Except as indicated, this page deals with English or British convention.
ISO 8601 describes the nature of the Gregorian Calendar. ISO 8601:2000 assigned 1875 to the signing of the "Convention du mètre" in Paris, thereby only loosely indicating the positioning of the sequence of dates on the sequence of natural days. ISO 8601:2004 assigns 1875-05-20 to the signing of the Convention.
In this respect, the Papal Bull of Pope Gregory XIII only said, in effect, "omit ten dates"; it relied otherwise on the existing positioning. Likewise for the British Calendar (New Style) Act (1750 c.23), which omitted eleven dates.
Dionysius Exiguus positioned the Anno Domini numbering on the sequence of natural years.
Julius Caesar decreed, via the length of the Year of Confusion (B.C. 46; 445 days (432?); annis confusionis), the exact positioning of the year boundaries of the Julian Calendar on the sequence of natural days, though it seems possible that the positioning got nudged by the Augustine month-length changes about four decades later (according to the Calendar FAQ that reform is, anyway, a fabrication).
The positioning of a calendar should be defined either by reference to another calendar (which must itself be well-defined) or by reference to an event for which the moment of occurrence (to within a few hours or better) can be determined whenever it may be required. The latter means an astronomical event defined by orbital mechanics; I suggest the famous Total Solar Eclipse of 1999-08-11.
The date, in Gregorian and other calendars past and present, can change at different times of the Local or legal GMT (=UT) day. Here the times are given in signed form hh:mm with zero being at midnight, either Local or Greenwich. Local times may be affected by Summer Time.
|-??:?? ¤||GMT||Date first appears, in the Far East (South Pacific)|
|-12:00 ¶||GMT||Date appears, in Zone +12, Standard Time|
|-12:00||Local||Royal Navy Date, until change ordered on 1805-10-11|
|-08:21||GMT||Jewish Moon prediction, -0600 @ Jerusalem|
|-06:00||Local ¤||Jewish/Islamic conventions|
|Sunset §||n/a||Islamic/Jewish conventions; also Roman|
Gregorian Date, customary civil usage
GMT Date - by UK Law from 1880
þ Astronomers' GMT Date, from 1925-01-01 GMT
Some transport may operate internally on 03:00-27:00,
so that the working day has a constant date
|+04:30||Local ¤||Transport for London, Oyster day, 00:00/04:30-28:29+|
Broadcasting may operate internally on 06:00-30:00,
so that the working day has a constant date
|+10:00 ¤||GMT||Date finally appears, in the Far West, Zone -10|
|+12:00||GMT||þ Astronomers' GMT Date, before 1925-01-01 GMT|
|+12:00 ¶||GMT||Date would appear, in Zone -12, Standard Time|
§ : Sunset definitions vary
¤ : Summer Time may affect this
¶ : Zones ±12 are not the actual limits
þ : or Astronomical Date (GMAT)
In earlier times, European conventions varied considerably, including starting the day at noon. I know nothing about conventions in regions not mentioned or implied.
Here only Gregorian/Julian based calendars are considered.
|Year Type||First Day of Year|
|Present Civil||January 1st|
|Astronomers', before 1925||January 1.5 GMT (London noon Jan 1st) *|
|English, c12 to 1751||March 25th - Lady Day, Annunciation|
|Present UK Financial||April 6th - Lady Day + 12 sic|
|Old Celtic||November 1st - All-Hallows' Day|
|Roman Catholic Church||First Sunday in Advent (Nov 27th - Dec 3rd)|
|English, c7 to c12||December 25th - Christmas Day|
|Roman Catholic ecclesiastical, to 1910||December 25th - January 1st from 1911|
|British Regnal||Anniversary of the Accession of the Monarch|
|Royal Navy, at sea||December 31st, local noon (to 1805)|
Likewise, conventions varied elsewhere.
See the Calendar FAQ, Sec 2.9, for more.
British legislation is dated by Regnal Year, counting from the day of the Accession of the Monarch.
For a list, see Harvey, via Date and Time Index and Links.
For more on Date scales, see Date Miscellany I and CritDate (ZIP); these have more about some of the scales starting on these dates.
These are Zeroes of Day or Time; in many cases, Day scales are actually defined by Day 1 being the logically-following day. There is advantage in being free to use 0 to represent undefined / null / escape, in scales that do not need to extend back befor Day 1.
Where the origin is not given as GMT (=UT) or local, it merely means that I do not know; it may well be defined somewhere. The Calendar column applies to the date of origin.
|||Chronological JD||day||BC 4713-01-01||local 00:00||Julian|
|J J Scaliger||Julian Period||7980 years||BC 4713-01-01||Julian|
|John Herschel||Julian Day Count||day||BC 4713-01-01||12:00 GMT||Julian|
|JRS||program longcalc||second||AD 0000-03-01||00:00 local||Gregorian|
|Microsoft||.NET Framework Class Lib||100 ns||AD 0001-01-01||00:00 GMT+local||Gregorian|
|Borland||Delphi 1 TDateTime||day||AD 0001-01-00||00:00 local||Gregorian|
|Mike Cowlishaw||Rexx||1 µs||AD 0001-01-01||00:00||Gregorian|
|Rexx Base Format||day||AD 0001-01-00 ?||-||Gregorian|
|?||Paradox||ms||AD 0001-01-01 ?||00:00||Gregorian|
|MS Windows||TTimeStamp||100 ns||AD 0001-01-01||00:00||Gregorian|
|Dr Luigi Lilio||Lilian Day Count||day||AD 1582-10-14||-||Gregorian|
|ANSI||COBOL 85||day||AD 1601-01-00||-||Gregorian|
|Microsoft||Windows File Time||100 ns||AD 1601-01-01 ?||00:00 GMT||Gregorian|
|Microsoft||AD Integer8||100 ns||AD 1601-01-01 ?||00:00 GMT||Gregorian|
|ANSI ?||MUMPS $Horolog ||day||AD 1841-01-00||00:00||Gregorian|
|?||Modified Julian Date||day||AD 1858-11-17||00:00 GMT||Gregorian|
|?||Chronological MJD||day||AD 1858-11-17||00:00 local||Gregorian|
|DEC||VMS||100 ns||AD 1858-11-17||00:00 GMT||Gregorian|
|Lotus||123, Excel, etc.||day||AD 1899-12-30||00:00 local||Gregorian|
|Borland||Delphi 2+ TDateTime||day||AD 1899-12-30||00:00 local||Gregorian|
|ICL||George 3||day||AD 1899-12-31||-||Gregorian|
|IBM||MVS CICS DB2||day||AD 1899-12-31||-||Gregorian|
|IAU ||Dublin Julian Date||day||AD 1900-01-00||12:00 GMT?||Gregorian|
|Net RFC||NTP||second||AD 1900-01-01||00:00 GMT||Gregorian|
|Common Lisp||Universal Time||second||AD 1900-01-01||00:00 GMT||Gregorian|
|Apple||Old Mac Long Time||second||AD 1904-01-01||00:00 GMT||Gregorian|
|Apple||Mac Excel||day||AD 1904-01-01||-||Gregorian|
|Pick||Pick||day||AD 1967-12-31 ?||-||Gregorian|
|IBM||C, etc.||day||AD 1969-12-31||-||Gregorian|
|?||UNIX, C, Perl, etc.||second||AD 1970-01-01||00:00 GMT||Gregorian|
|?||Java, JS||millisecond||AD 1970-01-01||00:00 GMT||Gregorian|
|Apollo (now HP)||OS||||AD 1980-01-01||?||Gregorian|
|USA||GPS||week ||AD 1980-01-06||00:00 GMT||Gregorian|
|Steltor||UNIAPI_TIME||minute||AD 1991-01-01||00:00 GMT||Gregorian|
 Chronological Julian Date :
historians seem to treat Julian Day Number as starting at 0000h local
time on Julian BC 4713-01-01, rather than the proper noon GMT|
 About quarter-seconds, but true unit must be (220µs)/4 = 262144 µs
 No leap-seconds; always 86400 SI secs/day
 ISO 11756, InterSystems Caché Objectscript.
 Dublin 1955
[*] Units : ns = nanoseconds ; µs = microseconds
Astronomers' GMT used a 12-hour offset from legal GMT before 1925
The above are all, more or less, counts of units, although some are often subdivided.
Outlines of astronomy (4th edition) by Sir John F. W. Herschel, 1851, p.635, has further ancient dates.
See also in Wikipedia.
From above, it is possible to drag'n'drop Zero Date & Time together into the Date 1 field. In the calculation, Summer Time is inapplicable and Leap Seconds are ignored; rounding errors were possible.
The dates below are selected solely for their numerical significance.
The standard Julian Date, JD, changes at noon GMT. Modified Julian Date, MJD, is JD-2400000.5, and changes at midnight GMT.
Below, days are to be assumed as GMT 0000h - 2400h unless otherwise indicated. They can be taken to move round the world as daylight does; see Time Zone and Summer Time. Thus some MJD entries below are strictly CMJD.
* Hebrew days start at the 6 p.m. or sunset previous to
the MJD and Julian/Gregorian dates given here.|
~ Islamic Religious days start at sunset.
Dating Creation - Wikipedia
|-2400001||Mon||BC 4714-11-24||BC 4713-01-01||Julian Date Zero (from noon GMT)|
|-2052005||Sat||BC 3761-09-05||BC 3761-10-05||* Hebrew, Day 0|
|-2052003||Mon||BC 3761-09-07||BC 3761-10-07||* Hebrew, 1 Tishri, Year A.M. 1 starts|
|?||?||BC 3760-08-??||BC 3760-09-??||* Hebrew, Start of Creation,|
|-2051649||Fri||BC 3760-08-27||BC 3760-09-26||* Hebrew, Creation of Adam, 8am (local?); A.M. 1 ends|
|-678578||Fri||BC 0001-12-29||BC 0001-12-31||BC Julian ends|
|-678577||Sat||BC 0001-12-30||AD 0001-01-01||AD Julian begins|
|-678576||Sun||BC 0001-12-31||AD 0001-01-02||BC Gregorian ends|
|-678575||Mon||AD 0001-01-01||AD 0001-01-03||AD Gregorian begins|
|-605833||Sat||AD 0200-03-01||AD 0200-03-01||Gregorian and Julian agreement, first day|
|-569310||Wed||AD 0300-02-28||AD 0300-02-28||Gregorian and Julian agreement, last day|
|-451561||Fri||AD 0622-07-19||AD 0622-07-16||~ Moslem, MuHarram 1, 1 A.H. starts|
|-100841||Thu||AD 1582-10-04||Rome: Julian ends|
|-100841||Thu||AD 1582-10-14||Lilian Date Zero|
|-100840||Fri||AD 1582-10-15||Rome: Gregorian begins|
|-38780||Wed||AD 1752-09-02||Britain: Julian ends|
|-38779||Thu||AD 1752-09-14||Britain: Gregorian begins|
|0||Wed||AD 1858-11-17||AD 1858-11-05||Modified Julian Date Zero|
|+15018||Sat||AD 1899-12-30||.||Civil: Delphi 2+|
|+15018||Sat||AD 1899-12-30||.||Civil: PC Excel|
|+15019||Sun||AD 1899-12-31||.||Civil?: George 3|
|+16840||Fri||AD 1904-01-01||.||Civil: Mac Excel|
|+44239||Tue||AD 1980-01-01||.||Civil: MS-DOS internal Day 0|
|+50000||Tue||AD 1995-10-10||AD 1995-09-27||Civil: Certain Birthdays|
|10227×N||Wed||AD xxxx-11-05||Year (28×N+1858)|
|146097×N||Wed||AD xx58-11-17||Year (400×N+1858)|
|Tue||AD xx00-02-29||4/100/400 year rules|
BC dates are standard notation, for which BC 1 is followed by AD 1. Astronomers should subtract 1 from BC years, and replace "BC " with "-". Italic entries are subsidiary.
The tabulated differences change when a date wrongly occurs or is omitted in the "other" calendar. The safe algorithm for converting between Gregorian and Julian dates is by way of the daycount from some day in a year when the calendars were in agreement, AD 201-299.
In intermediate years (the non-Leap century-ending ones), look left for Jan and Feb and right for Mar to Dec.
Note that the difference was 10 days in 1582, when the Gregorian Calendar was first introduced; it was 11 days in 1752, when Britain made the change; and 13 days in 1918, when Russia changed to the Fregorian Calendar.
For the 99 years after the year AD N×100, the difference D is given by something including N - N Div 4. Note that 0 is not special for this Div function. The code using |0 will err when N+85 is negative.