Deep River Astronomy Club
The Origin of Julian Days
By Tom Alburger (Deep River Astronomy Club)
Astronomers, especially variable star observers, commonly use the chronological system of Julian Days to record their observations. Most astronomy texts explain that it is based on the number of days that have elapsed since noon universal time (UT), 1 January 4713 BCE (before current era, or B.C.); for example, 1 January 1996 CE (current era, or A.D.) is equivalent to Julian day 2,450,084.
But why and how did this dating system get started? Who was Julian, anyway? And why was 4713 BCE chosen as the starting date? If you think that "Julian" refers to the Julian calendar, named after the Roman emperor Julius Caesar, you are both right and wrong. The Julian calendar is involved, but only partly. This system was named after Julius Caesar, but not the one who ruled Rome and was assassinated by Brutus and others.
We tend to take our present Gregorian calendar system for granted. We add an extra day every four years to the usual 365, to account for the approximate 365.25 days it takes for Earth to orbit the sun. We divide the year into 12 months, based on the approximate 29.5 days of each lunar cycle. Although the number of days in each month varies from 31 to 30, or even 28, and we sometimes have to resort to the "Thirty days hath September ..." rhyme to remember the days in a particular month, it's a fairly regular system, once you get used to it.
The ancients knew the length of the year fairly accurately, as well as the length of the lunar cycle. Some societies set up calendars based on the moon, or lunar cycle, and some on the sun, or solar cycle. Many combined the two, inserting extra months when needed, to prevent the two cycles from getting out of phase, and to ensure seasons would begin the same time each year. Sometimes this led to confusion, since it was up to the ruler to determine the calendar, based on the advice given by ancient astronomers. When rulers changed, sometimes the calendar did, too.
In ancient Rome before 45 BCE, for example, the civil year was 355 days, with an extra or intercalary month (Intercalaris), inserted after Februarius every two years, but there was no rule to determine the length of this month. Officials before Julius Caesar varied the number of days in it, either for prolonging office terms, speeding up election dates, or through sheer incompetence. By Caesar's time spring equinox was falling in winter.
Sosigenes, the Greek astronomer, advised Julius Caesar on how to correct this problem. So, on 1 January 45 BCE Caesar decreed that the next year would be 445 days long. Contemporaries dubbed it the "year of confusion", although its purpose was to end confusion. The new Julian calendar was based on the solar year of 365 days, with an extra day every fourth year. The intercalary month was eliminated, and every other month, starting with January, would have 31 days. Other months, except February, would have 30 days. February would have 29, and 30 in leap years.
Two years later Julius got knifed, but his calendar ticked along until 7 BCE, when it was interfered with during the reign of Augustus. The fifth month, Quintilis (the year started with the spring equinox, in Martius, or March), which had been renamed Julius (our July) after the deceased Caesar, had 31 days. Augustus, or the Roman Senate, wanted a month named after the emperor, so the next month, Sextilis, with 30 days, was changed to Augustus, and given the same number of days as Julius. The extra day was taken from the final month, Februarius. This alteration left three months together, July, August, and September each with 31 days, contrary to logical alternating 31/30-day months of Sosigenes. To remedy this, September and November lost their 31st days to October and December. That's why we have to remember "thirty days hath September ..."
To make matters worse, after Julius died, the leap year was being added every third, instead of every fourth, year. Augustus put a stop to this by eliminating leap years from about 5 BCE to 8 CE (A.D.).
However, the Julian calendar year was 11 minutes, 4 seconds longer than the solar year. By 1582, this error has accumulated to 10 days, so the spring equinox (and consequently the Christian celebration of Easter), was falling about 11 March. To correct this anomaly Pope Gregory XIII, on the advice of his astronomers, took drastic action and decreed that 15 October 1582 would immediately follow 4 October, thus eliminating 10 days forever. To prevent the calendar from getting out of phase with the seasons again, he decreed that years ending in 00 would be 365 days long unless evenly divisible by 400.
At first, only Catholic countries adopted this Gregorian calendar. Due to anti-Catholic political sentiments, Protestant countries didn't make the change until about 1700, although leading Protestant astronomers, such as Tycho Brahe and Kepler, had earlier approved of the reform. The British Empire held out until 1752; China with its revolution changed in 1912, and Russia with its in 1917. There are institutions that still adhere to the Julian calendar (now 13 days behind the Gregorian), such as in the Orthodox Christian Church.
The years between 1582 and 1752 are still subject to confusion when comparing dates between England and the European continent. For example, the deaths of Shakespeare and Cervantes on the same day, 23 April 1616, did not really happen, for England was still using the Julian calendar. Shakespeare actually outlived the Spanish author by 10 days, dying on 3 May, according to the Gregorian calendar!
All these manipulations of the calendar create complex problems for historians. It might be enough to make them throw up their hands and say "Forget months and years, the only certainty is days!" Joseph Justus Scaliger (1540-1609), a French classical scholar, did just that in 1582, when he invented the Julian period, named after his father, Julius Caesar Scaliger. This was a period of 7,980 years, derived from the product of 28 times 19 times 15.
Why these numbers? Well, 28 refers to the number of years in the Julian calendar it takes for dates to fall again on the same days of the week, the so-called solar cycle. The figure 19 comes from the Metonic cycle of 19 years, devised by Meton of Athens in 432 BCE, although known in China as early as 2260 BCE. The basis of ancient Greek, Jewish, and other calendars, it shows the relationship between the lunar and solar year. In 19 years of exactly 365.25 days each (the Julian, or solar year), there are 235 lunar cycles, with seven of these years having a 13th, or embolistic, month. At the end of the cycle, the phases of the moon recur on a particular day in the solar year. The Metonic cycle was important because it established a lunar calendar having a definite rule for intercalary months, and didn't get out of phase with the cycle of tropical (seasonal) years.
The figure of 15 refers to the ancient Roman cycle of indiction, a 15-year period used for taxation. It was used by Emperor Constantine beginning in 312 CE, and continued not only during the Middle Ages, but was used in the Holy Roman Empire until Napoleon abolished it in 1806.
Why was 4713 BCE used as the starting date for the Julian period? Scaliger chose 12:00 UT, 1 January of that year for Julian day 0.0 because it was the nearest past year when all three cycles, solar, Metonic, and indiction, exactly coincided. The present Julian period will end at 12:00 UT, 31 December 3267.
So, a calendar system established by a classical scholar, to avoid the irregularities of old calendars when calculating differences between historical dates, is now used mainly for scientific purposes by astronomers, for calculating times of maximum and minimum brightness of variable stars. No matter to which purpose it is applied, its simple and direct count of days avoids the complexities of all previous (and possibly future) calendars.
This simplified explanation omits many details of the calendars described above. the following brief and incomplete list of sources will provide more information:
- "Julian period", in Encyclopaedia Britannica, 15th ed., 1987.
- "Calendar", in Encyclopaedia Britannica, 15th ed., 1987.
- "Calendar", in Collier's Encyclopedia, 1994.
- O'Neil, William Matthew. Time and the calendars. -- Sydney, N.S.W. : Sydney University Press, 1975. ISBN 0-424-00003-2.
- Whitrow, G.J. (Gerald James). Time in history : the evolution of our general awareness of time and temporal perspective. -- Oxford: Oxford University Press, 1988. ISBN 0-19-214361-7. LC 88-4245.
Here are some additional Internet sources for Julian day information:
Back to previous page