[r-t] multi-peal extent of major

[Frederick Karl Kepner DuPuy]

Hello, all. For me the notion of a complete extent is one of change
ringing's main pillars; the idea of ringing every possible
permutation once and only once is very satisfying to my OCD-ish mind.
So when I rang my first peal earlier this week, it was very important
to me that the stage be triples — as I see it, that's in a sense the
whole point of a peal!

With that under my belt, I have now rung extents of singles, minimus,
doubles, minor, and triples, more or less in order. The next step,
clearly, is an extent of major! Needless to say, though, ringing a
continuous 40320 isn't a reasonable option (the famous 1963 feat
notwithstanding), so I've been pondering an easier way to get an
extent of major. Why not ring a series of eight peals of major over
the course of a year such that, taken together, they make an extent?

(Before I go any further, let me add that this is all purely
theoretical speculation — I am 29 years old and hope to have many
years of ringing ahead of me. I'm obviously still a mere beginner and
don't have any sort of crazy ambition to overreach myself now; I'm
just daydreaming about something that's doubtless still decades away.)

OK, so what exactly would a multi-peal extent of major look like?
Firstly, each peal should be valid and publishable, and thus should
start and end in rounds. Ideally, however, the first non-rounds row of
each peal should be 'adjacent' to the last non-rounds row of the
previous peal — that is to say, it should be possible to go from the
one row to the next without a jump change. But before getting bogged
down wondering how difficult this would be, there's a bigger worry:
how to keep from repeating rows (apart from rounds).

For of course if you start from rounds, even if you allow yourself
interesting calls (besides ordinary bobs & singles), there's that
whole first lead to worry about! To get around this, I suppose you
could use a different method for each peal, if you can find 8
different methods whose first leads share no rows in common.

And there's another problem: since all peals will share rounds in
common (while it's only needed once for the extent), only one can be a
mere 5040. The others should all by rights be 5041 changes! But odd
(true) lengths are impossible to get with most methods, right? I
suppose it's all fungible, though — there could be seven 5040s and one
5047 (in grandsire or something like that where odd lengths can
occur).

Both for the sake of being able to start and finish without
row-repetition, and to make odd lengths more possible, it occurs to me
that the best solution would be to ring a principle rather than a
plain or treble-dodging method. Principles can come round at any row,
and I understand that sometimes idiosyncratic starts are used too. But
of course 8 is an even number. Are there any popular, simple,
appropriate major principles?

Anyhow, these are my musings. The ideal may be impossible, of course.
If the goal of being able in theory to skip (without jump changes)
between the last non-rounds row of one peal and the first non-rounds
row of the next one is too stringent, it could be relaxed. And as I
say, to avoid repeating any non-rounds rows, the two strategies which
occur to me are a) using 8 different methods; or b) using a principle;
but if neither of these suffices, the restrictions might have to be
further relaxed — perhaps it could be arranged such that no rows
*outside the plain course* occur more than once; but as much of the
plain course (of whatever method) as necessary could appear in each
peal. (The overall length would be longer then, of course.)

I'm probably not the first one to think about all this, though. What
conclusions have been reached on the question before? Has it ever
actually been done?

Many thanks,
Rick DuPuy