Anni certus modus apud solos semper Aegyptios fuit. aliarum gentium dispari numero pari errore nutabat, et ut contentus sim referendo paucarum morem regionum, Arcades annum suum tribus mensibus explicabant, Acarnanes sex, Graeci reliqui trecentis quinquaginta quattuor diebus annum proprium computabant. non igitur mirum in hac uarietate Romanos quoque olim auctore Romulo annum suum decem habuisse mensibus ordinatum, qui annus incipiebat a Martio, et conficiebatur diebus trecentis quattuor, ut sex quidem menses, id est Aprilis Iunius Sextilis September Nouember December, tricenum essent dierum, quattuor uero, Martius Maius Quintilis October, tricenis et singulis expedirentur, qui hodieque septimanas habent Nonas, ceteri quintanas. Haec fuit Romuli ordinatio, qui primum anni mensem genitori suo Marti dicauit. quem mensem anni primum fuisse uel ex hoc maxime probatur, quod ab ipso Quintilis quintus est, et deinceps pro numero nominabantur. huius etiam prima die ignem nouum Vestae aris accendebant, ut incipiente anno cura denuo seruandi nouati ignis inciperet. eodem quoque ingrediente mense tam in regia curiisque atque flaminum domibus laureae ueteres nouis laureis mutabantur. eodem quoque mense et publice et priuatim ad Annam Perennam sacrificatum itur, ut annare perennareque commode liceat. hoc mense mercedes exsoluebant magistris, quas completus annus deberi fecit, comitia auspicabantur, uectigalia locabant, et seruis cenas adponebant matronae, ut domini Saturnalibus: illae, ut principio anni ad promptum obsequium honore seruos inuitarent, hi, quia gratiam perfecti operis exsoluerent. Secundum mensem nominauit Aprilem, ut quidam putant cum adspiratione quasi Aphrilem, a spuma quam Graeci ἀφρὸν uocant, unde orta Venus creditur, et hanc Romuli asserunt fuisse rationem, ut primum quidem mensem a patre suo Marte, secundum ab Aeneae matre Venere nominaret, et hi potissimum anni principia seruarent, a quibus esset Romani nominis origo, cum hodieque in sacris Martem patrem, Venerem genetricem uocemus. alii putant Romulum uel altiore prudentia uel certi numinis prouidentia ita primos ordinasse menses ut, cum praecedens Marti esset dicatus deo plerumque hominum necatori, ut Homerus naturae conscius ait Ἆρες Ἄρες βροτολοιγὲ μιαιφόνε τειχεσιβλῆτα secundus Veneri dicaretur, quae uim eius quasi benefica leniret. sed Cingius in eo libro, quem de fastis reliquit, ait imperite quosdam opinari Aprilem mensem antiquos a Venere dixisse, cum nullus dies festus nullumque sacrificium insigne Veneri per hunc mensem a maioribus institutum sit, sed ne in carminibus quidem Saliorum Veneris ulla ut ceterorum caelestium laus celebretur. Cingio etiam Varro consentit adfirmans nomen Veneris ne sub regibus quidem apud Romanos uel Latinum uel Graecum fuisse, et ideo non potuisse mensem a Venere nominari. sed cum fere ante aequinoctium uernum triste sit caelum et nubibus obductum, sed et mare nauigantibus clausum, terrae etiam ipsae aut aqua aut pruina aut niuibus contegantur, eaque omnia uerno, id est hoc mense, aperiantur, arbores quoque nec minus cetera, quae continet terra, aperire se in germen incipiant: ab his omnibus mensem Aprilem dici merito credendum est, quasi Aperilem, sicut apud Athenienses ἀνθεστηριὼν idem mensis uocatur ab eo, quod hoc tempore cuncta florescant. non tamen negat Verrius Flaccus hoc die postea constitutum, ut matronae Veneri sacrum facerent, cuius rei causam, quia huic loco non conuenit, praetereundum est. Maium Romulus tertium posuit. de cuius nomine inter auctores lata dissensio est. nam Fulvius Nobilior in fastis, quos in aede Herculis Musarum posuit, Romulum dicit, postquam populum in maiores iunioresque diuisit, ut altera pars consilio, altera armis rem publicam tueretur, in honorem utriusque partis hunc Maium, sequentem Iunium mensem uocasse. sunt qui hunc mensem ad nostros fastos a Tusculanis transisse commemorent, apud quos nunc quoque uocatur deus Maius, qui est Iuppiter, a magnitudine scilicet ac maiestate dictus. Cingius mensem nominatum putat a Maia, quam Vulcani dicit uxorem, argumentoque utitur, quod flamen Vulcanalis Kalendis Maiis huic deae rem diuinam facit. sed Piso uxorem Vulcani Maiestam non Maiam dicit uocari. contendunt alii Maiam Mercurii matrem mensi nomen dedisse, hinc maxime probantes, quod hoc mense mercatores omnes Maiae pariter Mercurioque sacrificant. adfirmant quidam, quibus Cornelius Labeo consentit, hanc Maiam, cui mense Maio res diuina celebratur, terram esse hoc adeptam nomen a magnitudine, sicut et Mater Magna in sacris uocatur, adsertionemque aestimationis suae etiam hinc colligunt quod sus praegnans ei mactatur, quae hostia propria est terrae. et Mercurium ideo illi in sacris adiungi dicunt, quia uox nascenti homini terrae contactu datur: scimus autem Mercurium uocis et sermonis potentem. Iunius Maium sequitur aut ex parte populi, ut supra diximus, nominatus aut, ut Cingius arbitratur, quod Iunonius apud Latinos ante uocitatus, diuque apud Aricinos Praenestinosque hac appelatione in fastos relatus sit adeo ut, sicut Nisus in commentariis fastorum dicit apud maiores quoque nostros haec appelatio mensis diu manserit, sed post detritis quibusdam litteris ex Iunonio Iunius dictus sit. nam et aedes Iunoni Monetae Kalend. Iuniis dedicata est. non nulli putauerunt Iunium mensem a Iunio Bruto, qui primus Romae consul factus est, nominatum, quod hoc mense, id est Kalendis Iuniis, pulso Tarquinio sacrum Carnae deae in Caelio monte uoti reus fecerit. hanc deam uitalibus humanis praeesse credunt. ab ea denique petitur, ut iecinora et corda, quaeque sunt intrinsecus uiscera, salua conseruet: et quia cordis beneficio, cuius dissimulatione brutus habebatur, idoneus emendationi publici status extitit, hanc deam, quae uitalibus praeest, templo sacrauit. cui pulte fabacia et larido sacrificatur, quod his maxime rebus uires corporis roborentur. nam et Kalendae Iuniae fabariae uulgo uocantur, quia hoc mense adultae fabae diuinis rebus adhibentur. Sequitur Iulius qui, cum secundum Romuli ordinationem, Martio anni tenente principium, Quintilis a numero uocaretur, nihilo minus tamen etiam post praepositos a Numa Ianuarium ac Februarium retinuit nomen, cum non uideretur iam quintus esse sed septimus. sed postea in honorem Iulii Caesaris dictatoris, legem ferente M. Antonio M. filio consule, Iulius appellatus est, quod hoc mense a. d. quartum Idus Quintilis Iulius procreatus sit. Augustus deinde est, qui Sextilis antea uocabatur, donec honori Augusti daretur ex senatus consulto, cuius uerba subieci: cum imperator Caesar Augustus mense Sextili et primum consulatum inierit, et triumphos tres in urbem intulerit, et ex Ianiculo legiones deductae secutaeque sint eius auspicia ac fidem, sed et Aegyptus hoc mense in potestatem populi Romani redacta sit, finisque hoc mense bellis ciuilibus inpositus sit, atque ob has causas hic mensis huic imperio felicissimus sit ac fuerit, placere senatui ut hic mensis Augustus appelletur. item plebiscitum factum ob eandem rem Sexto Pacuuio tribuno plebem rogante. Mensis September principalem sui retinet appellationem: quem Germanici appellatione, Octobrem uero suo nomine Domitianus inuaserat. sed ubi infaustum uocabulum ex omni aere uel saxo placuit eradi, menses quoque usurpatione tyrannicae appellationis exuti sunt: cautio postea principum ceterorum, diri ominis infausta uitantium, mensibus a Septembri usque ad Decembrem prisca nomina reseruauit. Haec fuit a Romulo annua ordinata dimensio qui, sicut iam supra diximus, annum decem mensium, dierum uero quattuor et trecentorum habendum esse constituit, mensesque ita disposuit, ut quattuor ex his tricenos singulos, sex uero tricenos haberent dies. sed cum is numerus neque solis cursui neque lunae rationibus conueniret, non numquam usu ueniebat, ut frigus anni aestiuis mensibus et contra calor hiemalibus proueniret, quod ubi contigisset, tantum dierum sine ullo mensis nomine patiebantur absumi, quantum ad id anni tempus adduceret, quo caeli habitus instanti mensi aptus inueniretur. Sed secutus Numa, quantum sub caelo rudi et saeculo adhuc impolito solo ingenio magistro comprehendere potuit, uel quia Graecorum obseruatione forsan instructus est, quinquaginta dies addidit, ut in trecentos quinquaginta quattuor dies, quibus duodecim lunae cursus confici credidit, annus extenderetur. atque his quinquaginta diebus a se additis adiecit alios sex, retractos illis sex mensibus, qui triginta habebant dies, id est de singulis singulos, factosque quinquaginta et sex dies in duos nouos menses pari ratione diuisit, ac de duobus priorem Ianuarium nuncupauit, primumque anni esse uoluit, tamquam bicipitis dei mensem respicientem ac prospicientem transacti anni finem futurique principia, secundum dicauit Februo deo, qui lustrationum potens creditur. lustrari autem eo mense ciuitatem necesse erat, quo statuit ut iusta dis Manibus soluerentur. Numae ordinationem finitimi mox secuti totidem diebus totidemque mensibus, ut Pompilio placuit, annum suum computare coeperunt, sed hoc solo discrepabant, quod menses undetricenum tricenumque numero alternauerunt. paulo post Numa in honorem imparis numeri, secretum hoc et ante Pythagoram parturiente natura, unum adiecit diem, quem Ianuario dedit, ut tam in anno quam in mensibus singulis praeter unum Februarium impar numerus seruaretur. nam quia duodecim menses, si singuli aut pari aut impari numero putarentur, consummationem parem facerent, unus pari numero institutus uniuersam putationem imparem fecit. Ianuarius igitur Aprilis Iunius Sextilis September Nouember December undetricenis censebantur diebus, et quintanas Nonas habebant, ac post Idus in omnibus a. d. septimum decimum Kalendas computabatur. Martius uero Maius Quintilis et October dies tricenos singulos possidebant. Nonae in his septimanae erant, similiterque post Idus decem et septem dies in singulis usque ad sequentes Kalendas putabantur, sed solus Februarius uiginti et octo retinuit dies, quasi inferis et deminutio et par numerus conueniret. cum ergo Romani ex hac distributione Pompilii ad lunae cursum sicut Graeci annum proprium computarent, necessario et intercalarem mensem instituerunt more Graecorum. omni autem intercalationi mensis Februarius deputatus est, quoniam is ultimus anni erat, quod etiam ipsum de Graecorum imitatione faciebant. nam et illi ultimo anni sui mensi superfluos interserebant dies, ut refert Glaucippus, qui de sacris Atheniensium scripsit. uerum una re a Graecis differebant. nam illi confecto ultimo mense, Romani non confecto Februario sed post uicesimum et tertium eius diem intercalabant, terminalibus scilicet iam peractis. deinde reliquos Februarii mensis dies, qui erant quinque, post intercalationem subiungebant, credo uetere religionis suae more, ut Februarium omni modo Martius consequeretur. Verum fuit tempus, cum propter superstitionem intercalatio omnis omissa est, non numquam uero per gratiam sacerdotum, qui publicanis proferri uel imminui consulto anni dies uolebant, modo auctio modo retractio dierum proueniebat, et sub specie obseruationis emergebat maior confusionis occasio. sed postea C. Caesar omnem hanc inconstantiam temporum uagam adhuc et incertam in ordinem statae definitionis coegit, adnitente sibi M. Flauio scriba, qui scriptos dies singulos ita ad dictatorem retulit, ut et ordo eorum inueniri facillime posset, et inuento certus status perseueraret. ergo C. Caesar exordium nouae ordinationis initurus dies omnes, qui adhuc confusionem poterant facere, consumpsit, eaque re factum est, ut annus confusionis ultimus in quadringentos quadraginta tres dies protenderetur. post hoc imitatus Aegyptios, solos diuinarum rerum omnium conscios ad numerum solis, qui diebus tricenis sexaginta quinque et quadrante cursum conficit, annum dirigere contendit. Iulius ergo Caesar decem dies obseruationi ueteri super adiecit, ut annum trecenti sexaginta quinque dies, quibus sol lustrat zodiacum, efficerent, et ne quadrans deesset, statuit, ut quarto quoque anno sacerdotes, qui curabant mensibus ac diebus, unum intercalarent diem, eo scilicet mense ac loco, quo etiam apud ueteres mensis intercalabatur, id est ante quinque ultimos Februarii mensis dies, idque bissextum censuit nominandum. dies autem decem, quos ab eo additos diximus, hac ordinatione distribuit. in Ianuarium et Sextilem et Decembrem binos dies inseruit, in Aprilem autem Iunium Septembrem Nouembrem singulos, sed neque mensi Februario addidit diem, ne deum inferum religio immutaretur, et Martio Maio Quintili Octobri seruauit pristinum statum, quod satis pleno erant numero, id est dierum singulorum tricenorumque. nam ideo nouos dies circa finem cuiusque mensis inseruit, ubi finem omnium, quae in mense erant, repperit feriarum, adiectosque omnes a se dies fastos notauit, ut maiorem daret actionibus libertatem, et non solum nullum nefastum sed nec comitialem quemquam de adiectis diebus instituit, ne ambitionem magistratuum augeret adiectio. sic annum ciuilem Caesar habitis ad lunam dimensionibus constitutum edicto palam posito publicauit, et huc usque stare potuisset, ni sacerdotes sibi errorem nouum ex ipsa emendatione fecissent. nam cum oporteret diem, qui ex quadrantibus confit, quarto quoque anno confecto, antequam quintus inciperet intercalare, illi quarto non peracto sed incipiente intercalabant. hic error sex et triginta annis permansit, quibus annis intercalati sunt dies duodecim, cum debuerint intercalari nouem. sed hunc quoque errorem sero deprehensum correxit Augustus, qui annos duodecim sine intercalari die transigi iussit, ut illi tres dies, qui per annos triginta et sex uitio sacerdotalis festinationis excreuerant, sequentibus annis duodecim nullo die intercalato deuorarentur. post hoc unum diem secundum ordinationem Caesaris quinto quoque anno incipiente intercalari iussit, et omnem hunc ordinem aereae tabulae ad aeternam custodiam incisione mandauit.

Only the Egyptians have always had a fixed length for the year. Among other peoples the year wavered. Their numbers differed, but they were all equally wrong. To name the custom of just a few places, the Arcadians finished their year in three months and the Acarnanians in six. The rest of the Greeks counted their year as 354 days. So with all this variety it is no surprise that the Romans too once had a year of ten months, set up by Romulus. This year began with March and was 304 days long. Six months, April, June, Sextilis, September, November and December, had 30 days each. The other four, March, May, Quintilis and October, had 31 days each. These four still have their Nones on the seventh day; the rest have them on the fifth. This was Romulus's plan. He dedicated the first month of the year to his father, Mars. The clearest proof that March was the first month is that Quintilis is the fifth month counting from it, and the months after it were named by their numbers. On the first day of this month they also lit a new fire on the altars of Vesta, so that the duty of guarding the renewed fire would start again with the new year. As the month began, the old laurel branches in the Regia, the halls of the curiae and the houses of the flamens were also replaced with fresh ones. In this month, too, people sacrifice to Anna Perenna, both for the city and at home, so that they may get safely through the year and on through many years. In this month they paid teachers their fees, which fell due when the year was over. They took the auspices for the elections and leased out the collection of public revenues. And the mistresses of households served dinner to their slaves, as the masters did at the Saturnalia. The women did it at the start of the year, to win the slaves' willing service by honoring them. The masters did it to thank them for work completed. He named the second month Aprilis. Some think the name was first said with an h, as Aphrilis, from the sea foam that the Greeks call aphros, from which Venus is believed to have been born. They claim that this was Romulus's reasoning. He named the first month after his father, Mars, and the second after Venus, mother of Aeneas. He wanted the gods from whom the Roman people came to hold the start of the year above all. Even today, in our rites, we call Mars "Father" and Venus "Mother." Others think that Romulus, through deeper wisdom or the guidance of some god, arranged the first months so that one would balance the other. The first was dedicated to Mars, a god who most often brings death to men. As Homer, who understood his nature, says, "Ares, Ares, ruin of men, stained with blood, stormer of walls!" So the second was dedicated to Venus, whose kindness might soften his violence. But Cingius, in the book he left on the calendar, says that some people wrongly suppose the ancients named April after Venus. Our ancestors set up no feast day and no special sacrifice for Venus in this month. Even the hymns of the Salii, which praise the other gods of heaven, have no praise of Venus. Varro agrees with Cingius. He states that the Romans did not know the name of Venus, in Latin or in Greek, even in the time of the kings, and so the month could not have been named after her. Before the spring equinox the sky is usually gloomy and covered with clouds. The sea is closed to sailors, and the land itself is covered with water, frost or snow. All of this opens up in spring, that is, in this month. Trees, and everything else the earth holds, begin to open into bud. It is right to believe, then, that April takes its name from all this opening, as if it were Aperilis. In the same way the Athenians call this month Anthesterion, because everything flowers at this time. Verrius Flaccus, however, does not deny that a rule was made later that married women should sacrifice to Venus on this day. The reason does not belong here, so I will pass over it. Romulus put May third. Writers disagree widely about its name. Fulvius Nobilior wrote about it in the calendar he set up in the temple of Hercules of the Muses. He says that Romulus divided the people into the elders (maiores) and the younger men (iuniores). One group was to guard the state with counsel and the other with arms. In honor of each group Romulus called this month Maius and the next Iunius. Some say this month came into our calendar from the people of Tusculum. There Jupiter is still called the god Maius, a name that comes, of course, from his greatness (magnitudo) and majesty (maiestas). Cingius thinks the month was named after Maia, who he says was the wife of Vulcan. As proof he points out that the priest of Vulcan makes a sacrifice to this goddess on the first of May. But Piso says that Vulcan's wife was called Maiesta, not Maia. Others argue that Maia, the mother of Mercury, gave the month its name. Their main proof is that in this month all merchants sacrifice to Maia and Mercury together. Some people claim, and Cornelius Labeo agrees with them, that this Maia, who is worshiped in May, is the Earth. She got the name from her greatness (magnitudo), just as she is also called the Great Mother in our rites. They support their view with another fact: a pregnant sow is sacrificed to her, and that is the offering proper to the Earth. They also say that Mercury is joined with her in worship because a newborn child gets its voice from touching the earth, and we know that Mercury is the lord of voice and speech. June follows May. It was named either after a part of the people, as I said above, or, as Cingius judges, because the Latins once called it Iunonius [after Juno]. The people of Aricia and Praeneste long kept this name in their calendars. As Nisus says in his notes on the calendar, our ancestors too kept this name of the month for a long time, but later some letters wore away and Iunonius became Iunius. Indeed, the temple of Juno Moneta was dedicated on the first of June. Some have thought that June was named after Junius Brutus, who became the first consul of Rome. Their reason is that on the first of June, after Tarquin had been driven out, Brutus made a sacrifice he had vowed to the goddess Carna on the Caelian Hill. People believe that this goddess watches over the vital organs of the human body. People pray to her to keep the liver, the heart and the other organs inside the body healthy. Brutus was thought dull-witted because he hid his good sense. But his heart made him fit to set the state right, so he dedicated a temple to this goddess of the vital organs. She receives offerings of bean porridge and lard, because these foods above all build up the body's strength. And the first of June is commonly called the Bean Kalends, because in this month the ripe beans are used in religious rites. Next comes July. Under Romulus's plan, when March began the year, it was called Quintilis from its number, "fifth." It kept that name even after Numa put January and February in front of it, though it no longer seemed to be the fifth month but the seventh. But later it was named Iulius in honor of Julius Caesar the dictator, under a law proposed by the consul Mark Antony, son of Marcus. The reason was that Julius had been born in this month, on the fourth day before the Ides of Quintilis. August comes next. It was called Sextilis until a decree of the senate gave it in honor of Augustus. Here are the decree's words: "Whereas the emperor Caesar Augustus, in the month of Sextilis, entered on his first consulship and brought three triumphs into the city; and the legions were led down from the Janiculum to follow his command and his cause; and in this month Egypt was brought under the power of the Roman people, and in this month the civil wars were ended; and whereas for these reasons this month is and has been the most fortunate for this empire, the senate resolves that this month shall be called Augustus." A law of the people was also passed for the same purpose, on the motion of Sextus Pacuvius, tribune of the people. September keeps its original name. Domitian had taken it over for his title Germanicus, and October for his own name. But when it was decided to erase that ill-omened name from every bronze and stone, the months too were stripped of the tyrant's names he had seized. Later emperors were careful to avoid anything of such dreadful omen, and their caution kept the old names for the months from September to December. This was the measure of the year that Romulus set up. As I said before, he decided that the year should have ten months and 304 days. He arranged the months so that four of them had 31 days and six had 30. But this number fit neither the course of the sun nor the reckoning of the moon. So it sometimes happened that the cold of the year came in the summer months, and the heat in the winter ones. When that happened, they let as many days as needed go by without any month's name, until the year reached the point where the weather suited the month that was due. Then came Numa, and he added fifty days. He did this as far as he could understand things, in a rough land and an age still unrefined, with only his own mind for a teacher. Or perhaps he had learned from the observations of the Greeks. With these days the year was stretched to 354 days, the time in which he believed the moon completes twelve cycles. To these fifty days he added six more, taken from the six months that had 30 days, one from each. This made 56 days, which he divided equally into two new months. He named the first of the two Ianuarius and wanted it to be the first month of the year. It was the month of the two-headed god, who looks back at the end of the year just past and forward to the beginning of the next. He dedicated the second to the god Februus, who is believed to have power over purifications. The city had to be purified in the month in which he decreed that the rites owed to the spirits of the dead should be paid. The neighboring peoples soon followed Numa's plan. They began to count their year with as many days and as many months as Pompilius had chosen. They differed only in that their months alternated between 29 and 30 days. A little later Numa added one more day in honor of the odd number, a secret that nature revealed even before Pythagoras. He gave the day to January, so that the year and every month except February would have an odd number of days. If all twelve months had an even number of days, or all had an odd number, the total would be even. So one month with an even number made the whole count odd. So January, April, June, Sextilis, September, November and December were counted at 29 days. They had their Nones on the fifth, and in all of them the count after the Ides began at the seventeenth day before the next Kalends. March, May, Quintilis and October had 31 days. Their Nones fell on the seventh, and here too seventeen days were counted after the Ides up to the next Kalends. Only February kept 28 days, as if both a shortening and an even number suited the gods of the underworld. Under this arrangement of Pompilius the Romans counted their year by the course of the moon, as the Greeks did. So they also had to set up an intercalary month, after the Greek custom. Every intercalation was assigned to February, since it was the last month of the year. This too they did in imitation of the Greeks, who also put their extra days into the last month of their year, as Glaucippus reports in his book on the rites of the Athenians. But the Romans differed from the Greeks in one thing. The Greeks added their days once the last month was over. The Romans added theirs before February was over, after its twenty-third day, once the Terminalia had been held. Then they put the remaining five days of February after the extra period. I believe they did this by an old religious custom of theirs, so that March would always follow February. There was a time, though, when superstition led them to leave out the intercalation altogether. At other times the priests, as a favor to the tax contractors, deliberately lengthened or shortened the year. Days were added at one time and taken away at another, and under a show of strict observance the confusion only grew. Later, Gaius Caesar brought all this unsteady reckoning of time, still wandering and uncertain, into a fixed and settled order. He was helped by Marcus Flavius, a scribe, who wrote out each of the days for the dictator. Their order could then be found very easily and, once found, would stay fixed. So when Gaius Caesar was about to start the new order, he used up all the days that could still cause confusion. As a result the last year of confusion was stretched to 443 days. After this he followed the Egyptians, the only people who understand all things divine. He set out to fit the year to the number of the sun, which completes its course in 365 and a quarter days. So Julius Caesar added ten days to the old count, to make a year of 365 days, the time the sun takes to pass through the zodiac. And so that the quarter day would not be lost, he ruled that every fourth year the priests in charge of the months and days should insert one day. It went into the same month and place where the ancients had inserted the extra month, before the last five days of February. He decreed that it should be called the bissextum. He shared out the ten added days like this. He put two days each into January, Sextilis and December, and one each into April, June, September and November. He added no day to February, so as not to disturb the worship of the gods of the underworld. And he left March, May, Quintilis and October as they had been, because their number was already full, at 31 days. He put the new days near the end of each month, after all the month's festivals were over. He marked all the days he added as court days, to give more room for legal business. None of the added days was closed to business, and none was even made an assembly day, so that the extra days would not feed the ambitions of officials. In this way Caesar set up the civil year, with its measurements set in relation to the moon [for the months], and published it in an edict posted for all to see. It could have stood unchanged to this day if the priests had not made a new error out of the reform itself. The day made up of the four quarter days should have been inserted when each fourth year was complete, before the fifth began. Instead, the priests inserted it at the start of the fourth year, not at its end. This error lasted for 36 years. In those years twelve days were inserted, when there should have been nine. This error too was discovered late, and Augustus corrected it. He ordered twelve years to pass without an extra day. In this way the three surplus days, which had piled up over 36 years through the priests' haste, would be used up in the next twelve years with no day inserted. After that he ordered one day to be inserted, as Caesar had arranged, at the start of every fifth year. He had this whole arrangement engraved on a bronze tablet, to keep it safe forever.