We've been doing some practice towards a Spliced Surprise Major project, and at the moment this involves trying to ring a quarter of Preston, Ipswich and Dunster. I agree that this is a strange combination of methods, but all will be revealed eventually. Preston is familiar as one of the difficult methods from Norman Smith's 23-spliced - familiar, that is, in the sense of knowing about it, rather than being experts at ringing it. Ipswich is also a Norman Smith's method. Dunster is better known in its variation with plain hunting at the lead end, which is Deva; this has become fairly popular and is associated with Simon Linford's Project Pickled Egg. It's Bristol above the treble, and Superlative below with plain hunting at the half lead.

We decided to ring a cyclic 7-part, and my computer came up with a number of compositions, including the following two which are intriguingly similar.

1344 Spliced Surprise Major (3m)              1344 Spliced Surprise Major (3m)
S.J.Gay                                       S.J.Gay
         2345678                                       2345678
----------------                              ----------------
Dunster  8674523                              Preston  5738264
Dunster- 2357486                              Preston- 7864523
Ipswich  6485723                              Ipswich  3526478
Ipswich- 2378564                              Ipswich- 7842635
Preston  8634257                              Dunster  5634278
Preston  4567823                              Dunster- 7823456
----------------                              ----------------
7 part                                        7 part

We've tried both compositions a couple of times, but we've settled on the second one, because having a bob attached to every change of method seems to reduce the risk of miscalls (!).

I've been ringing the tenors, and I've found myself doing a lot of coursing - more than I expected, considering that my general expectation is that a cyclic composition would have wild and difficult coursing orders with the tenors all over the place. One thing about cyclic compositions is that all the handbell pairs ring the same work as each other - for example, a part-end of 17823456 means that in the second part, 5-6 ring what 7-8 rang in the first part, and 3-4 ring what 5-6 rang in the first part. So if it's true that there is a significant amount of coursing for 7-8, then the other pairs get it as well, and this is a helpful feature for everyone.

Here's a table of the lead ends, methods, and which pairs are coursing, throughout the composition. Actually it's not all of the lead ends - we can consider the leads in pairs between bobs.

This table immediately explains why we break down in the 5th part! It's the first time that no-one is coursing.

In total each pair rings 20 leads of coursing, which is almost half of the quarter. For 3-4 and 5-6, 16 of these leads are in a continuous block. (The 16 continuous leads of coursing for 7-8 wrap around the beginning and end of the quarter, so they are not experienced in the same way). And for the last 4 leads of the 2nd part, all the pairs are coursing.

Preston is the most difficult method, and it's the only one with leads in which none of the pairs are coursing: at the beginning of the 5th, 6th and 7th parts. This suggests that we should focus our learning on the leads of Preston that we ring in these parts.

These observations raise the question of how much coursing it's possible to get in a cyclic composition. I might return to it in a future article.

Update: we managed to ring the quarter the next time we tried it.