A medieval calendar dating to 813 shows that scholars were already aware that the calendar in use did not precisely match the movements of the Sun. Preserved in Florence, the manuscript records a three-day discrepancy between conventional calendar dates and astronomical events, centuries before the problem was finally addressed by the Gregorian calendar reform.
The calendar is preserved in a sacramentary held by the Laurentian Library (Biblioteca Medicea Laurenziana) in Florence. It was examined by Francesco Vizza, associate researcher emeritus at Italy’s National Research Council (CNR) and former director of its Institute of Chemistry of Organometallic Compounds, with the support of Giuseppe Giari, archivist of the Opera di Santa Maria del Fiore.
According to the researchers, the calendar demonstrates a surprisingly precise understanding of a problem inherent in the Julian calendar. Its compilers recognised that dates used for determining events such as the equinoxes and solstices no longer corresponded exactly with when those astronomical events actually occurred.
“The calendar was found in a sacramentary from the Opera del Duomo, and although it had already been noted in 1757 by the Jesuit astronomer Leonardo Ximenes, its significance for the history of astronomy and timekeeping had never been adequately appreciated,” Vizza explains. “Historians have traditionally maintained that only starting in the 13th century did full awareness of the discrepancy between the Julian calendar and the solar year develop.”
Why was the Julian calendar wrong?
Julius Caesar depicted in the 1170s – The J. Paul Getty Museum, Los Angeles, Ms. 64, fol. 6v
Introduced under Julius Caesar in 46 BC, the Julian calendar was a major improvement in timekeeping. It calculated the year as averaging 365.25 days, achieved through a system of three ordinary 365-day years followed by a leap year containing an additional day.
The problem is that the actual tropical year — essentially the time it takes the seasons to complete their cycle — is slightly shorter, at approximately 365 days, 5 hours and 49 minutes. That makes the Julian calendar year roughly 11 minutes too long.
Eleven minutes sounds insignificant, but the difference accumulates. After about 128 years, the calendar falls approximately one full day out of alignment with the solar year. Over several centuries, astronomical events such as the spring equinox therefore occur progressively earlier according to the calendar.
“Although it represented an extraordinarily effective reform, it was slightly longer than the actual year and thus accumulated an error that, over the centuries, caused the equinoxes and solstices to gradually advance. This discrepancy would only be corrected in 1582 with the Gregorian reform devised by Luigi Lilio, the calendar adopted today by almost all humanity,” continues Vizza.
This growing discrepancy was particularly important in medieval Europe because the calculation of Easter was tied to the spring equinox. Christian tradition placed the equinox on 21 March, following the system associated with the Council of Nicaea in 325. But as the centuries passed, the astronomical equinox was gradually moving away from that conventional date.
The calendar from 813 is particularly significant because it gives both the conventional dates used in ecclesiastical timekeeping and dates corresponding to astronomical phenomena.
The clearest example appears in March. The traditional spring equinox remains listed on 21 March, but the Sun’s entry into Aries, corresponding to the actual astronomical equinox, is recorded on 18 March — three days earlier.
The same three-day advance is recorded for the other equinoxes and solstices.
According to the CNR researchers, these entries demonstrate that whoever compiled the calendar had not merely noticed that something was wrong. They had quantified the accumulated discrepancy with considerable accuracy. The three-day difference is broadly consistent with modern astronomical calculations for the early ninth century.
Awareness that the Julian calendar did not perfectly match the solar year was not entirely unknown in the early Middle Ages. The eighth-century English scholar Bede had recognised that there was a discrepancy, although his estimate of its extent was inaccurate. The astronomer Al-Battani, who lived from around 858 to 929, would later provide a more precise assessment.
The 813 calendar, however, provides evidence that scholars were already recording the difference between conventional calendar dates and observed astronomical phenomena with notable precision during the Carolingian period.
A wider Carolingian network?
The Florence calendar may also shed light on the circulation of astronomical and calendrical knowledge across Carolingian Europe. Researchers note that it predates by about three decades a similar calendar from Prüm, in present-day Germany, dated to around 840. The two calendars display striking similarities.
One possibility is that both ultimately derive from a common source, perhaps the Reichskalender, or “imperial calendar”, reconstructed by historian Arno Borst. Alternatively, their similarities could reflect a broader network through which astronomical and computistical knowledge circulated among intellectual centres within the Carolingian Empire.
Whatever its precise origins, the calendar offers evidence that medieval scholars were closely comparing their inherited system of timekeeping with astronomical observations. The problem they identified would continue to grow for centuries.
By the 16th century, the accumulated discrepancy had reached roughly ten days. The Gregorian reform of 1582 removed those days from the calendar and introduced a modified system of leap years to prevent the same error from accumulating again.
The calendar preserved in Florence suggests that the fundamental problem behind that much later reform had been measured with considerable accuracy as early as 813.
Top Image: Tours. Municipal Library, Ms. 703 fol. 176v
A medieval calendar dating to 813 shows that scholars were already aware that the calendar in use did not precisely match the movements of the Sun. Preserved in Florence, the manuscript records a three-day discrepancy between conventional calendar dates and astronomical events, centuries before the problem was finally addressed by the Gregorian calendar reform.
The calendar is preserved in a sacramentary held by the Laurentian Library (Biblioteca Medicea Laurenziana) in Florence. It was examined by Francesco Vizza, associate researcher emeritus at Italy’s National Research Council (CNR) and former director of its Institute of Chemistry of Organometallic Compounds, with the support of Giuseppe Giari, archivist of the Opera di Santa Maria del Fiore.
According to the researchers, the calendar demonstrates a surprisingly precise understanding of a problem inherent in the Julian calendar. Its compilers recognised that dates used for determining events such as the equinoxes and solstices no longer corresponded exactly with when those astronomical events actually occurred.
“The calendar was found in a sacramentary from the Opera del Duomo, and although it had already been noted in 1757 by the Jesuit astronomer Leonardo Ximenes, its significance for the history of astronomy and timekeeping had never been adequately appreciated,” Vizza explains. “Historians have traditionally maintained that only starting in the 13th century did full awareness of the discrepancy between the Julian calendar and the solar year develop.”
Why was the Julian calendar wrong?
Introduced under Julius Caesar in 46 BC, the Julian calendar was a major improvement in timekeeping. It calculated the year as averaging 365.25 days, achieved through a system of three ordinary 365-day years followed by a leap year containing an additional day.
The problem is that the actual tropical year — essentially the time it takes the seasons to complete their cycle — is slightly shorter, at approximately 365 days, 5 hours and 49 minutes. That makes the Julian calendar year roughly 11 minutes too long.
Eleven minutes sounds insignificant, but the difference accumulates. After about 128 years, the calendar falls approximately one full day out of alignment with the solar year. Over several centuries, astronomical events such as the spring equinox therefore occur progressively earlier according to the calendar.
“Although it represented an extraordinarily effective reform, it was slightly longer than the actual year and thus accumulated an error that, over the centuries, caused the equinoxes and solstices to gradually advance. This discrepancy would only be corrected in 1582 with the Gregorian reform devised by Luigi Lilio, the calendar adopted today by almost all humanity,” continues Vizza.
This growing discrepancy was particularly important in medieval Europe because the calculation of Easter was tied to the spring equinox. Christian tradition placed the equinox on 21 March, following the system associated with the Council of Nicaea in 325. But as the centuries passed, the astronomical equinox was gradually moving away from that conventional date.
A three-day difference in 813
The calendar from 813 is particularly significant because it gives both the conventional dates used in ecclesiastical timekeeping and dates corresponding to astronomical phenomena.
The clearest example appears in March. The traditional spring equinox remains listed on 21 March, but the Sun’s entry into Aries, corresponding to the actual astronomical equinox, is recorded on 18 March — three days earlier.
The same three-day advance is recorded for the other equinoxes and solstices.
According to the CNR researchers, these entries demonstrate that whoever compiled the calendar had not merely noticed that something was wrong. They had quantified the accumulated discrepancy with considerable accuracy. The three-day difference is broadly consistent with modern astronomical calculations for the early ninth century.
Awareness that the Julian calendar did not perfectly match the solar year was not entirely unknown in the early Middle Ages. The eighth-century English scholar Bede had recognised that there was a discrepancy, although his estimate of its extent was inaccurate. The astronomer Al-Battani, who lived from around 858 to 929, would later provide a more precise assessment.
The 813 calendar, however, provides evidence that scholars were already recording the difference between conventional calendar dates and observed astronomical phenomena with notable precision during the Carolingian period.
A wider Carolingian network?
The Florence calendar may also shed light on the circulation of astronomical and calendrical knowledge across Carolingian Europe. Researchers note that it predates by about three decades a similar calendar from Prüm, in present-day Germany, dated to around 840. The two calendars display striking similarities.
One possibility is that both ultimately derive from a common source, perhaps the Reichskalender, or “imperial calendar”, reconstructed by historian Arno Borst. Alternatively, their similarities could reflect a broader network through which astronomical and computistical knowledge circulated among intellectual centres within the Carolingian Empire.
Whatever its precise origins, the calendar offers evidence that medieval scholars were closely comparing their inherited system of timekeeping with astronomical observations. The problem they identified would continue to grow for centuries.
By the 16th century, the accumulated discrepancy had reached roughly ten days. The Gregorian reform of 1582 removed those days from the calendar and introduced a modified system of leap years to prevent the same error from accumulating again.
The calendar preserved in Florence suggests that the fundamental problem behind that much later reform had been measured with considerable accuracy as early as 813.
Top Image: Tours. Municipal Library, Ms. 703 fol. 176v
Subscribe to Medievalverse
Related Posts