar has only 97 leap-years in a period of 400 years, the mean Gregorian year is (303 × 365 ÷ 97 × 366) ÷ 400, that is 365d. 5h 49′12″, or only 24″ more than the mean tropical year. This difference in 60 years would amount to 24′, and in 60 times 60, or 3600 years, to 24 hours, or a day. Hence the French astronomer, Delambre, has proposed that the years 3600, 7200, 10,800, and all multiples of 3600 should not be leap years. The Gregorian calendar was introduced in the greater part of Italy, as well as in Spain and Portugal, on the day named in the bull. In France, two months after, by an edict of Henry III, the 9th of December was followed by the 20th. The Catholic parts of Switzerland, Germany, and the Low Countries, adopted the correction in 1583, Poland in 1586, Hungary in 1587. The Protestant parts of Europe resisted what they called a Papistical invention for more than a century. At last, in 1700, Protestant Germany, as well as Denmark and Holland, allowed reason to prevail over prejudice; and the Protestant cantons of Switzerland copied their example the following year.

In England the Gregorian calendar was first adopted in 1752, and in Sweden in 1753. In Russia, and those countries which belong to the Greek church, the Julian year, or old style as it is called, still prevails.

In this article free use has been made of Ideler's work Lehrbuch der Chronologie. For other information connected with the Roman measurement of time, see Astronomia; Dies; Horologium; Lustrum; Nundinae; Saeculum.

The Author's Note:

1 Hence there are found attached to the successive days in the old calendars the recurring series of letters A, B, C, D, E, F, G, H, no doubt for the purpose of fixing the nundines in the week of eight days; precisely in the same way in which the first seven letters are still employed in ecclesiastical calendars, to mark the days of the Christian week.

Thayer's Notes:

a Not in the least; astronomical evidence and the testimony of Plutarch are against it. See my note ad loc.

b Smith's article, quite properly, does not go into the whole question of chronology: i.e., how accurate is our dating of the events of Roman history, what is it based on, and how the Romans dated them. It does seem to me appropriate, however, at least to mention it. For an initial orientation to Roman chronology and some of its problems, I recommend this excellent page at Livius.

c The alert reader will have noticed that our article does not give an ancient source for this statement. The reason for that is a very good one: that Augustus corrected the problem in 8 B.C. is a matter of inference. The contorsions involved in reaching this conclusion may be found, stated in full glorious detail, here.

d This is just one of three instances of a month of Germanicus found in the authors. The text we now have of Suetonius says the month was September — the natural target, following on the already renamed July and August — but provides two conflicting explanations, or two separate instances:

Calig. XV.2 states that Caligula named September after his own father Germanicus, the general.

Dom. XIII.3 (backed up by Macrobius, Sat. I.12.36) states that both Germanicus and Domitianus were names given to September and October by Domitian, after himself.

Tacitus gives us the third instance:

AnnalsXVI.12 states that senatorial flatterers gave the name Germanicus to June after one of the names of Nero.

Finally, Cassius Dio tells us (LXXII.15) that Commodus named all twelve months after himself; the collection of names didn't include Germanicus, though.

At any rate, for photographic evidence of the month name Germanicus, see this inscription in Foligno.

e There are many pages online that explain the Gregorian calendar, which is really outside the scope of a site like mine on Roman Antiquity, so you're on your own here — except for this excellent comprehensive site on the Gregorian reform where among many things you can read the Latin text, or a French translation, of Pope Gregory's bull Inter gravissimas. An English translation of the same bull can be found here.

[image ALT: Valid HTML 4.01.]

Page updated: 15 Apr 20