At some point in elementary school, your science teacher probably explained to you that there are 365 days in a year because that's how long it takes for Earth to complete one full rotation¹ around the sun. What they might not have specified², however, is that it's not exactly 365 days—it's actually closer to 365.2421 days.

上小学时，你们的科学老师很可能会向你们解释一年有365天，因为这是地球绕太阳公转一圈的时间。不过，他们可能没有详细说明，地球公转一圈并不是精确的365天，实际上说是365.2421天更准确一些。

So, if we want our calendar year to begin right when Earth begins a new rotation around the sun, we have to account for (roughly) an extra quarter of a day each year, or one day every four years. History.com reports that the Egyptians had already been doing this for a while before Europe finally caught on in 46 BC, when Roman dictator Julius Caesar and astronomer³ Sosigenes put their heads together to come up with what we now call the Julian calendar, which includes 12 months, 365 days, and an additional "leap day" every four years on February 29.

因此，如果我们希望年历的开端正好是地球开始新一轮公转的时候，我们就必须每年多算（大约）四分之一天，或每四年多算一天。历史网报道称，埃及人先采用了这种算法，后来到了公元前46年欧洲人才开始这么算。当时罗马独裁者尤利乌斯·恺撒（恺撒大帝）和天文学家索西琴尼商讨后制定了我们今天所谓的儒略历，儒略历共有12个月、365天，每四年加上一个闰日（2月29日）。

But rounding 0.2421 up to 0.25 each year created an issue, because it didn't quite add up to a full day every four years—and that tiny discrepancy⁴ meant that after 128 years, the calendar year ended up starting a day before Earth had completed its rotation around the sun. By the 14th century, the calendar year was starting a whopping 10 days before Earth finished its orbit.

但是把每年的0.2421天约等于0.25天也产生了一个问题，因为按照0.2421算的话，每四年要增加的并不是一整天。这个微小的差额意味着128年后，年历就会在地球完成公转前一天开始。到了14世纪，年历开始的时间比地球完成公转的时间早了10天之多。

In 1582, Pope Gregory XIII sought to correct the error by suggesting that we simply skip a leap day every so often. His Gregorian calendar, which we still use today, mandates⁵ that we omit the leap day during years evenly divisible by 100 but not by 400. For instance, the year 2000 included a leap day because it's divisible by 100 and 400; the year 2100, on the other hand, will not include a leap day, since it's evenly divisible by 100, but not by 400.

1582年，教皇格列高利十三世想要改正这一错误，于是建议我们时不时跳过一个闰日。他的格里历（我们今天仍在使用的公历）规定，在能够被100整除但不能被400整除的年份跳过闰日。举例来说，2000年有一个闰日，因为2000能被100和400整除；而2100年没有闰日，因为它能被100整除，但不能被400整除。

Gregory XIII's correction to Caesar's overcorrection is itself a bit of an under-correction, so we'll probably need to reevaluate our leap day protocol⁶ again in about 10,000 years.

教皇格列高利十三世对恺撒大帝过度校正的更正本身是一种校正不足，所以大约1万年后，我们将很可能需要对我们的闰日规则进行重新评估。