[r-t] Yorkshire royal composition

[Mark Davies]

Ander writes,

> Interestingly, my first realization that 65432 was a good part end came 
> from SMC32 (quite a few years ago).  If you ask for lots of runs in 
> certain methods, they pop up very naturally.

Indeed, although this example is reasonably predictable. If you want to 
maximise 2345 and 3456 runs in a two-part, this is the only part-end which 
will produce runs in the second part wherever there is one in the first.

Methods with a degree of falseness can lead to more bizarre part ends. A 
simple example for London S8:

5088 London Surprise Major, arr MBD (SMC32)
  23456   M  B  W  H
  42356            -
  46532   -  3
  35264   -     -
  53462   2        2
3 part
Contains:
  19 xxxx5678
  5 xxxx6578
  10 8765xxxx
  5 5678xxxx
  10 LB5
  61 LB4

This kind of part end can be good for handbells too, but is fairly 
unfamiliar territory for the conductor. Or what about:

5088 London Surprise Major, arr MBD (SMC32)
  23456   M  B  W  H
  26543   -  3
  45362   -     -
  25463   2        3
3 part
Contains:
  19 xxxx5678
  5 xxxx6578
  10 8765xxxx
  5 5678xxxx
  7 LB5
  47 LB4

> An unusual case of a new insight actually coming from a computer search

I find this not so unusual! Searches often come up with interesting 
surprises.

MBD