[r-t] 36 Surprise Minor all-the-work

[Richard Smith]

John Warboys wrote:

> I believe the maximum number of methods for an all-the-work regular length
> peal of Surprise Minor 'from the book' was 35 (composition by Richard
> Pearce).

I presume this wasn't in whole courses?  So far as I'm
aware, you can't get 35 in whole courses without using the
new 3-course splice.

I think this table is correct for the number of surprise
methods 'from the book' that you can get atw in whole
courses in n extents.

  extents  |  1   2   3   4   5   6   7   8   9  10  11
  ---------+-------------------------------------------
  methods  |  5  12  16  22  26  31  35  38  39  40  41

Although this makes it look as though there are three
intractible methods that simply have to be tacked on to the
end of the extent, this isn't really true -- Norwich is the
only such method.  One of the other two comes from the
difficulty in knowing how many times to apply the 3-course
cross splice involving Warkworth.

  (6+3) Beverley     <->   York           (3)
    (4) Westminster  <->   Netherseale  (6+0)
    (1) Warkworth    <->   Bacup          (3)

The numbers in paretheses are the number of courses of each
that are needed.  As you have six courses to play with,
either you are short a course of Westminster, or you are
short a course of Bacup.

The other problem is that you want seven courses of
Cambridge (four with Cambridge over and three with Carlisle
over).  This mean you have five wasted courses.

> I have now used Richard Smith's 3-course splice to produce a
> composition containing 36 methods, this can be found on my website:

When I first found this splice, I wondered whether it might
be possible to get 37 methods, atw, into an ordinary length
peal.  Unfortunately I could never quite get it to fit
together.  My starting point was an extent of 35 in whole
courses that had an extent of Stamford (some of which was
crossed into the London backwork to give Wells).  This left
3 spare courses-worth of the Wells underwork which I hoped
to 3-lead splice to give two of London, Lightfoot and
Rosendale (none of which were present in the extent).
There was flexibility as to whether two or three courses had
the London overwork.

This seemed a promising start, and I hoped it would be quite
straightforward to get two extra methods, atw, into the
extent.  Unforunately not.

This morning I wrote a short computer program to
exhaustively check to see whether I could take two (three)
courses of Wells and four (three) courses of Stamford and
use three-lead splices to introduce London and Rosendale,
both atw.  Each of the four methods has five bells ringing
five different place bells giving 100 different pieces of
work by different bells to be included.  The most I could
get was 96.  (I didn't try compositions that rang isolated
leads the wrong way around.  This might buy enough to get
the last four pieces of work, but it would make gluing the
pieces together again very tricky.)

I wonder whether it's possible to get some extra leeway
using Durham/York, Alnwick/Newcastle and Canterbury/Morpeth,
but I can't see how to get a 35-method base composition with
any spare courses of these.

Richard