Reform of the Calendar: Mean Value of the Year

Abstract

I SEE there is to be a meeting at Rome in 1922 to consider questions concerning the calendar. I should like to direct attention to the fact, apparently little known—I, at least, have never seen it in any book—that if we make the year equal to 365 218/900 days we get a very good approximation, and one which can be applied by omitting leap-years at certain complete centuries, something like what is being done under the present Gregorian rule. If we say that “a century-year shall be a leap-year only if it gives a remainder of 2 or 7 when divided by 9,” we have a rule which is much more approximate than the Gregorian rule, and one which has been followed de facto since 1582 (year of the Gregorian reform). The new rule would not differ in its application from the Gregorian rule before the year 2400. The Gregorian year, 365 97/400 days, differs from the true tropical year by 26 seconds; if the above modified rule were introduced the difference would be reduced to 2 seconds.