*not*leap years) and then adding them back in every 400 years, so 2000

*was*a leap year. Our leap year in 2000 is something we won't see again until 2400! This gives us a length of the year equal to 365 + 1/4 - 1/100 + 1/400 = 365.2425 days, which means the calendar will be off by one day every 3333 years -- not too shabby!

Calendar reform had a symbiotic relationship with the infant science of astronomy. It's really just a weird coincidence that a simple leap year works so well, but not well enough that it didn't need to be fixed later on. Would science have developed any differently if the year were more exactly a simple fraction of the number of days, so that the Gregorian reform had not been necessary? And what if the opposite had been true? What if the year weren't anywhere close to a simple fraction of the number of days? Would that have led to much greater calendrical confusion over the centuries?

Given all of this confusion, what unit of time do scientists use in their own work? Days? Years? Hours? Minutes? The answer is none of the above.