No I do not reject your thoughts on the responsiveness of the authorities to
popular calendar traditions (female or male, though I get the point that a
lot of tradition has been lost due to male ignor-ance of female wisdom). The
article, in Archaeoastronomy #13, that I cited, deals with examples of this
responsiveness, documenting how the old pagan mid-quarter-day festivals were
too strong for the church to ignore. They tried to coopt them. The one that
has given them most trouble (Samhain) still lingers on as Halloween
(fittingly, since this is when the Celts honored the living wisdom of their
dead ancestors).

My point about solstices (now that I know that this IS what you were talking
about) and equinoxes, is as follows: (Heh I'm trying to make my sentences sho-
-rter,..and my lines. Tee Hee!)

Solstices and equinoxes divide the year into four roughly equal periods. The
Celtic mid-quarter days are apparently deliberately placed in the middle of
each of these four periods (seasons). The continuance of these mid-quarter
day festivals into the Middle Ages is evidence for a "folk" calendar-lore
which has to be observationally based (vis. McCluskey) and/or rule-based
(as I posit) since these festivals remained more accurate than the Julian
Calendar (astronomically speaking, with respect to ANY measure of the sun's
real, annual, tropical behaviour).

If some of this folk-lore was rule-based (as I propose) then the division of
the year by eight significant days (four festivals and four solar phenomena)
suggests the following:

An old tradition of 8 leap-days* in 33 years (probably handed down by
oral tradition, from a pre-Druidic time when both the summer solstice year
and the vernal equinox year followed this rule) could have been most easily
applied by adding a 366th. day every four and one-eighth years. While our
Roman calendars have always kept the leap-day in February (though there is
ambiguity about which day in February is actually the extra one!), I am
suggesting that a Celtic calendar-keeper could have honored his more
accurate tradition by simply moving his "leap-day"* (or computational
equivalent) through his calendar one eighth of a year every time it was
applied (8 times in 33 years).

Let us say for example that the last occasion for a "leap-day"* (or extra
366th. day) was last Beltane (mid-quarter day between Vernal Equinox and
Summer Solstice), then his next insertion of a leap-day* would occur four
years on, but at the Summer-Solstice (the next in rotation of the eight
significant days). After another four years he could insert the next leap-
day at Lughnasa (the mid-quarter-day halfway between, the Summer-Solstice
and the Fall Equinox). In this way his "leap-day" would make a full rotation
of the calendar in exactly 33 calendar-years (returning, in our example, to
Beltane, thirty-three years on).

Now, my point about an "astronomical" preference for the Vernal equinox is
that, while the summer solstice and the vernal equinox followed this rule
of eight 366th.-days every thirty-three years, around Stonehenge and Old
Kingdom times (the beginnings of rule-based calendars?) the situation has
changed over the millenia. Currently, the Vernal Equinox still follows
this rule, but the Summer Solstice now follows a rule closer to 7 leap
days in 29 years. Moreover, the Summer Solstice year (like the Winter
Solstice year and the Fall Equinox year) will continue to change more
rapidly than the Vernal Equinox year for millenia to come. This is just
how the astronomy works out (see Jean Meeus "Astronomical Algorithms",
1991, chapter 9) but no modern astronomer** has told you, until now!

"You asked and you received"!
 

* Whenever I use the term "leap-day", I am talking about the extra day that
needs to be allowed for, about every four years, in any solar calendar
which cycles through its significant named days with a usual periodicity of
365 days, but which tries to stay in step with some aspect of the sun's real
behaviour with this occasional 366th. day (either using a rule or by result
of some observation of the sun). Examples in current solar calendars are,
our extra day in February, and the 30th. day of Esfahan in the Persian/Iran
calendar. Note that a calendar may not name every day in its year. It may be
sufficient only to know the number of days between significant festivals. I
have never seen evidence that the celts named every SOLAR calendar-day, with
months or numbers like us. Some celts may (like the Old Kingdom Egyptians)
have maintained a simultaneous, independent, lunar calendar with month names
etc.. Please do not confuse such remanants of celtic lunar calendars (e.g.
the Coligny plate) with the evidence of the solar mid-quarter days.

** Even Meeus, only gives formulae for determining the instant, on the
Dynamical time-scale, of each equinox and solstice, but anyone with a
knowledge of calculus can differentiate Meeus' formulae and discover the
truth about the lengths of these four "real tropical years" and how they
are currently varying. I suggest you do this yourself, or have a friend
who is mathematically competent do it, as an independent test of my say-so.
It is not trivial, and requires smoothing over several decades to show
a consistent average. To apply Meeus' formula to times other than the
twentieth century you have to correct for the changing length of the
(Universal) real calendar day. You must therefore add terms (similarily
smoothed) for the changing Delta-T, before differentiating (see Meeus
chapter 9 for several choices of formulae for Delta-T).
 
For a qualitative demonstration (no horrible calculus!) read my essay
on the web at http://serendipity.magnet.ch/cassidy/tropanom.html. and
while you're at it read its parent .............../err_trop.html.

Yours sincerely,  Simon Cassidy, 1053 47th. St. Emeryville Ca. 94608 USA
phone 510-547-0684. copyright 1996 Simon Cassidy all rights reserved.