To prevent similar inaccuracies in the future, the Gregorian calendar provided that once in four centuries the additional day to make a leap-year should be omitted, the date selected for such omission being the last year of every fourth century.

Thus the years 1500, 1900, and 2300, A.D., would not be leap-years.

By this arrangement an approximate rectification of the calendar was effected, though even this does not make it absolutely exact.
Such a rectification as this was obviously desirable, but there was really no necessity for the omission of the ten days from the calendar.
The equinoctial day had shifted so that in the year 1582 it fell on the 10th of March and September.

There was no reason why it should not have remained there.

It would greatly have simplified the task of future historians had Gregory contented himself with providing for the future stability of the calendar without making the needless shift in question.
We are so accustomed to think of the 21st of March and 21st of September as the natural periods of the equinox, that we are likely to forget that these are purely arbitrary dates for which the 10th might have been substituted without any inconvenience or inconsistency.
But the opposition to the new calendar, to which reference has been made, was not based on any such considerations as these.