When in about 47 B.C. Julius Caesar asked the astronomer, Sosigenes, to devise a new calendar for the Roman world, he had returned from Egypt where he had been favorably impressed by the calendar there in use. This calendar provided for 12 months of 30 days each, for a total of 360, with 5 days added at the end of the year. Although the Egyptian calendar was slipping out of sync with the solar year (and the seasons), it was a vast improvement over the Roman calendar then in effect. The latter attempted to divide the solar year into lunar months, a hopeless task, and as a result the calendar of Rome was about 90 days adrift from the solar year, despite many adjustments.
With the help of Sosigenes, Caesar adopted a calendar in 46 B.C. for the Roman Empire that effected two principal changes. First, he instituted leap year every four years to accommodate the fact that the solar year is approximately 365 1/4 days in length. Second, he spread the extra five days (six in leap year) among the months, although this distribution did not exactly match the system of days and months that is in effect today.
Julius Caesar did two other things with the Roman calendar. Because the calendar in 46 B.C. was about 90 days out of phase with the seasons, he ordained one “catch-up” adjustment, and the year 46 B.C. had 455 days. It was known as “the year of confusion.” He also moved the beginning of the year from March 1 to January 1. Julius Caesar was assassinated shortly thereafter, and in 44 B.C. the month then known as Quintilis (5th) was changed to Julius (July) to honor him. His calendar is known today as the Julian Calendar.
Augustus Caesar followed Julius as emperor, and in 8 B.C. the month Sextilis (6th) was renamed Augustus (August) to honor him, and it was at that time that the numbers of days in the respective months were reordered to what they are now, e.g., 31 days in July, and 31 days in August.
Under the reign of Emperor Constantine in 321 A.D., the Roman Empire adopted the seven-day week. Previous to that time, each month had been divided into market-day intervals of seven days or eight days each, so that each month contained four such intervals. A 30-day month, for example, would have two 7s and two 8s. It was probably about this time that the calendar generally observed reverted to a March 25 to March 24 year instead of a January to December year; during the Middle Ages most European countries used either March 1 or March 25 as the first day of the calendar year.
During the time of the Julian Calendar, however, another problem was exacerbating confusion over dates. The policy of having a leap year every four years is based on a year of 365 1/4 days, or 365 days, 6 hours. Actually, the length of the solar year is 365 days, 5 hours, 48 minutes, 46 seconds. So the leap year every fourth year constitutes an overcorrection of 11 minutes, 14 seconds, which adds up to a full day in about 128 years. From a mathematical standpoint, the ideal solution would be to drop the leap year every 32nd time (every 128 years).
In 1582 AD., Pope Gregory XIII devised a formula almost as precise -- and much easier to remember. The new formula for determination of leap year was and is: a leap year occurs in each year divisible evenly by four, except those years evenly divisible by 100. However, if a year is divisible by 400 it is a leap year, notwithstanding the fact that it is divisible by 100. This formula has the effect of omitting a leap year every 133 years, and the difference between this 133 and the theoretically almost exact 128, or 5, will not catch up with us for about 3,000 years, at which time an ad hoc correction of one day will have to be made.
A onetime adjustment of only 10 days (compared with Julius Caesar’s 90 days), corrected the accumulated astronomical time difference up to 1582. The 10 day period was calculated from 325 A.D., when the vernal equinox fell on March 21, rather than from 46 B.C. (370 years earlier) when presumably the vernal equinox occurred a few days later. The 1257 years from 325 to 1582 signified a correction of 9.8 or 10 days, so Pope Gregory [?] decreed that the day following October 4, 1582, would be October 15, 1582. He also restored January 1 as the first day and December 31 as the last day of the year, and this is known as the Gregorian Calendar.
Most Catholic countries immediately adopted the Gregorian Calendar in 1582, but other countries did so at various times over the next few centuries. Great Britain, including the colonies, waited until 1752. By that time, since 1700 had not been a leap year in these areas, the catch-up correction was 11 days instead of 10 -- thus 14 September followed 2 September. Some other countries also were stragglers; for instance, Russia did not adopt the changes until 1918.
The adjustments had three aspects, and the one with really no scientific basis is the one which causes the most difficulties to genealogists. The adjustments were:
1. Small changes were made in the formula determining which years are leap years to approximate more closely the astronomical calendar.
2. Eleven days were dropped from the calendar as a onetime correction to the Julian calendar, and the day following 2 September 1752 was 14 September 1752.
 3. The beginning and end dates of the year were changed to January 1 to December 31, instead of March 25 to March 24.
The third adjustment was made for convenience, but not for any scientific reason, and it is this change which can cause genealogists the most difficulty. We will use an example.
Abraham Gourden and Sarah Hodder were married at Boston, 12 May 1698, by the well known minister, Cotton Mather. Their first child, Elizabeth, was born 27 February 1698.* As every genealogist should know (but may forget), there was no cause for embarrassment in the Gourden and Hodder families; baby Elizabeth was born precisely 9 1/2 months after the wedding. The two dates are in the Julian Calendar, or Old Style. In the Gregorian Calendar, or New Style, the February date would be 1699.
During the period of colonization in America until 1752, then, the first day of the year was considered to be March 25, while on the European continent the year began on January 1. It is obvious from writings available to genealogists that the system caused confusion even in those days. The date 27 February 1698 can be seen written in any one of the following ways:
1. 27 February 1698
2. 27 February 1699
3. 27th 12th mo. 1698
4. 27 February 1698/9
Either of the first two dates (out of context) is ambiguous, since the reader does not know whether the date was written strictly according to the formal English calendar (example #1) or was “clarified” by someone to conform with the January to December year (#2). Example #3 is a style that was used frequently in the 17th and 18th centuries and is still ambiguous. The last item is known as “double-dating.” it is clear, unambiguous, and should be used whenever possible. On the other hand, a date found in original documents should never be “clarified” or interpreted to show double-dating unless the researcher has sufficient evidence for doing so. Otherwise there may be clarifications upon clarifications, creating errors. in the example given earlier in this article, a researcher might decide to copy the date as 27 February 1699, to make it easier for the reader to understand when the child was born. The next transcriber may decide to use a double-dated form, i.e., 1699/1700, erroneously, and the error may be perpetuated by those after him. As to the other two aspects of the Gregorian adjustments, as long as the genealogist questions such dates as 29 February 1800 and 10 September 1752, there should be less confusion.
* “Boston Births, Baptisms, Marriages, and Deaths, 1630-1699” in Report of the Record Commissioners, Vol. 9 (1883, 1 rep. 1978), pp. 241, 243.
Sources: Keith Gordon Irwin, The 365 Days (1963); Donald H. Menzel, “Calendar History” in The World Book Encyclopedia, Vol. (1972); Agnes Kirsopp Michels, The Calendar of the Roman Republic (1967); and Frank Parise, The Book of Calendars (1982)
Editor’s Note: Researchers may also consult “Dates and the Calendar” in Genealogy as Pastime and Profession, by Donald Lines Jacobus (1930, rev. 1968), “Charts and Charting" in Know Your Ancestors, by Ethel W. Williams (19ô0), an “The 1752 Calendar Change” by Paul W. Prindle in TAG 40 October 1964):246-48,41(April 1965):98,for further discussion on this topic. A one paragraph explanation also appears in Shaking Your Family Tree, by Ralph: J. Crandall (1986), on page 103.