Peterborough junction numbers

[CrackersA361]

Could anyone shed some light on the logic of the apparently random distribution of junction numbers on the roads encircling Peterborough?

Starting on the A1139 and heading anti-clockwise there is a junction 1 and continuing along the south and eastern sides of the city upto junction 7, all seems fine, then there is a roundabout with the Eye Road, then at the north-eastern corner is junction 20 with the A47...

Continuing anti-clockwise along the northern section, the A47, the junction numbers begin to descend from junction 20 to junction 15 in the north-western corner.

Finally the western half of the odd shape ring is the A1260, and heading south, after A47 junction 15, there is a junction 33, followed by 32, and 31, before it meets the A1139 again at junction 3!

What a mess! Was something else planned here or have multiple map makers all made the same error?

[Burwellian]

If you take the A1260 as a spur of the A1139; Junctions 3.1, 3.2 and 3.3; maybe they thought they'd need the 3a, 3b etc on the mainline, so numbered it like so (take the decimal points out, obviously)? In a similar way to how some motorways are numbered as spurs of others (e.g. M275 from the M27, etc)

As for the A47, not part of it's own system at all? Though if that's the case, it seems the rest of it isn't on the ground, nor on maps at all...

[si404]

every junction number is unique. every road has sequential numbering. there are gaps.

[Paul J]

There's a previous thread about it here with a map of the city centre junctions here:Click

[Nicholas]

Junction number locations (as seen on roadsUK):

[Halstead]

I remembered asking the same question two years before but to no avail. One of my raw ideas were to revise a roly-poly shaped numbering system for the GSJ's on the A1139, A47 and A1260, omitting the inferior junctions but still if this happened it wouldn't help at all.