Retorts

Millennium Countdown

M. Strawson's letter [Retorts, July] delighted me because it chronicled some of the amusing absurdities in our counting and measuring habits for which I had no space in my article.

The Julian Day numbering system of astronomy doesn't need to be related to the Gregorian Calendar at all, or any other calendar. It is used as a calendar in its own right for some astronomical applications and was devised precisely to avoid reference to years and months with their accumulated complications. However, being human, we feel more at home relating to the calendar dates we use in everyday life, so we like to have a Gregorian date for each Julian Day number. As with the Dionysian year numbers, the numbers we apply to the days of the month are counting-number labels not number-line units -- 31 days have not elapsed until the end of the 31st day.

For our predilection, the Astronomical Ephemeris provides tables from which we can calculate the Julian Day number for any calendar day since 1697BC and includes corrections for the Gregorian century-leap-year refinements and the ten days subtracted in 1582 when the Gregorian Calendar was introduced.

These tables assume that AD1 was the first year of the Christian-era calendar, and that BC1 was the year preceding it, which is consistent with the historical evidence about the Dionysian system.

The tables refer to BC0, AD0 and day-zero of each month. But these have nothing to do with the "zero" value for the first unit when measuring a continuum using the number line. They are tabular conventions to ensure accurate calculation when applying additions and subtractions to the base figures given. (If you add 31 days to January 1 you get 32 days, which won't do, so you add it to January 0.)

January 0 is, of course, really December 31 on these tables. And AD0 is really BC1. This can be seen from the fact that the tables yield 1721057 as the Julian Day number for January 0, AD0, which is the same JD number as that for January 0, BC1, and for January 0, BC0. And December 31, AD0 has the same JD number as January 0, AD1 (JD 1721423).

A point of interest is that the Julian Day system measures time from noon instead of midnight. This is a historical convention of astronomers based in Europe to avoid a change of date at midnight in the middle of their working period. Thus at 0 hours Universal Time, 8 July, 1998 the Julian Day number is 2451002.5 (solar days elapsed since noon Universal Time 1 January 4713 BC). As I said in my article, if you want to measure time elapsed from a particular point in time, it doesn't matter where the point is, as long as you define it.

A final point of clarification: the Julian Day numbers of astronomy have nothing to do with the Julian Calendar of Julius Caesar -- the civil calendar in force in Europe for over 1,600 years. Joseph Scaliger named his consecutive day numbering system after his father Julius -- as if we didn't have enough confusion.

Bill Keir, Hokianga