This year, 1999, is a "normal" year since an extra day is not inserted at the end of February. In 1996 an extra day was placed in the shortest month of the year. Officially, 29 February is called an "intercalary day", but this day is commonly called "leap day" or by some other name such as Saddie Hawkins Day according to local legend. For many centuries, this day was not considered to have legal status. Why are some years designated as leap years with an extra day, while most years are not?
To answer this question, we will need to consider the role of the calendar. Various calendars have been devised to mark time for practical purposes, to include agriculture, commerce, taxation and religious observances. These calendars attempt to use some recognizable recurring event, such as the return of the sun to its high position in the sky or a full moon. Solar calendars track the periodic movements of the sun across the sky through its annual cycle. The solar year represents the time elapsed between one vernal equinox and the next. The solstices are other important events in the solar calendar, marking the point where the sun's path is either the highest or lowest in the sky. The lunar calendar, from which the month originated, is based on the lunar cycle of recurring phases of the moon. Unfortunately, the two cycles are not commensurate, since twelve complete lunar months of a somewhat nonuniform number of days (because the lunar orbit about the earth is not exactly circular) do not fit the solar year exactly. Furthermore, these cycles do not have periods that fit nicely into an integer number of days - a feature that humans would like to have when constructing a civil calendar with simple whole numbers to identify days. Various early calendars used elaborate adjustment schemes that were unsatisfactory.
In the first century B.C., Julius Caesar decreed calendrical reform with a 365 day year with a sequence of twelve months having essentially the same arrangement as our present calendar. One of his contributions to the calendar scheme involved the inclusion of an extra day to the end of February (the end of the old Roman year) every fourth year. A later requirement was made that the leap year was to occur on those years evenly divisible by four. This quadrennial adjustment was required to realign the calendar with the sun after four years because the time elapsed for the recurrence of the equinox, which defined the solar year, was found to be approximately 365.25 days -- or a quarter of a day more than the normal 365 day year.
However, by the 16th century, astronomers and church clerics had become distressed that the observance of spring equinox was on about 10 March rather than the traditional 21 March. This traditional and desired date represented the date of vernal equinox at the beginning of the Christian era. The astronomers correctly reasoned that a more precise length of the solar year was 365.244 days (or 365 days, 5 hours, 48 minutes and 46 seconds), which was 1 minute and 14 seconds shorter than the 365.25 days used in the Julian calendar. The insertion of the extra day every fourth year was a small overcorrection that produced serious cumulative errors after several centuries.
To reconcile this problem for computing the dates of Easter, the recommendations by the astronomer commissioned by Pope Gregory XIII in 1572 included the requirement that only those centurial years divisible evenly by 400 would be leap years, while the other centurial years (e.g., 1900) would not. So, next year, the year 2000 will mark the first time since 1600 that a centurial year will be a leap year. This Gregorian scheme is still not exact but will result in an error of less than 1 day in 3000 years.
Another feature of the Gregorian calendar reform adopted by Roman Catholic countries was that 10 days in October 1572 were skipped for realignment purposes to ensure that the vernal equinox would fall on 21 March. The British Empire (to include the American colonies) did not convert from the Julian to the Gregorian calendar until 1752, while eastern European countries did not change until after World War I.
The calendrical corrections just described explain why the exact times of when the astronomical seasons begin (on the solstices and equinoxes) do not occur at precisely the same time every year, but undergo a 6 hour drift for three years before they revert to an earlier time. Do the leap day and leap year have any weather significance? Weather records are kept for this day, and because of the uniqueness of the day, any new daily record established on this date is typically heralded. One European legend held that leap years were cold years, probably because of the inclusion of another day during winter. On the other hand, some people would argue that February monthly temperatures at some locations may be slightly warmer than normal, because of the addition of an extra day that would be typically more March-like. In most cases, the variations between leap years and normal years appear to be too small to establish any definitive relationship.