[r-t] 3-cycle leadheads

[Mark Davies]

I wrote this on r-c but (perhaps not unexpectedly) didn't get any 
response there.

I had forgotten, although it is fairly obvious and presumably 
well-known, that at a stage where the number of working bells is 
non-prime, there will exist sets of courses which share some of the same 
leads but are otherwise different. This occurs because there are 
different ways of reaching the "short-course" leads (those which occur 
in the course at factors of the stage number).

For example the following (treble-dominated Royal) course includes three 
leadheads from the Plain Bob course, but is otherwise different:

   234567890 *
   089234567
   756089234
   423756089 *
   908423756
   675908423
   342675908 *
   890342675
   567890342
   234567890

Has this property ever been used, for instance in a composition of
spliced? I feel sure that the feature must be of use, and should be 
exploitable in some interesting fashion.

MBD