We all knew that the term "blue moon" had to refer to something that happens only very rarely. So its designation as the second full moon in a month was fairly satisfying. That's bound to be a rare occurrence, since most months encompass only one lunar cycle.
Alas, that widely dispersed definition is apparently wrong, and has been for half a century. The mistake came into view as the editors of Sky and Telescope magazine prepared an article for their March edition that drew on the above "blue moon" definition, first aired in a 1946 article in their publication. The author of that earlier piece had himself referred to a "blue moon" reference in the 1937 Maine Farmers' Almanac.
But an alert co-author of this year's article, rechecking sources, realized the second full moon in a month didn't fit, astronomically, with the '37 almanac reference - or with many others he dug up. There must be another explanation for the term. Further sleuthing and reasoning led to the conclusion the term originally referred to the occurrence of four moons in a season - winter, spring, summer, or fall. Normally, there are only three. And the third moon in a four-moon season is, for reasons best explained by the almanac's editors, the blue moon. The next one, by the way, comes in February of 2000. So what you observed at the end of last month didn't really qualify.
Admittedly, though, the second-full-moon-in-a-month definition is much easier to grasp, and it's been around a good while - it even got enshrined in Trivial Pursuit.
But what the new (make that original) definition has going for it is even greater rarity. And isn't that what "once in a blue moon" is all about?